Pyramid Change History

1.10.8 (2021-02-28)

1.10.7 (2021-02-20)

1.10.6 (2021-02-20)

1.10.5 (2020-11-08)

  • Support Python 3.8 and 3.9.

  • Deprecate the default hashalg='md5' on pyramid.authentication.AuthTktCookieHelper to encourage users to be explicit if they want their code to work without changes when upgrading to Pyramid 2.0.

1.10.4 (2019-04-15)

1.10.3 (2019-04-12)

1.10.2 (2019-01-30)

  • Fix a bug in pyramid.testing.DummySecurityPolicy in which principals_allowed_by_permission would return all principals instead of an empty list if permissive is False. See

  • Fix a bug in which pyramid.exceptions.ConfigurationConflictError may not render the appropriate error message on certain conflicts that were not sortable on Python 3 due to differing types. See

  • Avoid configuring logging in the monitor process using the logging config intended for the application. This avoids opening files for writing in both processes which can cause issues on some systems. See

1.10.1 (2018-11-06)

  • Fix an issue when passing a duck-typed registry object into pyramid.testing.setUp(registry=...) in which the registry wasn't properly fixed prior to invoking actions. See

1.10 (2018-10-31)

  • No major changes from 1.10b1.

1.10b1 (2018-10-28)

1.10a1 (2018-10-15)


Bug Fixes

  • Set appropriate code and title attributes on the HTTPClientError and HTTPServerError exception classes. This prevents inadvertently returning a 520 error code. See

  • Replace webob.acceptparse.MIMEAccept from WebOb with webob.acceptparse.create_accept_header in the HTTP exception handling code. The old MIMEAccept has been deprecated. The new methods follow the RFC's more closely. See

  • Catch extra errors like AttributeError when unpickling "trusted" session cookies with bad pickle data in them. This would occur when sharing a secret between projects that shouldn't actually share session cookies, like when reusing secrets between projects in development. See


  • The pyramid.intefaces.ISession interface will move to require JSON-serializable objects in Pyramid 2.0. See "Upcoming Changes to ISession in Pyramid 2.0" in the "Sessions" chapter of the documentation for more information about this change. See

  • The pyramid.session.signed_serialize and pyramid.session.signed_deserialize functions will be removed in Pyramid 2.0, along with the removal of pyramid.session.UnencryptedCookieSessionFactoryConfig which was deprecated in Pyramid 1.5. Please switch to using the SignedCookieSessionFactory, copying the code, or another session implementation if you're still using these features. See

  • Media ranges are deprecated in the accept argument of pyramid.config.Configurator.add_route. Use a list of explicit media types to add_route to support multiple types.

  • Media ranges are deprecated in the accept argument of pyramid.config.Configurator.add_view. There is no replacement for ranges to add_view, but after much discussion the workflow is fundamentally ambiguous in the face of various client-supplied values for the Accept header. See

Backward Incompatibilities

  • On Python 3.4+ the repoze.lru dependency is dropped. If you were using this package directly in your apps you should make sure that you are depending on it directly within your project. See

  • Remove the permission argument from pyramid.config.Configurator.add_route. This was an argument left over from a feature removed in Pyramid 1.5 and has had no effect since then. See

  • Modify the builtin session implementations to set SameSite='Lax' on cookies. This affects pyramid.session.BaseCookieSessionFactory, pyramid.session.SignedCookieSessionFactory, and pyramid.session.UnencryptedCookieSessionFactoryConfig. See

  • Variables defined in the [pshell] section of the settings will no longer override those set by the setup function. See

  • pyramid.config.Configurator.add_notfound_view uses default redirect class exception pyramid.httpexceptions.HTTPTemporaryRedirect instead of previous pyramid.httpexceptions.HTTPFound. See

  • Removed pyramid.config.Configurator.set_request_property which had been deprecated since Pyramid 1.5. Instead use pyramid.config.Configurator.add_request_method with reify=True or property=True. See

  • Removed the principal keyword argument from which had been deprecated since Pyramid 1.6 and replaced by the userid argument. See

  • Removed the pyramid.tests subpackage that used to contain the Pyramid test suite. These changes also changed the format of the repository to move the code into a src folder. See

Documentation Changes

1.9 (2017-06-26)

  • No major changes from 1.9b1.

  • Updated documentation links for to use HTTPS.

1.9b1 (2017-06-19)

1.9a2 (2017-05-09)

Backward Incompatibilities

  • request.exception and request.exc_info will only be set if the response was generated by the EXCVIEW tween. This is to avoid any confusion where a response was generated elsewhere in the pipeline and not in direct relation to the original exception. If anyone upstream wants to catch and render responses for exceptions they should set request.exception and request.exc_info themselves to indicate the exception that was squashed when generating the response.

    Similar behavior occurs with request.invoke_exception_view in which the exception properties are set to reflect the exception if a response is successfully generated by the method.

    This is a very minor incompatibility. Most tweens right now would give priority to the raised exception and ignore request.exception. This change just improves and clarifies that bookkeeping by trying to be more clear about the relationship between the response and its squashed exception. See and

1.9a1 (2017-05-01)

Major Features

  • The file format used by all p* command line scripts such as pserve and pshell, as well as the pyramid.paster.bootstrap function is now replaceable thanks to a new dependency on plaster.

    For now, Pyramid is still shipping with integrated support for the PasteDeploy INI format by depending on the plaster_pastedeploy binding library. This may change in the future.


  • Added an execution policy hook to the request pipeline. An execution policy has the ability to control creation and execution of the request objects before they enter the rest of the pipeline. This means for a single request environ the policy may create more than one request object.

    The first library to use this feature is pyramid_retry.


  • CSRF support has been refactored out of sessions and into its own independent API in the pyramid.csrf module. It supports a pluggable pyramid.interfaces.ICSRFStoragePolicy which can be used to define your own mechanism for generating and validating CSRF tokens. By default, Pyramid continues to use the pyramid.csrf.LegacySessionCSRFStoragePolicy that uses the request.session.get_csrf_token and request.session.new_csrf_token APIs under the hood to preserve compatibility. Two new policies are shipped as well, pyramid.csrf.SessionCSRFStoragePolicy and pyramid.csrf.CookieCSRFStoragePolicy which will store the CSRF tokens in the session and in a standalone cookie, respectively. The storage policy can be changed by using the new pyramid.config.Configurator.set_csrf_storage_policy config directive.

    CSRF tokens should be used via the new pyramid.csrf.get_csrf_token, pyramid.csrf.new_csrf_token and pyramid.csrf.check_csrf_token APIs in order to continue working if the storage policy is changed. Also, the pyramid.csrf.get_csrf_token function is injected into templates to be used conveniently in UI code.

    See and

Minor Features

  • Support an open_url config setting in the pserve section of the config file. This url is used to open a web browser when pserve --browser is invoked. When this setting is unavailable the pserve script will attempt to guess the port the server is using from the server:<server_name> section of the config file but there is no requirement that the server is being run in this format so it may fail. See

  • The pyramid.config.Configurator can now be used as a context manager which will automatically push/pop threadlocals (similar to config.begin() and config.end()). It will also automatically perform a config.commit() and thus it is only recommended to be used at the top-level of your app. See

  • The threadlocals are now available inside any function invoked via config.include. This means the only config-time code that cannot rely on threadlocals is code executed from non-actions inside the main. This can be alleviated by invoking config.begin() and config.end() appropriately or using the new context manager feature of the configurator. See

Bug Fixes

  • HTTPException's accepts a detail kwarg that may be used to pass additional details to the exception. You may now pass objects so long as they have a valid __str__ method. See

  • Fix a reference cycle causing memory leaks in which the registry would keep a Configurator instance alive even after the configurator was discarded. Another fix was also added for the global_registries object in which the registry was stored in a closure preventing it from being deallocated. See

  • Fix a bug directly invoking pyramid.scripts.pserve.main with the --reload option in which sys.argv is always used in the subprocess instead of the supplied argv. See


  • Pyramid currently depends on plaster_pastedeploy to simplify the transition to plaster by maintaining integrated support for INI files. This dependency on plaster_pastedeploy should be considered subject to Pyramid's deprecation policy and may be removed in the future. Applications should depend on the appropriate plaster binding to satisfy their needs.

  • Retrieving CSRF token from the session has been deprecated in favor of equivalent methods in the pyramid.csrf module. The CSRF methods (ISession.get_csrf_token and ISession.new_csrf_token) are no longer required on the ISession interface except when using the default pyramid.csrf.LegacySessionCSRFStoragePolicy.

    Also, pyramid.session.check_csrf_token is now located at pyramid.csrf.check_csrf_token.

    See and

Documentation Changes

1.8 (2017-01-21)

  • No major changes from 1.8b1.

1.8b1 (2017-01-17)


Documentation Changes

1.8a1 (2016-12-25)

Backward Incompatibilities

  • Support for the IContextURL interface that was deprecated in Pyramid 1.3 has been removed. See

  • Following the Pyramid deprecation period (1.6 -> 1.8), daemon support for pserve has been removed. This includes removing the daemon commands (start, stop, restart, status) as well as the following arguments: --daemon, --pid-file, --log-file, --monitor-restart, --status, --user, --group, --stop-daemon

    To run your server as a daemon you should use a process manager instead of pserve.


  • pcreate is now interactive by default. You will be prompted if a file already exists with different content. Previously if there were similar files it would silently skip them unless you specified --interactive or --overwrite. See

  • Removed undocumented argument cachebust_match from pyramid.static.static_view. This argument was shipped accidentally in Pyramid 1.6. See

  • Change static view to avoid setting the Content-Encoding response header to an encoding guessed using Python's mimetypes module. This was causing clients to decode the content of gzipped files when downloading them. The client would end up with a foo.txt.gz file on disk that was already decoded, thus should really be foo.txt. Also, the Content-Encoding should only have been used if the client itself broadcast support for the encoding via Accept-Encoding request headers. See

  • Settings are no longer accessible as attributes on the settings object (e.g. This was deprecated in Pyramid 1.2. See


  • Python 3.6 compatibility.

  • pcreate learned about --package-name to allow you to create a new project in an existing folder with a different package name than the project name. See

  • The _get_credentials private method of BasicAuthAuthenticationPolicy has been extracted into standalone function extract_http_basic_credentials in pyramid.authentication module, this function extracts HTTP Basic credentials from a request object, and returns them as a named tuple. See

  • Pyramid 1.4 silently dropped a feature of the configurator that has been restored. It's again possible for action discriminators to conflict across different action orders. See

  • pyramid.paster.bootstrap and its sibling pyramid.scripting.prepare can now be used as context managers to automatically invoke the closer and pop threadlocals off of the stack to prevent memory leaks. See

  • Added pyramid.config.Configurator.add_exception_view and the pyramid.view.exception_view_config decorator. It is now possible using these methods or via the new exception_only=True option to add_view to add a view which will only be matched when handling an exception. Previously any exception views were also registered for a traversal context that inherited from the exception class which prevented any exception-only optimizations. See

  • Added the exception_only boolean to pyramid.interfaces.IViewDeriverInfo which can be used by view derivers to determine if they are wrapping a view which only handles exceptions. This means that it is no longer necessary to perform request-time checks for request.exception to determine if the view is handling an exception - the pipeline can be optimized at config-time. See

  • pserve should now work with gevent and other workers that need to monkeypatch the process, assuming the server and / or the app do so as soon as possible before importing the rest of pyramid. See

  • Pyramid no longer copies the settings object passed to the pyramid.config.Configurator(settings=). The original dict is kept. See

  • The csrf trusted origins setting may now be a whitespace-separated list of domains. Previously only a python list was allowed. Also, it can now be set using the PYRAMID_CSRF_TRUSTED_ORIGINS environment variable similar to other settings. See

  • pserve --reload now uses the hupper <> library to monitor file changes. This comes with many improvements:

    • If the watchdog package is installed then monitoring will be done using inotify instead of cpu and disk-intensive polling.

    • The monitor is now a separate process that will not crash and starts up before any of your code.

    • The monitor will not restart the process after a crash until a file is saved.

    • The monitor works on windows.

    • You can now trigger a reload manually from a pyramid view or any other code via hupper.get_reloader().trigger_reload(). Kind of neat.

    • You can trigger a reload by issuing a SIGHUP to the monitor process.


  • A new [pserve] section is supported in your config files with a watch_files key that can configure pserve --reload to monitor custom file paths. See

  • Allow streaming responses to be made from subclasses of pyramid.httpexceptions.HTTPException. Previously the response would be unrolled while testing for a body, making it impossible to stream a response. See

  • Update starter, alchemy and zodb scaffolds to support IPv6 by using the new listen directives in waitress. See

  • All p* scripts now use argparse instead of optparse. This improves their --help output as well as enabling nicer documentation of their options. See

  • Any deferred configuration action registered via config.action may now depend on threadlocal state, such as asset overrides, being active when the action is executed. See

  • Asset specifications for directories passed to config.add_translation_dirs now support overriding the entire asset specification, including the folder name. Previously only the package name was supported and the folder would always need to have the same name. See

  • config.begin() will propagate the current threadlocal request through as long as the registry is the same. For example:

    request = Request.blank(...)
    config.begin(request)  # pushes a request
    config.begin()         # propagates the previous request through unchanged
    assert get_current_request() is request


  • Added a new callback option to config.set_default_csrf_options which can be used to determine per-request whether CSRF checking should be enabled to allow for a mix authentication methods. Only cookie-based methods generally require CSRF checking. See

Bug Fixes

  • Fixed bug in proutes such that it now shows the correct view when a class and attr is involved. See:

  • Fix a FutureWarning in Python 3.5 when using re.split on the format setting to the proutes script. See

  • Fix a RuntimeWarning emitted by WebOb when using arbitrary objects as the userid in the AuthTktAuthenticationPolicy. This is now caught by the policy and the object is serialized as a base64 string to avoid the cryptic warning. Since the userid will be read back as a string on subsequent requests a more useful warning is emitted encouraging you to use a primitive type instead. See

  • Pyramid 1.6 introduced the ability for an action to invoke another action. There was a bug in the way that config.add_view would interact with custom view derivers introduced in Pyramid 1.7 because the view's discriminator cannot be computed until view derivers and view predicates have been created in earlier orders. Invoking an action from another action would trigger an unrolling of the pipeline and would compute discriminators before they were ready. The new behavior respects the order of the action and ensures the discriminators are not computed until dependent actions from previous orders have executed. See

  • Fix bug in i18n where the default domain would always use the Germanic plural style, even if a different plural function is defined in the relevant messages file. See

  • The config.override_asset method now occurs during pyramid.config.PHASE1_CONFIG such that it is ordered to execute before any calls to config.add_translation_dirs. See


Documentation Changes

1.7 (2016-05-19)

1.7b4 (2016-05-12)

  • Fixed the exception view tween to re-raise the original exception if no exception view could be found to handle the exception. This better allows tweens further up the chain to handle exceptions that were left unhandled. Previously they would be converted into a PredicateMismatch exception if predicates failed to allow the view to handle the exception. See

  • Exposed the pyramid.interfaces.IRequestFactory interface to mirror the public pyramid.interfaces.IResponseFactory interface.

1.7b3 (2016-05-10)

  • Fix request.invoke_exception_view to raise an HTTPNotFound exception if no view is matched. Previously None would be returned if no views were matched and a PredicateMismatch would be raised if a view "almost" matched (a view was found matching the context). See

  • Add defaults for py.test configuration and coverage to all three scaffolds, and update documentation accordingly. See

  • Add linkcheck to Makefile for Sphinx. To check the documentation for broken links, use the command make linkcheck SPHINXBUILD=$VENV/bin/sphinx-build. Also removed and fixed dozens of broken external links.

  • Fix the internal runner for scaffold tests to ensure they work with pip and py.test. See

1.7b2 (2016-05-01)

  • Removed inclusion of pyramid_tm in development.ini for alchemy scaffold See

  • A default permission set via config.set_default_permission will no longer be enforced on an exception view. This has been the case for a while with the default exception views (config.add_notfound_view and config.add_forbidden_view), however for any other exception view a developer had to remember to set permission=NO_PERMISSION_REQUIRED or be surprised when things didn't work. It is still possible to force a permission check on an exception view by setting the permission argument manually to config.add_view. This behavior is consistent with the new CSRF features added in the 1.7 series. See

1.7b1 (2016-04-25)

1.7a2 (2016-04-19)


  • Automatic CSRF checks are now disabled by default on exception views. They can be turned back on by setting the appropriate require_csrf option on the view. See

  • The automatic CSRF API was reworked to use a config directive for setting the options. The pyramid.require_default_csrf setting is no longer supported. Instead, a new config.set_default_csrf_options directive has been introduced that allows the developer to specify the default value for require_csrf as well as change the CSRF token, header and safe request methods. The pyramid.csrf_trusted_origins setting is still supported. See

Bug fixes

1.7a1 (2016-04-16)

Backward Incompatibilities

  • Following the Pyramid deprecation period (1.4 -> 1.6), AuthTktAuthenticationPolicy's default hashing algorithm is changing from md5 to sha512. If you are using the authentication policy and need to continue using md5, please explicitly set hashalg to 'md5'.

    This change does mean that any existing auth tickets (and associated cookies) will no longer be valid, and users will no longer be logged in, and have to login to their accounts again.


  • The check_csrf_token function no longer validates a csrf token in the query string of a request. Only headers and request bodies are supported. See


  • Added a new setting, pyramid.require_default_csrf which may be used to turn on CSRF checks globally for every POST request in the application. This should be considered a good default for websites built on Pyramid. It is possible to opt-out of CSRF checks on a per-view basis by setting require_csrf=False on those views. See

  • Added a require_csrf view option which will enforce CSRF checks on any request with an unsafe method as defined by RFC2616. If the CSRF check fails a BadCSRFToken exception will be raised and may be caught by exception views (the default response is a 400 Bad Request). This option should be used in place of the deprecated check_csrf view predicate which would normally result in unexpected 404 Not Found response to the client instead of a catchable exception. See and

  • Added an additional CSRF validation that checks the origin/referrer of a request and makes sure it matches the current request.domain. This particular check is only active when accessing a site over HTTPS as otherwise browsers don't always send the required information. If this additional CSRF validation fails a BadCSRFOrigin exception will be raised and may be caught by exception views (the default response is 400 Bad Request). Additional allowed origins may be configured by setting pyramid.csrf_trusted_origins to a list of domain names (with ports if on a non standard port) to allow. Subdomains are not allowed unless the domain name has been prefixed with a .. See

  • Added a new pyramid.session.check_csrf_origin API for validating the origin or referrer headers against the request's domain. See

  • Pyramid HTTPExceptions will now take into account the best match for the clients Accept header, and depending on what is requested will return text/html, application/json or text/plain. The default for / is still text/html, but if application/json is explicitly mentioned it will now receive a valid JSON response. See

  • A new event and interface (BeforeTraversal) has been introduced that will notify listeners before traversal starts in the router. See and

  • Add a new "view deriver" concept to Pyramid to allow framework authors to inject elements into the standard Pyramid view pipeline and affect all views in an application. This is similar to a decorator except that it has access to options passed to config.add_view and can affect other stages of the pipeline such as the raw response from a view or prior to security checks. See

  • Allow a leading = on the key of the request param predicate. For example, '=abc=1' is equivalent down to request.params['=abc'] == '1'. See

  • A new request.invoke_exception_view(...) method which can be used to invoke an exception view and get back a response. This is useful for rendering an exception view outside of the context of the excview tween where you may need more control over the request. See

  • Allow using variable substitutions like %(LOGGING_LOGGER_ROOT_LEVEL)s for logging sections of the .ini file and populate these variables from the pserve command line -- e.g.: pserve development.ini LOGGING_LOGGER_ROOT_LEVEL=DEBUG See

Documentation Changes

  • A complete overhaul of the docs:

    • Use pip instead of easy_install.

    • Become opinionated by preferring Python 3.4 or greater to simplify installation of Python and its required packaging tools.

    • Use venv for the tool, and virtual environment for the thing created, instead of virtualenv.

    • Use py.test and pytest-cov instead of nose and coverage.

    • Further updates to the scaffolds as well as tutorials and their src files.


  • A complete overhaul of the alchemy scaffold as well as the Wiki2 SQLAlchemy + URLDispatch tutorial to introduce more modern features into the usage of SQLAlchemy with Pyramid and provide a better starting point for new projects. See

Bug Fixes

  • Fix pserve --browser to use the --server-name instead of the app name when selecting a section to use. This was only working for people who had server and app sections with the same name, for example [app:main] and [server:main]. See


1.6 (2016-01-03)


1.6b3 (2015-12-17)

Backward Incompatibilities

  • Remove the cachebust option from config.add_static_view. See config.add_cache_buster for the new way to attach cache busters to static assets. See

  • Modify the pyramid.interfaces.ICacheBuster API to be a simple callable instead of an object with match and pregenerate methods. Cache busters are now focused solely on generation. Matching has been dropped.

    Note this affects usage of pyramid.static.QueryStringCacheBuster and pyramid.static.ManifestCacheBuster.



Bug Fixes

1.6b2 (2015-10-15)


  • Allow asset specifications to be supplied to pyramid.static.ManifestCacheBuster instead of requiring a filesystem path.

1.6b1 (2015-10-15)

Backward Incompatibilities

  • IPython and BPython support have been removed from pshell in the core. To continue using them on Pyramid 1.6+ you must install the binding packages explicitly:

    $ pip install pyramid_ipython
    $ pip install pyramid_bpython
  • Remove default cache busters introduced in 1.6a1 including PathSegmentCacheBuster, PathSegmentMd5CacheBuster, and QueryStringMd5CacheBuster. See


Bug Fixes


  • Deprecate pserve --monitor-restart in favor of user's using a real process manager such as Systemd or Upstart as well as Python-based solutions like Circus and Supervisor. See

1.6a2 (2015-06-30)

Bug Fixes

1.6a1 (2015-04-15)


  • pcreate will now ask for confirmation if invoked with an argument for a project name that already exists or is importable in the current environment. See and

  • Make it possible to subclass pyramid.request.Request and also use pyramid.request.Request.add_request.method. See

  • The pyramid.config.Configurator has grown the ability to allow actions to call other actions during a commit-cycle. This enables much more logic to be placed into actions, such as the ability to invoke other actions or group them for improved conflict detection. We have also exposed and documented the config phases that Pyramid uses in order to further assist in building conforming addons. See

  • Add pyramid.request.apply_request_extensions function which can be used in testing to apply any request extensions configured via config.add_request_method. Previously it was only possible to test the extensions by going through Pyramid's router. See

  • pcreate when run without a scaffold argument will now print information on the missing flag, as well as a list of available scaffolds. See and

  • Added support / testing for 'pypy3' under Tox and Travis. See

  • Automate code coverage metrics across py2 and py3 instead of just py2. See

  • Cache busting for static resources has been added and is available via a new argument to pyramid.config.Configurator.add_static_view: cachebust. Core APIs are shipped for both cache busting via query strings and path segments and may be extended to fit into custom asset pipelines. See and

  • Add pyramid.config.Configurator.root_package attribute and init parameter to assist with includeable packages that wish to resolve resources relative to the package in which the Configurator was created. This is especially useful for addons that need to load asset specs from settings, in which case it is may be natural for a developer to define imports or assets relative to the top-level package. See

  • Added line numbers to the log formatters in the scaffolds to assist with debugging. See

  • Add new HTTP exception objects for status codes 428 Precondition Required, 429 Too Many Requests and 431 Request Header Fields Too Large in pyramid.httpexceptions. See

  • The pshell script will now load a PYTHONSTARTUP file if one is defined in the environment prior to launching the interpreter. See

  • Make it simple to define notfound and forbidden views that wish to use the default exception-response view but with altered predicates and other configuration options. The view argument is now optional in config.add_notfound_view and config.add_forbidden_view.. See

  • Greatly improve the readability of the pcreate shell script output. See

  • Improve robustness to timing attacks in the AuthTktCookieHelper and the SignedCookieSessionFactory classes by using the stdlib's hmac.compare_digest if it is available (such as Python 2.7.7+ and 3.3+). See

  • Assets can now be overidden by an absolute path on the filesystem when using the config.override_asset API. This makes it possible to fully support serving up static content from a mutable directory while still being able to use the request.static_url API and config.add_static_view. Previously it was not possible to use config.add_static_view with an absolute path and generate urls to the content. This change replaces the call, config.add_static_view('/abs/path', 'static'), with config.add_static_view('myapp:static', 'static') and config.override_asset(to_override='myapp:static/', override_with='/abs/path/'). The myapp:static asset spec is completely made up and does not need to exist - it is used for generating urls via request.static_url('myapp:static/foo.png'). See

  • Added pyramid.config.Configurator.set_response_factory and the response_factory keyword argument to the Configurator for defining a factory that will return a custom Response class. See

  • Allow an iterator to be returned from a renderer. Previously it was only possible to return bytes or unicode. See

  • pserve can now take a -b or --browser option to open the server URL in a web browser. See

  • Overall improvements for the proutes command. Added --format and --glob arguments to the command, introduced the method column for displaying available request methods, and improved the view output by showing the module instead of just __repr__. See

  • Support keyword-only arguments and function annotations in views in Python 3. See

  • request.response will no longer be mutated when using the pyramid.renderers.render_to_response() API. It is now necessary to pass in a response= argument to render_to_response if you wish to supply the renderer with a custom response object for it to use. If you do not pass one then a response object will be created using the application's IResponseFactory. Almost all renderers mutate the request.response response object (for example, the JSON renderer sets request.response.content_type to application/json). However, when invoking render_to_response it is not expected that the response object being returned would be the same one used later in the request. The response object returned from render_to_response is now explicitly different from request.response. This does not change the API of a renderer. See

  • The append_slash argument of `Configurator().add_notfound_view() will now accept anything that implements the IResponse interface and will use that as the response class instead of the default HTTPFound. See

Bug Fixes


  • The pserve command's daemonization features have been deprecated. This includes the [start,stop,restart,status] subcommands as well as the --daemon, --stop-server, --pid-file, and --status flags.

    Please use a real process manager in the future instead of relying on the pserve to daemonize itself. Many options exist including your Operating System's services such as Systemd or Upstart, as well as Python-based solutions like Circus and Supervisor.


  • Renamed the principal argument to to userid in order to clarify its intended purpose. See



  • Update scaffold generating machinery to return the version of pyramid and pyramid docs for use in scaffolds. Updated starter, alchemy and zodb templates to have links to correctly versioned documentation and reflect which pyramid was used to generate the scaffold.

  • Removed non-ascii copyright symbol from templates, as this was causing the scaffolds to fail for project generation.

  • You can now run the scaffolding func tests via tox py2-scaffolds and tox py3-scaffolds.

1.5 (2014-04-08)

  • Python 3.4 compatibility.

  • Avoid crash in pserve --reload under Py3k, when iterating over possibly mutated sys.modules.

  • UnencryptedCookieSessionFactoryConfig failed if the secret contained higher order characters. See

  • Fixed a bug in UnencryptedCookieSessionFactoryConfig and SignedCookieSessionFactory where timeout=None would cause a new session to always be created. Also in SignedCookieSessionFactory a reissue_time=None would cause an exception when modifying the session. See

  • Updated docs and scaffolds to keep in step with new 2.0 release of Lingua. This included removing all setup.cfg files from scaffolds and documentation environments.

1.5b1 (2014-02-08)


  • We no longer eagerly clear request.exception and request.exc_info in the exception view tween. This makes it possible to inspect exception information within a finished callback. See

1.5a4 (2014-01-28)


  • Updated scaffolds with new theme, fixed documentation and sample project.

Bug Fixes

  • Depend on a newer version of WebOb so that we pull in some crucial bug-fixes that were showstoppers for functionality in Pyramid.

  • Add a trailing semicolon to the JSONP response. This fixes JavaScript syntax errors for old IE versions. See

  • Fix a memory leak when the configurator's set_request_property method was used or when the configurator's add_request_method method was used with the property=True attribute. See .

1.5a3 (2013-12-10)


  • An authorization API has been added as a method of the request: request.has_permission.

    request.has_permission is a method-based alternative to the API and works exactly the same. The older API is now deprecated.

  • Property API attributes have been added to the request for easier access to authentication data: request.authenticated_userid, request.unauthenticated_userid, and request.effective_principals.

    These are analogues, respectively, of,, and They operate exactly the same, except they are attributes of the request instead of functions accepting a request. They are properties, so they cannot be assigned to. The older function-based APIs are now deprecated.

  • Pyramid's console scripts (pserve, pviews, etc) can now be run directly, allowing custom arguments to be sent to the python interpreter at runtime. For example:

    python -3 -m pyramid.scripts.pserve development.ini
  • Added a specific subclass of HTTPBadRequest named pyramid.exceptions.BadCSRFToken which will now be raised in response to failures in check_csrf_token. See

  • Added a new SignedCookieSessionFactory which is very similar to the UnencryptedCookieSessionFactoryConfig but with a clearer focus on signing content. The custom serializer arguments to this function should only focus on serializing, unlike its predecessor which required the serializer to also perform signing. See . Note that cookies generated using SignedCookieSessionFactory are not compatible with cookies generated using UnencryptedCookieSessionFactory, so existing user session data will be destroyed if you switch to it.

  • Added a new BaseCookieSessionFactory which acts as a generic cookie factory that can be used by framework implementors to create their own session implementations. It provides a reusable API which focuses strictly on providing a dictionary-like object that properly handles renewals, timeouts, and conformance with the ISession API. See

  • The anchor argument to pyramid.request.Request.route_url and pyramid.request.Request.resource_url and their derivatives will now be escaped via URL quoting to ensure minimal conformance. See

  • Allow sending of _query and _anchor options to pyramid.request.Request.static_url when an external URL is being generated. See

  • You can now send a string as the _query argument to pyramid.request.Request.route_url and pyramid.request.Request.resource_url and their derivatives. When a string is sent instead of a list or dictionary. it is URL-quoted however it does not need to be in k=v form. This is useful if you want to be able to use a different query string format than x-www-form-urlencoded. See

  • pyramid.testing.DummyRequest now has a domain attribute to match the new WebOb 1.3 API. Its value is

Bug Fixes

  • Fix the pcreate script so that when the target directory name ends with a slash it does not produce a non-working project directory structure. Previously saying pcreate -s starter /foo/bar/ produced different output than saying pcreate -s starter /foo/bar. The former did not work properly.

  • Fix the principals_allowed_by_permission method of ACLAuthorizationPolicy so it anticipates a callable __acl__ on resources. Previously it did not try to call the __acl__ if it was callable.

  • The pviews script did not work when a url required custom request methods in order to perform traversal. Custom methods and descriptors added via pyramid.config.Configurator.add_request_method will now be present, allowing traversal to continue. See

  • Remove unused renderer argument from Configurator.add_route.

  • Allow the BasicAuthenticationPolicy to work with non-ascii usernames and passwords. The charset is not passed as part of the header and different browsers alternate between UTF-8 and Latin-1, so the policy now attempts to decode with UTF-8 first, and will fallback to Latin-1. See

  • The @view_defaults now apply to notfound and forbidden views that are defined as methods of a decorated class. See


  • Added a "Quick Tutorial" to go with the Quick Tour

  • Removed mention of pyramid_beaker from docs. Beaker is no longer maintained. Point people at pyramid_redis_sessions instead.

  • Add documentation for pyramid.interfaces.IRendererFactory and pyramid.interfaces.IRenderer.

Backwards Incompatibilities

  • The key/values in the _query parameter of request.route_url and the query parameter of request.resource_url (and their variants), used to encode a value of None as the string 'None', leaving the resulting query string to be a=b&key=None. The value is now dropped in this situation, leaving a query string of a=b&key=. See


  • Deprecate the pyramid.interfaces.ITemplateRenderer interface. It was ill-defined and became unused when Mako and Chameleon template bindings were split into their own packages.

  • The pyramid.session.UnencryptedCookieSessionFactoryConfig API has been deprecated and is superseded by the pyramid.session.SignedCookieSessionFactory. Note that while the cookies generated by the UnencryptedCookieSessionFactoryConfig are compatible with cookies generated by old releases, cookies generated by the SignedCookieSessionFactory are not. See

  • The API is now deprecated. Instead, use the newly-added has_permission method of the request object.

  • The API is now deprecated. Instead, use the newly-added effective_principals attribute of the request object.

  • The API is now deprecated. Instead, use the newly-added authenticated_userid attribute of the request object.

  • The API is now deprecated. Instead, use the newly-added unauthenticated_userid attribute of the request object.


  • Pyramid now depends on WebOb>=1.3 (it uses webob.cookies.CookieProfile from 1.3+).

1.5a2 (2013-09-22)


  • Users can now provide dotted Python names to as the factory argument the Configurator methods named add_{view,route,subscriber}_predicate (instead of passing the predicate factory directly, you can pass a dotted name which refers to the factory).

Bug Fixes

  • Fix an exception in pyramid.path.package_name when resolving the package name for namespace packages that had no __file__ attribute.

Backwards Incompatibilities

  • Pyramid no longer depends on or configures the Mako and Chameleon templating system renderers by default. Disincluding these templating systems by default means that the Pyramid core has fewer dependencies and can run on future platforms without immediate concern for the compatibility of its templating add-ons. It also makes maintenance slightly more effective, as different people can maintain the templating system add-ons that they understand and care about without needing commit access to the Pyramid core, and it allows users who just don't want to see any packages they don't use come along for the ride when they install Pyramid.

    This means that upon upgrading to Pyramid 1.5a2+, projects that use either of these templating systems will see a traceback that ends something like this when their application attempts to render a Chameleon or Mako template:

    ValueError: No such renderer factory .pt


    ValueError: No such renderer factory .mako


    ValueError: No such renderer factory .mak

    Support for Mako templating has been moved into an add-on package named pyramid_mako, and support for Chameleon templating has been moved into an add-on package named pyramid_chameleon. These packages are drop-in replacements for the old built-in support for these templating langauges. All you have to do is install them and make them active in your configuration to register renderer factories for .pt and/or .mako (or .mak) to make your application work again.

    To re-add support for Chameleon and/or Mako template renderers into your existing projects, follow the below steps.

    If you depend on Mako templates:

    • Make sure the pyramid_mako package is installed. One way to do this is by adding pyramid_mako to the install_requires section of your package's file and afterwards rerunning develop:

              'pyramid_mako',         # new dependency
    • Within the portion of your application which instantiates a Pyramid pyramid.config.Configurator (often the main() function in your project's file), tell Pyramid to include the pyramid_mako includeme:

      config = Configurator(.....)

    If you depend on Chameleon templates:

    • Make sure the pyramid_chameleon package is installed. One way to do this is by adding pyramid_chameleon to the install_requires section of your package's file and afterwards rerunning develop:

              'pyramid_chameleon',         # new dependency
    • Within the portion of your application which instantiates a Pyramid ~pyramid.config.Configurator (often the main() function in your project's file), tell Pyramid to include the pyramid_chameleon includeme:

      config = Configurator(.....)

    Note that it's also fine to install these packages into older Pyramids for forward compatibility purposes. Even if you don't upgrade to Pyramid 1.5 immediately, performing the above steps in a Pyramid 1.4 installation is perfectly fine, won't cause any difference, and will give you forward compatibility when you eventually do upgrade to Pyramid 1.5.

    With the removal of Mako and Chameleon support from the core, some unit tests that use the pyramid.renderers.render* methods may begin to fail. If any of your unit tests are invoking either pyramid.renderers.render() or pyramid.renderers.render_to_response() with either Mako or Chameleon templates then the pyramid.config.Configurator instance in effect during the unit test should be also be updated to include the addons, as shown above. For example:

    class ATest(unittest.TestCase):
        def setUp(self):
            self.config = pyramid.testing.setUp()
        def test_it(self):
            result = pyramid.renderers.render('mypkg:templates/home.mako', {})


    class ATest(unittest.TestCase):
        def setUp(self):
            self.config = pyramid.testing.setUp()
        def test_it(self):
            result = pyramid.renderers.render('mypkg:templates/', {})
  • If you're using the Pyramid debug toolbar, when you upgrade Pyramid to 1.5a2+, you'll also need to upgrade the pyramid_debugtoolbar package to at least version 1.0.8, as older toolbar versions are not compatible with Pyramid 1.5a2+ due to the removal of Mako support from the core. It's fine to use this newer version of the toolbar code with older Pyramids too.

  • Removed the request.response_* varying attributes. These attributes have been deprecated since Pyramid 1.1, and as per the deprecation policy, have now been removed.

  • request.response will no longer be mutated when using the pyramid.renderers.render() API. Almost all renderers mutate the request.response response object (for example, the JSON renderer sets request.response.content_type to application/json), but this is only necessary when the renderer is generating a response; it was a bug when it was done as a side effect of calling pyramid.renderers.render().

  • Removed the bfg2pyramid fixer script.

  • The event is now sent after response callbacks are executed. It previously executed before response callbacks were executed. Rationale: it's more useful to be able to inspect the response after response callbacks have done their jobs instead of before.

  • Removed the class named pyramid.view.static that had been deprecated since Pyramid 1.1. Instead use pyramid.static.static_view with use_subpath=True argument.

  • Removed the pyramid.view.is_response function that had been deprecated since Pyramid 1.1. Use the pyramid.request.Request.is_response method instead.

  • Removed the ability to pass the following arguments to pyramid.config.Configurator.add_route: view, view_context. view_for, view_permission, view_renderer, and view_attr. Using these arguments had been deprecated since Pyramid 1.1. Instead of passing view-related arguments to add_route, use a separate call to pyramid.config.Configurator.add_view to associate a view with a route using its route_name argument. Note that this impacts the pyramid.config.Configurator.add_static_view function too, because it delegates to add_route.

  • Removed the ability to influence and query a pyramid.request.Request object as if it were a dictionary. Previously it was possible to use methods like __getitem__, get, items, and other dictlike methods to access values in the WSGI environment. This behavior had been deprecated since Pyramid 1.1. Use methods of request.environ (a real dictionary) instead.

  • Removed ancient backwards compatibility hack in pyramid.traversal.DefaultRootFactory which populated the __dict__ of the factory with the matchdict values for compatibility with BFG 0.9.

  • The renderer_globals_factory argument to the pyramid.config.Configurator` constructor and its ``setup_registry method has been removed. The set_renderer_globals_factory method of pyramid.config.Configurator has also been removed. The (internal) pyramid.interfaces.IRendererGlobals interface was also removed. These arguments, methods and interfaces had been deprecated since 1.1. Use a BeforeRender event subscriber as documented in the "Hooks" chapter of the Pyramid narrative documentation instead of providing renderer globals values to the configurator.


  • The pyramid.config.Configurator.set_request_property method now issues a deprecation warning when used. It had been docs-deprecated in 1.4 but did not issue a deprecation warning when used.

1.5a1 (2013-08-30)


  • A new http exception subclass named pyramid.httpexceptions.HTTPSuccessful was added. You can use this class as the context of an exception view to catch all 200-series "exceptions" (e.g. "raise HTTPOk"). This also allows you to catch only the HTTPOk exception itself; previously this was impossible because a number of other exceptions (such as HTTPNoContent) inherited from HTTPOk, but now they do not.

  • You can now generate "hybrid" urldispatch/traversal URLs more easily by using the new route_name, route_kw and route_remainder_name arguments to request.resource_url and request.resource_path. See the new section of the "Combining Traversal and URL Dispatch" documentation chapter entitled "Hybrid URL Generation".

  • It is now possible to escape double braces in Pyramid scaffolds (unescaped, these represent replacement values). You can use \{\{a\}\} to represent a "bare" {{a}}. See

  • Add localizer and locale_name properties (reified) to the request. See Note that the pyramid.i18n.get_localizer and pyramid.i18n.get_locale_name functions now simply look up these properties on the request.

  • Add pdistreport script, which prints the Python version in use, the Pyramid version in use, and the version number and location of all Python distributions currently installed.

  • Add the ability to invert the result of any view, route, or subscriber predicate using the not_ class. For example:

    from pyramid.config import not_
    @view_config(route_name='myroute', request_method=not_('POST'))
    def myview(request): ...

    The above example will ensure that the view is called if the request method is not POST (at least if no other view is more specific).

    The pyramid.config.not_ class can be used against any value that is a predicate value passed in any of these contexts:

    • pyramid.config.Configurator.add_view

    • pyramid.config.Configurator.add_route

    • pyramid.config.Configurator.add_subscriber

    • pyramid.view.view_config


  • scripts/ add support for submitting PUT and PATCH requests. See add support for submitting OPTIONS and PROPFIND requests, and allow users to specify basic authentication credentials in the request via a --login argument to the script. See

  • ACLAuthorizationPolicy supports __acl__ as a callable. This removes the ambiguity between the potential AttributeError that would be raised on the context when the property was not defined and the AttributeError that could be raised from any user-defined code within a dynamic property. It is recommended to define a dynamic ACL as a callable to avoid this ambiguity. See

  • Allow a protocol-relative URL (e.g. // to be passed to pyramid.config.Configurator.add_static_view. This allows externally-hosted static URLs to be generated based on the current protocol.

  • The AuthTktAuthenticationPolicy has two new options to configure its domain usage:

    • parent_domain: if set the authentication cookie is set on the parent domain. This is useful if you have multiple sites sharing the same domain.

    • domain: if provided the cookie is always set for this domain, bypassing all usual logic.

    See, and

  • The AuthTktAuthenticationPolicy now supports IPv6 addresses when using the include_ip=True option. This is possibly incompatible with alternative auth_tkt implementations, as the specification does not define how to properly handle IPv6. See

  • Make it possible to use variable arguments via pyramid.paster.get_appsettings. This also allowed the generated initialize_db script from the alchemy scaffold to grow support for options in the form a=1 b=2 so you can fill in values in a parameterized .ini file, e.g. initialize_myapp_db etc/development.ini a=1 b=2. See

  • The request.session.check_csrf_token() method and the check_csrf view predicate now take into account the value of the HTTP header named X-CSRF-Token (as well as the csrf_token form parameter, which they always did). The header is tried when the form parameter does not exist.

  • View lookup will now search for valid views based on the inheritance hierarchy of the context. It tries to find views based on the most specific context first, and upon predicate failure, will move up the inheritance chain to test views found by the super-type of the context. In the past, only the most specific type containing views would be checked and if no matching view could be found then a PredicateMismatch would be raised. Now predicate mismatches don't hide valid views registered on super-types. Here's an example that now works:

    class IResource(Interface):
    def get(context, request):
    @view_config(context=IResource, request_method='POST')
    def post(context, request):
    @view_config(context=IResource, request_method='DELETE')
    def delete(context, request):
    class MyResource:
    @view_config(context=MyResource, request_method='POST')
    def override_post(context, request):

    Previously the override_post view registration would hide the get and delete views in the context of MyResource -- leading to a predicate mismatch error when trying to use GET or DELETE methods. Now the views are found and no predicate mismatch is raised. See and and

  • The pserve command now takes a -v (or --verbose) flag and a -q (or --quiet) flag. Output from running pserve can be controlled using these flags. -v can be specified multiple times to increase verbosity. -q sets verbosity to 0 unconditionally. The default verbosity level is 1.

  • The alchemy scaffold tests now provide better coverage. See

  • The pyramid.config.Configurator.add_route method now supports being called with an external URL as pattern. See and the documentation section in the "URL Dispatch" chapter entitled "External Routes" for more information.

Bug Fixes

  • It was not possible to use pyramid.httpexceptions.HTTPException as the context of an exception view as very general catchall for http-related exceptions when you wanted that exception view to override the default exception view. See

  • When the pyramid.reload_templates setting was true, and a Chameleon template was reloaded, and the renderer specification named a macro (e.g., renderings of the template after the template was reloaded due to a file change would produce the entire template body instead of just a rendering of the macro. See

  • Fix an obscure problem when combining a virtual root with a route with a *traverse in its pattern. Now the traversal path generated in such a configuration will be correct, instead of an element missing a leading slash.

  • Fixed a Mako renderer bug returning a tuple with a previous defname value in some circumstances. See for more information.

  • Make the pyramid.config.assets.PackageOverrides object implement the API for __loader__ objects specified in PEP 302. Proxies to the __loader__ set by the importer, if present; otherwise, raises NotImplementedError. This makes Pyramid static view overrides work properly under Python 3.3 (previously they would not). See for more information.

  • mako_templating: added defensive workaround for non-importability of mako due to upstream markupsafe dropping Python 3.2 support. Mako templating will no longer work under the combination of MarkupSafe 0.17 and Python 3.2 (although the combination of MarkupSafe 0.17 and Python 3.3 or any supported Python 2 version will work OK).

  • Spaces and dots may now be in mako renderer template paths. This was broken when support for the new makodef syntax was added in 1.4a1. See

  • pyramid.debug_authorization=true will now correctly print out Allowed for views registered with NO_PERMISSION_REQUIRED instead of invoking the permits method of the authorization policy. See

  • Pyramid failed to install on some systems due to being packaged with some test files containing higher order characters in their names. These files have now been removed. See

  • pyramid.testing.DummyResource didn't define __bool__, so code under Python 3 would use __len__ to find truthiness; this usually caused an instance of DummyResource to be "falsy" instead of "truthy". See

  • The alchemy scaffold would break when the database was MySQL during tables creation. See

  • The current_route_url method now attaches the query string to the URL by default. See

  • Make pserve.cherrypy_server_runner Python 3 compatible. See

Backwards Incompatibilities

  • Modified the current_route_url method in pyramid.Request. The method previously returned the URL without the query string by default, it now does attach the query string unless it is overriden.

  • The route_url and route_path APIs no longer quote / to %2F when a replacement value contains a /. This was pointless, as WSGI servers always unquote the slash anyway, and Pyramid never sees the quoted value.

  • It is no longer possible to set a locale_name attribute of the request, nor is it possible to set a localizer attribute of the request. These are now "reified" properties that look up a locale name and localizer respectively using the machinery described in the "Internationalization" chapter of the documentation.

  • If you send an X-Vhm-Root header with a value that ends with a slash (or any number of slashes), the trailing slash(es) will be removed before a URL is generated when you use use request.resource_url or request.resource_path. Previously the virtual root path would not have trailing slashes stripped, which would influence URL generation.

  • The pyramid.interfaces.IResourceURL interface has now grown two new attributes: virtual_path_tuple and physical_path_tuple. These should be the tuple form of the resource's path (physical and virtual).

1.4 (2012-12-18)


  • Fix functional tests in the ZODB tutorial

1.4b3 (2012-12-10)

  • Packaging release only, no code changes. 1.4b2 was a brownbag release due to missing directories in the tarball.

1.4b2 (2012-12-10)


  • Scaffolding is now PEP-8 compliant (at least for a brief shining moment).

  • Tutorial improvements.

Backwards Incompatibilities

  • Modified the _depth argument to pyramid.view.view_config to accept a value relative to the invocation of view_config itself. Thus, when it was previously expecting a value of 1 or greater, to reflect that the caller of view_config is 1 stack frame away from venusian.attach, this implementation detail is now hidden.

  • Modified the _backframes argument to pyramid.util.action_method in a similar way to the changes described to _depth above. This argument remains undocumented, but might be used in the wild by some insane person.

1.4b1 (2012-11-21)


  • Small microspeed enhancement which anticipates that a pyramid.response.Response object is likely to be returned from a view. Some code is shortcut if the class of the object returned by a view is this class. A similar microoptimization was done to pyramid.request.Request.is_response.

  • Make it possible to use variable arguments on p* commands (pserve, pshell, pviews, etc) in the form a=1 b=2 so you can fill in values in parameterized .ini file, e.g. pshell etc/development.ini http_port=8080. See

  • A somewhat advanced and obscure feature of Pyramid event handlers is their ability to handle "multi-interface" notifications. These notifications have traditionally presented multiple objects to the subscriber callable. For instance, if an event was sent by code like this:

    registry.notify(event, context)

    In the past, in order to catch such an event, you were obligated to write and register an event subscriber that mentioned both the event and the context in its argument list:

    @subscriber([SomeEvent, SomeContextType])
    def asubscriber(event, context):

    In many subscriber callables registered this way, it was common for the logic in the subscriber callable to completely ignore the second and following arguments (e.g. context in the above example might be ignored), because they usually existed as attributes of the event anyway. You could usually get the same value by doing event.context or similar.

    The fact that you needed to put an extra argument which you usually ignored in the subscriber callable body was only a minor annoyance until we added "subscriber predicates", used to narrow the set of circumstances under which a subscriber will be executed, in a prior 1.4 alpha release. Once those were added, the annoyance was escalated, because subscriber predicates needed to accept the same argument list and arity as the subscriber callables that they were configured against. So, for example, if you had these two subscriber registrations in your code:

    @subscriber([SomeEvent, SomeContextType])
    def asubscriber(event, context):
    def asubscriber(event):

    And you wanted to use a subscriber predicate:

    @subscriber([SomeEvent, SomeContextType], mypredicate=True)
    def asubscriber1(event, context):
    @subscriber(SomeOtherEvent, mypredicate=True)
    def asubscriber2(event):

    If an existing mypredicate subscriber predicate had been written in such a way that it accepted only one argument in its __call__, you could not use it against a subscription which named more than one interface in its subscriber interface list. Similarly, if you had written a subscriber predicate that accepted two arguments, you couldn't use it against a registration that named only a single interface type.

    For example, if you created this predicate:

    class MyPredicate(object):
        # portions elided...
        def __call__(self, event):
            return self.val ==

    It would not work against a multi-interface-registered subscription, so in the above example, when you attempted to use it against asubscriber1, it would fail at runtime with a TypeError, claiming something was attempting to call it with too many arguments.

    To hack around this limitation, you were obligated to design the mypredicate predicate to expect to receive in its __call__ either a single event argument (a SomeOtherEvent object) or a pair of arguments (a SomeEvent object and a SomeContextType object), presumably by doing something like this:

    class MyPredicate(object):
        # portions elided...
        def __call__(self, event, context=None):
            return self.val ==

    This was confusing and bad.

    In order to allow people to ignore unused arguments to subscriber callables and to normalize the relationship between event subscribers and subscriber predicates, we now allow both subscribers and subscriber predicates to accept only a single event argument even if they've been subscribed for notifications that involve multiple interfaces. Subscribers and subscriber predicates that accept only one argument will receive the first object passed to notify; this is typically (but not always) the event object. The other objects involved in the subscription lookup will be discarded. You can now write an event subscriber that accepts only event even if it subscribes to multiple interfaces:

    @subscriber([SomeEvent, SomeContextType])
    def asubscriber(event):
        # this will work!

    This prevents you from needing to match the subscriber callable parameters to the subscription type unnecessarily, especially when you don't make use of any argument in your subscribers except for the event object itself.

    Note, however, that if the event object is not the first object in the call to notify, you'll run into trouble. For example, if notify is called with the context argument first:

    registry.notify(context, event)

    You won't be able to take advantage of the event-only feature. It will "work", but the object received by your event handler won't be the event object, it will be the context object, which won't be very useful:

    @subscriber([SomeContextType, SomeEvent])
    def asubscriber(event):
        # bzzt! you'll be getting the context here as ``event``, and it'll
        # be useless

    Existing multiple-argument subscribers continue to work without issue, so you should continue use those if your system notifies using multiple interfaces and the first interface is not the event interface. For example:

    @subscriber([SomeContextType, SomeEvent])
    def asubscriber(context, event):
        # this will still work!

    The event-only feature makes it possible to use a subscriber predicate that accepts only a request argument within both multiple-interface subscriber registrations and single-interface subscriber registrations. You needn't make slightly different variations of predicates depending on the subscription type arguments. Instead, just write all your subscriber predicates so they only accept event in their __call__ and they'll be useful across all registrations for subscriptions that use an event as their first argument, even ones which accept more than just event.

    However, the same caveat applies to predicates as to subscriber callables: if you're subscribing to a multi-interface event, and the first interface is not the event interface, the predicate won't work properly. In such a case, you'll need to match the predicate __call__ argument ordering and composition to the ordering of the interfaces. For example, if the registration for the subscription uses [SomeContext, SomeEvent], you'll need to reflect that in the ordering of the parameters of the predicate's __call__ method:

    def __call__(self, context, event):
        return event.request.path.startswith(self.val)

    tl;dr: 1) When using multi-interface subscriptions, always use the event type as the first subscription registration argument and 2) When 1 is true, use only event in your subscriber and subscriber predicate parameter lists, no matter how many interfaces the subscriber is notified with. This combination will result in the maximum amount of reusability of subscriber predicates and the least amount of thought on your part. Drink responsibly.

Bug Fixes

  • A failure when trying to locate the attribute __text__ on route and view predicates existed when the debug_routematch setting was true or when the pviews command was used. See


  • Sync up tutorial source files with the files that are rendered by the scaffold that each uses.

1.4a4 (2012-11-14)


  • pyramid.authentication.AuthTktAuthenticationPolicy has been updated to support newer hashing algorithms such as sha512. Existing applications should consider updating if possible for improved security over the default md5 hashing.

  • Added an effective_principals route and view predicate.

  • Do not allow the userid returned from the authenticated_userid or the userid that is one of the list of principals returned by effective_principals to be either of the strings system.Everyone or system.Authenticated when any of the built-in authorization policies that live in pyramid.authentication are in use. These two strings are reserved for internal usage by Pyramid and they will not be accepted as valid userids.

  • Slightly better debug logging from pyramid.authentication.RepozeWho1AuthenticationPolicy.

  • used to return True if no view could be found. It now raises a TypeError exception in that case, as it doesn't make sense to assert that a nonexistent view is execution-permitted. See

  • Allow a _depth argument to pyramid.view.view_config, which will permit limited composition reuse of the decorator by other software that wants to provide custom decorators that are much like view_config.

  • Allow an iterable of decorators to be passed to pyramid.config.Configurator.add_view. This allows views to be wrapped by more than one decorator without requiring combining the decorators yourself.

Bug Fixes

  • In the past if a renderer returned None, the body of the resulting response would be set explicitly to the empty string. Instead, now, the body is left unchanged, which allows the renderer to set a body itself by using e.g. request.response.body = b'foo'. The body set by the renderer will be unmolested on the way out. See

  • In uncommon cases, the pyramid_excview_tween_factory might have inadvertently raised a KeyError looking for request_iface as an attribute of the request. It no longer fails in this case. See

  • Be more tolerant of potential error conditions in match_param and physical_path predicate implementations; instead of raising an exception, return False.

  • pyramid.view.render_view was not functioning properly under Python 3.x due to a byte/unicode discrepancy. See


  • pyramid.authentication.AuthTktAuthenticationPolicy will emit a warning if an application is using the policy without explicitly passing a hashalg argument. This is because the default is "md5" which is considered theoretically subject to collision attacks. If you really want "md5" then you must specify it explicitly to get rid of the warning.


  • All of the tutorials that use pyramid.authentication.AuthTktAuthenticationPolicy now explicitly pass sha512 as a hashalg argument.


  • Move TopologicalSorter from pyramid.config.util to pyramid.util, move CyclicDependencyError from pyramid.config.util to pyramid.exceptions, rename Singleton to Sentinel and move from pyramid.config.util to pyramid.util; this is in an effort to move that stuff that may be an API one day out of pyramid.config.util, because that package should never be imported from non-Pyramid code. TopologicalSorter is still not an API, but may become one.

  • Get rid of shady monkeypatching of pyramid.request.Request and pyramid.response.Response done within the of Pyramid. Webob no longer relies on this being done. Instead, the ResponseClass attribute of the Pyramid Request class is assigned to the Pyramid response class; that's enough to satisfy WebOb and behave as it did before with the monkeypatching.

1.4a3 (2012-10-26)

Bug Fixes

  • The match_param predicate's text method was fixed to sort its values. Part of

  • 1.4a pyramid.scripting.prepare behaved differently than 1.3 series function of same name. In particular, if passed a request, it would not set the registry attribute of the request like 1.3 did. A symptom would be that passing a request to pyramid.paster.bootstrap (which uses the function) that did not have a registry attribute could assume that the registry would be attached to the request by Pyramid. This assumption could be made in 1.3, but not in 1.4. The assumption can now be made in 1.4 too (a registry is attached to a request passed to bootstrap or prepare).

  • When registering a view configuration that named a Chameleon ZPT renderer with a macro name in it (e.g. renderer='some/ as well as a view configuration without a macro name in it that pointed to the same template (e.g. renderer='some/'), internal caching could confuse the two, and your code might have rendered one instead of the other.


  • Allow multiple values to be specified to the request_param view/route predicate as a sequence. Previously only a single string value was allowed. See

  • Comments with references to documentation sections placed in scaffold .ini files.

  • Added an HTTP Basic authentication policy at pyramid.authentication.BasicAuthAuthenticationPolicy.

  • The Configurator testing_securitypolicy method now returns the policy object it creates.

  • The Configurator testing_securitypolicy method accepts two new arguments: remember_result and forget_result. If supplied, these values influence the result of the policy's remember and forget methods, respectively.

  • The DummySecurityPolicy created by testing_securitypolicy now sets a forgotten value on the policy (the value True) when its forget method is called.

  • The DummySecurityPolicy created by testing_securitypolicy now sets a remembered value on the policy, which is the value of the principal argument it's called with when its remember method is called.

  • New physical_path view predicate. If specified, this value should be a string or a tuple representing the physical traversal path of the context found via traversal for this predicate to match as true. For example: physical_path='/' or physical_path='/a/b/c' or physical_path=('', 'a', 'b', 'c'). This is not a path prefix match or a regex, it's a whole-path match. It's useful when you want to always potentially show a view when some object is traversed to, but you can't be sure about what kind of object it will be, so you can't use the context predicate. The individual path elements in between slash characters or in tuple elements should be the Unicode representation of the name of the resource and should not be encoded in any way.

1.4a2 (2012-09-27)

Bug Fixes


  • A new pyramid.session.check_csrf_token convenience function was added.

  • A check_csrf view predicate was added. For example, you can now do config.add_view(someview, check_csrf=True). When the predicate is checked, if the csrf_token value in request.params matches the CSRF token in the request's session, the view will be permitted to execute. Otherwise, it will not be permitted to execute.

  • Add Base.metadata.bind = engine to alchemy template, so that tables defined imperatively will work.


  • update wiki2 SQLA tutorial with the changes required after inserting Base.metadata.bind = engine into the alchemy scaffold.

1.4a1 (2012-09-16)

Bug Fixes

  • Forward port from 1.3 branch: When no authentication policy was configured, a call to would unconditionally return the empty list. This was incorrect, it should have unconditionally returned [Everyone], and now does.

  • Explicit url dispatch regexes can now contain colons.

  • On at least one 64-bit Ubuntu system under Python 3.2, using the view_config decorator caused a RuntimeError: dictionary changed size during iteration exception. It no longer does. See for more information.

  • In Mako Templates lookup, check if the uri is already adjusted and bring it back to an asset spec. Normally occurs with inherited templates or included components.

  • In Mako Templates lookup, check for absolute uri (using mako directories) when mixing up inheritance with asset specs.

  • HTTP Accept headers were not being normalized causing potentially conflicting view registrations to go unnoticed. Two views that only differ in the case ('text/html' vs. 'text/HTML') will now raise an error.

  • Forward-port from 1.3 branch: when registering multiple views with an accept predicate in a Pyramid application running under Python 3, you might have received a TypeError: unorderable types: function() < function() exception.


  • Python 3.3 compatibility.

  • Configurator.add_directive now accepts arbitrary callables like partials or objects implementing __call__ which dont have __name__ and __doc__ attributes. See and

  • Third-party custom view, route, and subscriber predicates can now be added for use by view authors via pyramid.config.Configurator.add_view_predicate, pyramid.config.Configurator.add_route_predicate and pyramid.config.Configurator.add_subscriber_predicate. So, for example, doing this:

    config.add_view_predicate('abc', my.package.ABCPredicate)

    Might allow a view author to do this in an application that configured that predicate:


    Similar features exist for add_route, and add_subscriber. See "Adding A Third Party View, Route, or Subscriber Predicate" in the Hooks chapter for more information.

    Note that changes made to support the above feature now means that only actions registered using the same "order" can conflict with one another. It used to be the case that actions registered at different orders could potentially conflict, but to my knowledge nothing ever depended on this behavior (it was a bit silly).

  • Custom objects can be made easily JSON-serializable in Pyramid by defining a __json__ method on the object's class. This method should return values natively serializable by json.dumps (such as ints, lists, dictionaries, strings, and so forth).

  • The JSON renderer now allows for the definition of custom type adapters to convert unknown objects to JSON serializations.

  • As of this release, the request_method predicate, when used, will also imply that HEAD is implied when you use GET. For example, using @view_config(request_method='GET') is equivalent to using @view_config(request_method=('GET', 'HEAD')). Using @view_config(request_method=('GET', 'POST') is equivalent to using @view_config(request_method=('GET', 'HEAD', 'POST'). This is because HEAD is a variant of GET that omits the body, and WebOb has special support to return an empty body when a HEAD is used.

  • config.add_request_method has been introduced to support extending request objects with arbitrary callables. This method expands on the previous config.set_request_property by supporting methods as well as properties. This method now causes less code to be executed at request construction time than config.set_request_property in version 1.3.

  • Don't add a ? to URLs generated by request.resource_url if the query argument is provided but empty.

  • Don't add a ? to URLs generated by request.route_url if the _query argument is provided but empty.

  • The static view machinery now raises (rather than returns) HTTPNotFound and HTTPMovedPermanently exceptions, so these can be caught by the Not Found View (and other exception views).

  • The Mako renderer now supports a def name in an asset spec. When the def name is present in the asset spec, the system will render the template def within the template and will return the result. An example asset spec is package:path/to/template#defname.mako. This will render the def named defname inside the template.mako template instead of rendering the entire template. The old way of returning a tuple in the form ('defname', {}) from the view is supported for backward compatibility,

  • The Chameleon ZPT renderer now accepts a macro name in an asset spec. When the macro name is present in the asset spec, the system will render the macro listed as a define-macro and return the result instead of rendering the entire template. An example asset spec: package:path/to/ This will render the macro defined as macroname within the template instead of the entire templae.

  • When there is a predicate mismatch exception (seen when no view matches for a given request due to predicates not working), the exception now contains a textual description of the predicate which didn't match.

  • An add_permission directive method was added to the Configurator. This directive registers a free-standing permission introspectable into the Pyramid introspection system. Frameworks built atop Pyramid can thus use the permissions introspectable category data to build a comprehensive list of permissions supported by a running system. Before this method was added, permissions were already registered in this introspectable category as a side effect of naming them in an add_view call, this method just makes it possible to arrange for a permission to be put into the permissions introspectable category without naming it along with an associated view. Here's an example of usage of add_permission:

    config = Configurator()
  • The UnencryptedCookieSessionFactoryConfig now accepts signed_serialize and signed_deserialize hooks which may be used to influence how the sessions are marshalled (by default this is done with HMAC+pickle).

  • pyramid.testing.DummyRequest now supports methods supplied by the pyramid.util.InstancePropertyMixin class such as set_property.

  • Request properties and methods added via config.set_request_property or config.add_request_method are now available to tweens.

  • Request properties and methods added via config.set_request_property or config.add_request_method are now available in the request object returned from pyramid.paster.bootstrap.

  • request.context of environment request during bootstrap is now the root object if a context isn't already set on a provided request.

  • The pyramid.decorator.reify function is now an API, and was added to the API documentation.

  • Added the pyramid.testing.testConfig context manager, which can be used to generate a configurator in a test, e.g. with testing.testConfig(...):.

  • Users can now invoke a subrequest from within view code using a new request.invoke_subrequest API.


  • The pyramid.config.Configurator.set_request_property has been documentation-deprecated. The method remains usable but the more featureful pyramid.config.Configurator.add_request_method should be used in its place (it has all of the same capabilities but can also extend the request object with methods).

Backwards Incompatibilities

  • The Pyramid router no longer adds the values bfg.routes.route or bfg.routes.matchdict to the request's WSGI environment dictionary. These values were docs-deprecated in repoze.bfg 1.0 (effectively seven minor releases ago). If your code depended on these values, use request.matched_route and request.matchdict instead.

  • It is no longer possible to pass an environ dictionary directly to pyramid.traversal.ResourceTreeTraverser.__call__ (aka ModelGraphTraverser.__call__). Instead, you must pass a request object. Passing an environment instead of a request has generated a deprecation warning since Pyramid 1.1.

  • Pyramid will no longer work properly if you use the webob.request.LegacyRequest as a request factory. Instances of the LegacyRequest class have a request.path_info which return a string. This Pyramid release assumes that request.path_info will unconditionally be Unicode.

  • The functions from pyramid.chameleon_zpt and pyramid.chameleon_text named get_renderer, get_template, render_template, and render_template_to_response have been removed. These have issued a deprecation warning upon import since Pyramid 1.0. Use pyramid.renderers.get_renderer(), pyramid.renderers.get_renderer().implementation(), pyramid.renderers.render() or pyramid.renderers.render_to_response respectively instead of these functions.

  • The pyramid.configuration module was removed. It had been deprecated since Pyramid 1.0 and printed a deprecation warning upon its use. Use pyramid.config instead.

  • The pyramid.paster.PyramidTemplate API was removed. It had been deprecated since Pyramid 1.1 and issued a warning on import. If your code depended on this, adjust your code to import pyramid.scaffolds.PyramidTemplate instead.

  • The pyramid.settings.get_settings() API was removed. It had been printing a deprecation warning since Pyramid 1.0. If your code depended on this API, use pyramid.threadlocal.get_current_registry().settings instead or use the settings attribute of the registry available from the request (request.registry.settings).

  • These APIs from the pyramid.testing module were removed. They have been printing deprecation warnings since Pyramid 1.0:

    • registerDummySecurityPolicy, use pyramid.config.Configurator.testing_securitypolicy instead.

    • registerResources (aka registerModels, use pyramid.config.Configurator.testing_resources instead.

    • registerEventListener, use pyramid.config.Configurator.testing_add_subscriber instead.

    • registerTemplateRenderer (aka registerDummyRenderer`), use pyramid.config.Configurator.testing_add_template instead.

    • registerView, use pyramid.config.Configurator.add_view instead.

    • registerUtility, use pyramid.config.Configurator.registry.registerUtility instead.

    • registerAdapter, use pyramid.config.Configurator.registry.registerAdapter instead.

    • registerSubscriber, use pyramid.config.Configurator.add_subscriber instead.

    • registerRoute, use pyramid.config.Configurator.add_route instead.

    • registerSettings, use pyramid.config.Configurator.add_settings instead.

  • In Pyramid 1.3 and previous, the __call__ method of a Response object was invoked before any finished callbacks were executed. As of this release, the __call__ method of a Response object is invoked after finished callbacks are executed. This is in support of the request.invoke_subrequest feature.

  • The 200-series exception responses named HTTPCreated, HTTPAccepted, HTTPNonAuthoritativeInformation, HTTPNoContent, HTTPResetContent, and HTTPPartialContent in pyramid.httpexceptions no longer inherit from HTTPOk. Instead they inherit from a new base class named HTTPSuccessful. This will have no effect on you unless you've registered an exception view for HTTPOk and expect that exception view to catch all the aforementioned exceptions.


  • Added an "Upgrading Pyramid" chapter to the narrative documentation. It describes how to cope with deprecations and removals of Pyramid APIs and how to show Pyramid-generated deprecation warnings while running tests and while running a server.

  • Added a "Invoking a Subrequest" chapter to the documentation. It describes how to use the new request.invoke_subrequest API.


  • Pyramid now requires WebOb 1.2b3+ (the prior Pyramid release only relied on 1.2dev+). This is to ensure that we obtain a version of WebOb that returns request.path_info as text.

1.3 (2012-03-21)

Bug Fixes

  • When pyramid.wsgi.wsgiapp2 calls the downstream WSGI app, the app's environ will no longer have (deprecated and potentially misleading) bfg.routes.matchdict or bfg.routes.route keys in it. A symptom of this bug would be a wsgiapp2-wrapped Pyramid app finding the wrong view because it mistakenly detects that a route was matched when, in fact, it was not.

  • The fix for issue (which made it possible for instance methods to be used as view callables) introduced a backwards incompatibility when methods that declared only a request argument were used. See

1.3b3 (2012-03-17)

Bug Fixes

  • config.add_view(<aninstancemethod>) raised AttributeError involving __text__. See

  • Remove references to do-nothing pyramid.debug_templates setting in all Pyramid-provided .ini files. This setting previously told Chameleon to render better exceptions; now Chameleon always renders nice exceptions regardless of the value of this setting.


  • The alchemy scaffold now shows an informative error message in the browser if the person creating the project forgets to run the initialization script.

  • The alchemy scaffold initialization script is now called initialize_<projectname>_db instead of populate_<projectname>.


  • Wiki tutorials improved due to collaboration at PyCon US 2012 sprints.

1.3b2 (2012-03-02)

Bug Fixes

  • The method pyramid.request.Request.partial_application_url is no longer in the API docs. It was meant to be a private method; its publication in the documentation as an API method was a mistake, and it has been renamed to something private.

  • When a static view was registered using an absolute filesystem path on Windows, the request.static_url function did not work to generate URLs to its resources. Symptom: "No static URL definition matching c:\foo\bar\baz".

  • Make all tests pass on Windows XP.

  • Bug in ACL authentication checking on Python 3: the permits and principals_allowed_by_permission method of pyramid.authorization.ACLAuthenticationPolicy could return an inappropriate True value when a permission on an ACL was a string rather than a sequence, and then only if the ACL permission string was a substring of the permission value passed to the function.

    This bug effects no Pyramid deployment under Python 2; it is a bug that exists only in deployments running on Python 3. It has existed since Pyramid 1.3a1.

    This bug was due to the presence of an __iter__ attribute on strings under Python 3 which is not present under strings in Python 2.

1.3b1 (2012-02-26)

Bug Fixes

  • pyramid.config.Configurator.with_package didn't work if the Configurator was an old-style pyramid.configuration.Configurator instance.

  • Pyramid authorization policies did not show up in the introspector.


  • All references to the tmpl_context request variable were removed from the docs. Its existence in Pyramid is confusing for people who were never Pylons users. It was added as a porting convenience for Pylons users in Pyramid 1.0, but it never caught on because the Pyramid rendering system is a lot different than Pylons' was, and alternate ways exist to do what it was designed to offer in Pylons. It will continue to exist "forever" but it will not be recommended or mentioned in the docs.

1.3a9 (2012-02-22)


  • Add an introspection boolean to the Configurator constructor. If this is True, actions registered using the Configurator will be registered with the introspector. If it is False, they won't. The default is True. Setting it to False during action processing will prevent introspection for any following registration statements, and setting it to True will start them up again. This addition is to service a requirement that the debug toolbar's own views and methods not show up in the introspector.

  • New API: pyramid.config.Configurator.add_notfound_view. This is a wrapper for pyramid.Config.configurator.add_view which provides easy append_slash support and does the right thing about permissions. It should be preferred over calling add_view directly with context=HTTPNotFound as was previously recommended.

  • New API: pyramid.view.notfound_view_config. This is a decorator constructor like pyramid.view.view_config that calls pyramid.config.Configurator.add_notfound_view when scanned. It should be preferred over using pyramid.view.view_config with context=HTTPNotFound as was previously recommended.

  • New API: pyramid.config.Configurator.add_forbidden_view. This is a wrapper for pyramid.Config.configurator.add_view which does the right thing about permissions. It should be preferred over calling add_view directly with context=HTTPForbidden as was previously recommended.

  • New API: pyramid.view.forbidden_view_config. This is a decorator constructor like pyramid.view.view_config that calls pyramid.config.Configurator.add_forbidden_view when scanned. It should be preferred over using pyramid.view.view_config with context=HTTPForbidden as was previously recommended.

  • New APIs: pyramid.response.FileResponse and pyramid.response.FileIter, for usage in views that must serve files "manually".

Backwards Incompatibilities

  • Remove pyramid.config.Configurator.with_context class method. It was never an API, it is only used by pyramid_zcml and its functionality has been moved to that package's latest release. This means that you'll need to use the 0.9.2 or later release of pyramid_zcml with this release of Pyramid.

  • The introspector argument to the pyramid.config.Configurator constructor API has been removed. It has been replaced by the boolean introspection flag.

  • The pyramid.registry.noop_introspector API object has been removed.

  • The older deprecated set_notfound_view Configurator method is now an alias for the new add_notfound_view Configurator method. Likewise, the older deprecated set_forbidden_view is now an alias for the new add_forbidden_view. This has the following impact: the context sent to views with a (context, request) call signature registered via the set_notfound_view or set_forbidden_view will now be an exception object instead of the actual resource context found. Use request.context to get the actual resource context. It's also recommended to disuse set_notfound_view in favor of add_notfound_view, and disuse set_forbidden_view in favor of add_forbidden_view despite the aliasing.


  • The API documentation for pyramid.view.append_slash_notfound_view and pyramid.view.AppendSlashNotFoundViewFactory was removed. These names still exist and are still importable, but they are no longer APIs. Use pyramid.config.Configurator.add_notfound_view(append_slash=True) or pyramid.view.notfound_view_config(append_slash=True) to get the same behavior.

  • The set_forbidden_view and set_notfound_view methods of the Configurator were removed from the documentation. They have been deprecated since Pyramid 1.1.

Bug Fixes

  • The static file response object used by config.add_static_view opened the static file twice, when it only needed to open it once.

  • The AppendSlashNotFoundViewFactory used request.path to match routes. This was wrong because request.path contains the script name, and this would cause it to fail in circumstances where the script name was not empty. It should have used request.path_info, and now does.


  • Updated the "Creating a Not Found View" section of the "Hooks" chapter, replacing explanations of registering a view using add_view or view_config with ones using add_notfound_view or notfound_view_config.

  • Updated the "Creating a Not Forbidden View" section of the "Hooks" chapter, replacing explanations of registering a view using add_view or view_config with ones using add_forbidden_view or forbidden_view_config.

  • Updated the "Redirecting to Slash-Appended Routes" section of the "URL Dispatch" chapter, replacing explanations of registering a view using add_view or view_config with ones using add_notfound_view or notfound_view_config

  • Updated all tutorials to use pyramid.view.forbidden_view_config rather than pyramid.view.view_config with an HTTPForbidden context.

1.3a8 (2012-02-19)


  • The scan method of a Configurator can be passed an ignore argument, which can be a string, a callable, or a list consisting of strings and/or callables. This feature allows submodules, subpackages, and global objects from being scanned. See for more information about how to use the ignore argument to scan.

  • Better error messages when a view callable returns a value that cannot be converted to a response (for example, when a view callable returns a dictionary without a renderer defined, or doesn't return any value at all). The error message now contains information about the view callable itself as well as the result of calling it.

  • Better error message when a .pyc-only module is config.include -ed. This is not permitted due to error reporting requirements, and a better error message is shown when it is attempted. Previously it would fail with something like "AttributeError: 'NoneType' object has no attribute 'rfind'".

  • Add pyramid.config.Configurator.add_traverser API method. See the Hooks narrative documentation section entitled "Changing the Traverser" for more information. This is not a new feature, it just provides an API for adding a traverser without needing to use the ZCA API.

  • Add pyramid.config.Configurator.add_resource_url_adapter API method. See the Hooks narrative documentation section entitled "Changing How pyramid.request.Request.resource_url Generates a URL" for more information. This is not a new feature, it just provides an API for adding a resource url adapter without needing to use the ZCA API.

  • The system value req is now supplied to renderers as an alias for request. This means that you can now, for example, in a template, do req.route_url(...) instead of request.route_url(...). This is purely a change to reduce the amount of typing required to use request methods and attributes from within templates. The value request is still available too, this is just an alternative.

  • A new interface was added: pyramid.interfaces.IResourceURL. An adapter implementing its interface can be used to override resource URL generation when request.resource_url is called. This interface replaces the now-deprecated pyramid.interfaces.IContextURL interface.

  • The dictionary passed to a resource's __resource_url__ method (see "Overriding Resource URL Generation" in the "Resources" chapter) now contains an app_url key, representing the application URL generated during request.resource_url. It represents a potentially customized URL prefix, containing potentially custom scheme, host and port information passed by the user to request.resource_url. It should be used instead of request.application_url where necessary.

  • The request.resource_url API now accepts these arguments: app_url, scheme, host, and port. The app_url argument can be used to replace the URL prefix wholesale during url generation. The scheme, host, and port arguments can be used to replace the respective default values of request.application_url partially.

  • A new API named request.resource_path now exists. It works like request.resource_url but produces a relative URL rather than an absolute one.

  • The request.route_url API now accepts these arguments: _app_url, _scheme, _host, and _port. The _app_url argument can be used to replace the URL prefix wholesale during url generation. The _scheme, _host, and _port arguments can be used to replace the respective default values of request.application_url partially.

Backwards Incompatibilities

  • The pyramid.interfaces.IContextURL interface has been deprecated. People have been instructed to use this to register a resource url adapter in the "Hooks" chapter to use to influence request.resource_url URL generation for resources found via custom traversers since Pyramid 1.0.

    The interface still exists and registering such an adapter still works, but this interface will be removed from the software after a few major Pyramid releases. You should replace it with an equivalent pyramid.interfaces.IResourceURL adapter, registered using the new pyramid.config.Configurator.add_resource_url_adapter API. A deprecation warning is now emitted when a pyramid.interfaces.IContextURL adapter is found when request.resource_url is called.


  • Don't create a session instance in SQLA Wiki tutorial, use raw DBSession instead (this is more common in real SQLA apps).


  • Put pyramid.includes targets within ini files in scaffolds on separate lines in order to be able to tell people to comment out only the pyramid_debugtoolbar line when they want to disable the toolbar.


  • Depend on venusian >= 1.0a3 to provide scan ignore support.


  • Create a "MakoRendererFactoryHelper" that provides customizable settings key prefixes. Allows settings prefixes other than "mako." to be used to create different factories that don't use the global mako settings. This will be useful for the debug toolbar, which can currently be sabotaged by someone using custom mako configuration settings.

1.3a7 (2012-02-07)


  • More informative error message when a config.include cannot find an includeme. See

  • Internal: catch unhashable discriminators early (raise an error instead of allowing them to find their way into resolveConflicts).

  • The match_param view predicate now accepts a string or a tuple. This replaces the broken behavior of accepting a dict. See for more information.

Bug Fixes

Backwards Incompatibilities

  • The match_param view predicate no longer accepts a dict. This will have no negative affect because the implementation was broken for dict-based arguments.


  • Add a traversal hello world example to the narrative docs.

1.3a6 (2012-01-20)


  • New API: pyramid.config.Configurator.set_request_property. Add lazy property descriptors to a request without changing the request factory. This method provides conflict detection and is the suggested way to add properties to a request.

  • Responses generated by Pyramid's static_view now use a wsgi.file_wrapper (see when one is provided by the web server.

Bug Fixes

1.3a5 (2012-01-09)

Bug Fixes

Backwards Incompatibilities

  • The path_info route and view predicates now match against request.upath_info (Unicode) rather than request.path_info (indeterminate value based on Python 3 vs. Python 2). This has to be done to normalize matching on Python 2 and Python 3.

1.3a4 (2012-01-05)


  • New API: pyramid.request.Request.set_property. Add lazy property descriptors to a request without changing the request factory. New properties may be reified, effectively caching the value for the lifetime of the instance. Common use-cases for this would be to get a database connection for the request or identify the current user.

  • Use the waitress WSGI server instead of wsgiref in scaffolding.

Bug Fixes

  • The documentation of indicated that using it as a decorator with no arguments like this:

    def somefunc(event):

    Would register somefunc to receive all events sent via the registry, but this was untrue. Instead, it would receive no events at all. This has now been fixed and the code matches the documentation. See also

  • Literal portions of route patterns were not URL-quoted when route_url or route_path was used to generate a URL or path.

  • The result of route_path or route_url might have been unicode or str depending on the input. It is now guaranteed to always be str.

  • URL matching when the pattern contained non-ASCII characters in literal parts was indeterminate. Now the pattern supplied to add_route is assumed to be either: a unicode value, or a str value that contains only ASCII characters. If you now want to match the path info from a URL that contains high order characters, you can pass the Unicode representation of the decoded path portion in the pattern.

  • When using a traverse= route predicate, traversal would fail with a URLDecodeError if there were any high-order characters in the traversal pattern or in the matched dynamic segments.

  • Using a dynamic segment named traverse in a route pattern like this:

    config.add_route('trav_route', 'traversal/{traverse:.*}')

    Would cause a UnicodeDecodeError when the route was matched and the matched portion of the URL contained any high-order characters. See .

  • When using a *traverse stararg in a route pattern, a URL that matched that possessed a @@ in its name (signifying a view name) would be inappropriately quoted by the traversal machinery during traversal, resulting in the view not being found properly. See and .

Backwards Incompatibilities

  • String values passed to route_url or route_path that are meant to replace "remainder" matches will now be URL-quoted except for embedded slashes. For example:

    config.add_route('remain', '/foo*remainder')
    request.route_path('remain', remainder='abc / def')
    # -> '/foo/abc%20/%20def'

    Previously string values passed as remainder replacements were tacked on untouched, without any URL-quoting. But this doesn't really work logically if the value passed is Unicode (raw unicode cannot be placed in a URL or in a path) and it is inconsistent with the rest of the URL generation machinery if the value is a string (it won't be quoted unless by the caller).

    Some folks will have been relying on the older behavior to tack on query string elements and anchor portions of the URL; sorry, you'll need to change your code to use the _query and/or _anchor arguments to route_path or route_url to do this now.

  • If you pass a bytestring that contains non-ASCII characters to add_route as a pattern, it will now fail at startup time. Use Unicode instead.

1.3a3 (2011-12-21)


  • Added a prequest script (along the lines of paster request). It is documented in the "Command-Line Pyramid" chapter in the section entitled "Invoking a Request".

  • Add undocumented __discriminator__ API to derived view callables. e.g. adapters.lookup(...).__discriminator__(context, request). It will be used by superdynamic systems that require the discriminator to be used for introspection after manual view lookup.

Bug Fixes

  • Normalized exit values and -h output for all p* scripts (pviews, proutes, etc).


  • Added a section named "Making Your Script into a Console Script" in the "Command-Line Pyramid" chapter.

  • Removed the "Running Pyramid on Google App Engine" tutorial from the main docs. It survives on in the Cookbook ( Rationale: it provides the correct info for the Python 2.5 version of GAE only, and this version of Pyramid does not support Python 2.5.

1.3a2 (2011-12-14)


  • New API: pyramid.view.view_defaults. If you use a class as a view, you can use the new view_defaults class decorator on the class to provide defaults to the view configuration information used by every @view_config decorator that decorates a method of that class. It also works against view configurations involving a class made imperatively.

  • Added a backwards compatibility knob to pcreate to emulate paster create handling for the --list-templates option.

  • Changed scaffolding machinery around a bit to make it easier for people who want to have extension scaffolds that can work across Pyramid 1.0.X, 1.1.X, 1.2.X and 1.3.X. See the new "Creating Pyramid Scaffolds" chapter in the narrative documentation for more info.


  • Added documentation to "View Configuration" narrative documentation chapter about view_defaults class decorator.

  • Added API docs for view_defaults class decorator.

  • Added an API docs chapter for pyramid.scaffolds.

  • Added a narrative docs chapter named "Creating Pyramid Scaffolds".

Backwards Incompatibilities

  • The template_renderer method of pyramid.scaffolds.PyramidScaffold was renamed to render_template. If you were overriding it, you're a bad person, because it wasn't an API before now. But we're nice so we're letting you know.

1.3a1 (2011-12-09)


  • Python 3.2 compatibility.

  • New pyramid.compat module and API documentation which provides Python 2/3 straddling support for Pyramid add-ons and development environments.

  • A mako.directories setting is no longer required to use Mako templates Rationale: Mako template renderers can be specified using an absolute asset spec. An entire application can be written with such asset specs, requiring no ordered lookup path.

  • bpython interpreter compatibility in pshell. See the "Command-Line Pyramid" narrative docs chapter for more information.

  • Added get_appsettings API function to the pyramid.paster module. This function returns the settings defined within an [app:...] section in a PasteDeploy ini file.

  • Added setup_logging API function to the pyramid.paster module. This function sets up Python logging according to the logging configuration in a PasteDeploy ini file.

  • Configuration conflict reporting is reported in a more understandable way ("Line 11 in file..." vs. a repr of a tuple of similar info).

  • A configuration introspection system was added; see the narrative documentation chapter entitled "Pyramid Configuration Introspection" for more information. New APIs: pyramid.registry.Introspectable, pyramid.config.Configurator.introspector, pyramid.config.Configurator.introspectable, pyramid.registry.Registry.introspector.

  • Allow extra keyword arguments to be passed to the pyramid.config.Configurator.action method.

  • New APIs: pyramid.path.AssetResolver and pyramid.path.DottedNameResolver. The former can be used to resolve asset specifications, the latter can be used to resolve dotted names to modules or packages.

Bug Fixes

Backwards Incompatibilities

  • Pyramid no longer runs on Python 2.5 (which includes the most recent release of Jython and the Python 2.5 version of GAE as of this writing).

  • The paster command is no longer the documented way to create projects, start the server, or run debugging commands. To create projects from scaffolds, paster create is replaced by the pcreate console script. To serve up a project, paster serve is replaced by the pserve console script. New console scripts named pshell, pviews, proutes, and ptweens do what their paster <commandname> equivalents used to do. Rationale: the Paste and PasteScript packages do not run under Python 3.

  • The default WSGI server run as the result of pserve from newly rendered scaffolding is now the wsgiref WSGI server instead of the paste.httpserver server. Rationale: Rationale: the Paste and PasteScript packages do not run under Python 3.

  • The pshell command (see "paster pshell") no longer accepts a --disable-ipython command-line argument. Instead, it accepts a -p or --python-shell argument, which can be any of the values python, ipython or bpython.

  • Removed the pyramid.renderers.renderer_from_name function. It has been deprecated since Pyramid 1.0, and was never an API.

  • To use ZCML with versions of Pyramid >= 1.3, you will need pyramid_zcml version >= 0.8 and zope.configuration version >= 3.8.0. The pyramid_zcml package version 0.8 is backwards compatible all the way to Pyramid 1.0, so you won't be warned if you have older versions installed and upgrade Pyramid "in-place"; it may simply break instead.


  • Pyramid no longer depends on the zope.component package, except as a testing dependency.

  • Pyramid now depends on a zope.interface>=3.8.0, WebOb>=1.2dev, repoze.lru>=0.4, zope.deprecation>=3.5.0, translationstring>=0.4 (for Python 3 compatibility purposes). It also, as a testing dependency, depends on WebTest>=1.3.1 for the same reason.

  • Pyramid no longer depends on the Paste or PasteScript packages.


  • The SQLAlchemy Wiki tutorial has been updated. It now uses @view_config decorators and an explicit database population script.

  • Minor updates to the ZODB Wiki tutorial.

  • A narrative documentation chapter named "Extending Pyramid Configuration" was added; it describes how to add a new directive, and how use the pyramid.config.Configurator.action method within custom directives. It also describes how to add introspectable objects.

  • A narrative documentation chapter named "Pyramid Configuration Introspection" was added. It describes how to query the introspection system.


  • Rendered scaffolds have now been changed to be more relocatable (fewer mentions of the package name within files in the package).

  • The routesalchemy scaffold has been renamed alchemy, replacing the older (traversal-based) alchemy scaffold (which has been retired).

  • The starter scaffold now uses URL dispatch by default.

1.2 (2011-09-12)


1.2b3 (2011-09-11)

Bug Fixes

1.2b2 (2011-09-08)

Bug Fixes

  • The 1.2b1 tarball was a brownbag (particularly for Windows users) because it contained filenames with stray quotation marks in inappropriate places. We depend on setuptools-git to produce release tarballs, and when it was run to produce the 1.2b1 tarball, it didn't yet cope well with files present in git repositories with high-order characters in their filenames.


  • Minor tweaks to the "Introduction" narrative chapter example app and wording.

1.2b1 (2011-09-08)

Bug Fixes

  • Sometimes falling back from territory translations (de_DE) to language translations (de) would not work properly when using a localizer. See

  • The static file serving machinery could not serve files that started with a . (dot) character.

  • Static files with high-order (super-ASCII) characters in their names could not be served by a static view. The static file serving machinery inappropriately URL-quoted path segments in filenames when asking for files from the filesystem.

  • Within pyramid.traversal.traversal_path , canonicalize URL segments from UTF-8 to Unicode before checking whether a segment matches literally one of ., the empty string, or .. in case there's some sneaky way someone might tunnel those strings via UTF-8 that don't match the literals before decoded.


  • Added a "What Makes Pyramid Unique" section to the Introduction narrative chapter.

1.2a6 (2011-09-06)

Bug Fixes


  • Internalize code previously depended upon as imports from the paste.auth module (futureproof).

  • Replaced use of paste.urlparser.StaticURLParser with a derivative of Chris Rossi's "happy" static file serving code (futureproof).

  • Fixed test suite; on some systems tests would fail due to indeterminate test run ordering and a double-push-single-pop of a shared test variable.

Behavior Differences

  • An ETag header is no longer set when serving a static file. A Last-Modified header is set instead.

  • Static file serving no longer supports the wsgi.file_wrapper extension.

  • Instead of returning a 403 Forbidden error when a static file is served that cannot be accessed by the Pyramid process' user due to file permissions, an IOError (or similar) will be raised.


  • All scaffolds now send the cache_max_age parameter to the add_static_view method.

1.2a5 (2011-09-04)

Bug Fixes


  • The zope.configuration package is no longer a dependency.

1.2a4 (2011-09-02)


  • Support an onerror keyword argument to pyramid.config.Configurator.scan(). This onerror keyword argument is passed to venusian.Scanner.scan() to influence error behavior when an exception is raised during scanning.

  • The request_method predicate argument to pyramid.config.Configurator.add_view and pyramid.config.Configurator.add_route is now permitted to be a tuple of HTTP method names. Previously it was restricted to being a string representing a single HTTP method name.

  • Undeprecated pyramid.traversal.find_model, pyramid.traversal.model_path, pyramid.traversal.model_path_tuple, and pyramid.url.model_url, which were all deprecated in Pyramid 1.0. There's just not much cost to keeping them around forever as aliases to their renamed resource_* prefixed functions.

  • Undeprecated pyramid.view.bfg_view, which was deprecated in Pyramid 1.0. This is a low-cost alias to pyramid.view.view_config which we'll just keep around forever.


  • Pyramid now requires Venusian 1.0a1 or better to support the onerror keyword argument to pyramid.config.Configurator.scan.

1.2a3 (2011-08-29)

Bug Fixes

  • Pyramid did not properly generate static URLs using pyramid.url.static_url when passed a caller-package relative path due to a refactoring done in 1.2a1.

  • The settings object emitted a deprecation warning any time __getattr__ was called upon it. However, there are legitimate situations in which __getattr__ is called on arbitrary objects (e.g. hasattr). Now, the settings object only emits the warning upon successful lookup.


  • Use config.with_package in view_config decorator rather than manufacturing a new renderer helper (cleanup).

1.2a2 (2011-08-27)

Bug Fixes

  • When a renderers= argument is not specified to the Configurator constructor, eagerly register and commit the default renderer set. This permits the overriding of the default renderers, which was broken in 1.2a1 without a commit directly after Configurator construction.

  • Mako rendering exceptions had the wrong value for an error message.

  • An include could not set a root factory successfully because the Configurator constructor unconditionally registered one that would be treated as if it were "the word of the user".


  • A session factory can now be passed in using the dotted name syntax.

1.2a1 (2011-08-24)


  • The [pshell] section in an ini configuration file now treats a setup key as a dotted name that points to a callable that is passed the bootstrap environment. It can mutate the environment as necessary for great justice.

  • A new configuration setting named pyramid.includes is now available. It is described in the "Environment Variables and .ini Files Settings" narrative documentation chapter.

  • Added a route_prefix argument to the pyramid.config.Configurator.include method. This argument allows you to compose URL dispatch applications together. See the section entitled "Using a Route Prefix to Compose Applications" in the "URL Dispatch" narrative documentation chapter.

  • Added a constant for use in permission= statements to view configuration. This constant has a value of the string __no_permission_required__. This string value was previously referred to in documentation; now the documentation uses the constant.

  • Added a decorator-based way to configure a response adapter: pyramid.response.response_adapter. This decorator has the same use as pyramid.config.Configurator.add_response_adapter but it's declarative.

  • The event now has an attribute named rendering_val. This can be used to introspect the value returned by a view in a BeforeRender subscriber.

  • New configurator directive: pyramid.config.Configurator.add_tween. This directive adds a "tween". A "tween" is used to wrap the Pyramid router's primary request handling function. This is a feature may be used by Pyramid framework extensions, to provide, for example, view timing support and as a convenient place to hang bookkeeping code.

    Tweens are further described in the narrative docs section in the Hooks chapter, named "Registering Tweens".

  • New paster command paster ptweens, which prints the current "tween" configuration for an application. See the section entitled "Displaying Tweens" in the Command-Line Pyramid chapter of the narrative documentation for more info.

  • The Pyramid debug logger now uses the standard logging configuration (usually set up by Paste as part of startup). This means that output from e.g. debug_notfound, debug_authorization, etc. will go to the normal logging channels. The logger name of the debug logger will be the package name of the caller of the Configurator's constructor.

  • A new attribute is available on request objects: exc_info. Its value will be None until an exception is caught by the Pyramid router, after which it will be the result of sys.exc_info().

  • pyramid.testing.DummyRequest now implements the add_finished_callback and add_response_callback methods.

  • New methods of the pyramid.config.Configurator class: set_authentication_policy and set_authorization_policy. These are meant to be consumed mostly by add-on authors.

  • New Configurator method: set_root_factory.

  • Pyramid no longer eagerly commits some default configuration statements at Configurator construction time, which permits values passed in as constructor arguments (e.g. authentication_policy and authorization_policy) to override the same settings obtained via an "include".

  • Better Mako rendering exceptions via pyramid.mako_templating.MakoRenderingException

  • New request methods: current_route_url, current_route_path, and static_path.

  • New functions in pyramid.url: current_route_path and static_path.

  • The pyramid.request.Request.static_url API (and its brethren pyramid.request.Request.static_path, pyramid.url.static_url, and pyramid.url.static_path) now accept an asbolute filename as a "path" argument. This will generate a URL to an asset as long as the filename is in a directory which was previously registered as a static view. Previously, trying to generate a URL to an asset using an absolute file path would raise a ValueError.

  • The RemoteUserAuthenticationPolicy ``, ``AuthTktAuthenticationPolicy, and SessionAuthenticationPolicy constructors now accept an additional keyword argument named debug. By default, this keyword argument is False. When it is True, debug information will be sent to the Pyramid debug logger (usually on stderr) when the authenticated_userid or effective_principals method is called on any of these policies. The output produced can be useful when trying to diagnose authentication-related problems.

  • New view predicate: match_param. Example: a view added via config.add_view(aview, match_param='action=edit') will be called only when the request.matchdict has a value inside it named action with a value of edit.


  • The Pyramid "exception view" machinery is now implemented as a "tween" (pyramid.tweens.excview_tween_factory).

  • WSGIHTTPException (HTTPFound, HTTPNotFound, etc) now has a new API named "prepare" which renders the body and content type when it is provided with a WSGI environ. Required for debug toolbar.

  • Once __call__ or prepare is called on a WSGIHTTPException, the body will be set, and subsequent calls to __call__ will always return the same body. Delete the body attribute to rerender the exception body.

  • Previously the event wrapped a dictionary (it addressed it as its _system attribute). Now it is a dictionary (it inherits from dict), and it's the value that is passed to templates as a top-level dictionary.

  • The route_url, route_path, resource_url, static_url, and current_route_url functions in the pyramid.url package now delegate to a method on the request they've been passed, instead of the other way around. The pyramid.request.Request object now inherits from a mixin named pyramid.url.URLMethodsMixin to make this possible, and all url/path generation logic is embedded in this mixin.

  • Refactor pyramid.config into a package.

  • Removed the _set_security_policies method of the Configurator.

  • Moved the StaticURLInfo class from pyramid.static to pyramid.config.views.

  • Move the Settings class from pyramid.settings to pyramid.config.settings.

  • Move the OverrideProvider, PackageOverrides, DirectoryOverride, and FileOverride classes from pyramid.asset to pyramid.config.assets.


  • All Pyramid-related deployment settings (e.g. debug_all, debug_notfound) are now meant to be prefixed with the prefix pyramid.. For example: debug_all -> pyramid.debug_all. The old non-prefixed settings will continue to work indefinitely but supplying them may eventually print a deprecation warning. All scaffolds and tutorials have been changed to use prefixed settings.

  • The settings dictionary now raises a deprecation warning when you attempt to access its values via __getattr__ instead of via __getitem__.

Backwards Incompatibilities

  • If a string is passed as the debug_logger parameter to a Configurator, that string is considered to be the name of a global Python logger rather than a dotted name to an instance of a logger.

  • The pyramid.config.Configurator.include method now accepts only a single callable argument (a sequence of callables used to be permitted). If you are passing more than one callable to pyramid.config.Configurator.include, it will break. You now must now instead make a separate call to the method for each callable. This change was introduced to support the route_prefix feature of include.

  • It may be necessary to more strictly order configuration route and view statements when using an "autocommitting" Configurator. In the past, it was possible to add a view which named a route name before adding a route with that name when you used an autocommitting configurator. For example:

    config = Configurator(autocommit=True)
    config.add_view('my.pkg.someview', route_name='foo')
    config.add_route('foo', '/foo')

    The above will raise an exception when the view attempts to add itself. Now you must add the route before adding the view:

    config = Configurator(autocommit=True)
    config.add_route('foo', '/foo')
    config.add_view('my.pkg.someview', route_name='foo')

    This won't effect "normal" users, only people who have legacy BFG codebases that used an autocommitting configurator and possibly tests that use the configurator API (the configurator returned by pyramid.testing.setUp is an autocommitting configurator). The right way to get around this is to use a non-autocommitting configurator (the default), which does not have these directive ordering requirements.

  • The pyramid.config.Configurator.add_route directive no longer returns a route object. This change was required to make route vs. view configuration processing work properly.


  • Narrative and API documentation which used the route_url, route_path, resource_url, static_url, and current_route_url functions in the pyramid.url package have now been changed to use eponymous methods of the request instead.

  • Added a section entitled "Using a Route Prefix to Compose Applications" to the "URL Dispatch" narrative documentation chapter.

  • Added a new module to the API docs: pyramid.tweens.

  • Added a "Registering Tweens" section to the "Hooks" narrative chapter.

  • Added a "Displaying Tweens" section to the "Command-Line Pyramid" narrative chapter.

  • Added documentation for the pyramid.tweens and pyramid.includes configuration settings to the "Environment Variables and .ini Files Settings" chapter.

  • Added a Logging chapter to the narrative docs (based on the Pylons logging docs, thanks Phil).

  • Added a Paste chapter to the narrative docs (moved content from the Project chapter).

  • Added the pyramid.interfaces.IDict interface representing the methods of a dictionary, for documentation purposes only (IMultiDict and IBeforeRender inherit from it).

  • All tutorials now use - The route_url, route_path, resource_url, static_url, and current_route_url methods of the request rather than the function variants imported from pyramid.url.

  • The ZODB wiki tutorial now uses the pyramid_zodbconn package rather than the repoze.zodbconn package to provide ZODB integration.

Dependency Changes

  • Pyramid now relies on PasteScript >= 1.7.4. This version contains a feature important for allowing flexible logging configuration.


  • All scaffolds now use the pyramid_tm package rather than the repoze.tm2 middleware to manage transaction management.

  • The ZODB scaffold now uses the pyramid_zodbconn package rather than the repoze.zodbconn package to provide ZODB integration.

  • All scaffolds now use the pyramid_debugtoolbar package rather than the WebError package to provide interactive debugging features.

  • Projects created via a scaffold no longer depend on the WebError package at all; configuration in the production.ini file which used to require its error_catcher middleware has been removed. Configuring error catching / email sending is now the domain of the pyramid_exclog package (see

Bug Fixes

1.1 (2011-07-22)


  • Added the pyramid.renderers.null_renderer object as an API. The null renderer is an object that can be used in advanced integration cases as input to the view configuration renderer= argument. When the null renderer is used as a view renderer argument, Pyramid avoids converting the view callable result into a Response object. This is useful if you want to reuse the view configuration and lookup machinery outside the context of its use by the Pyramid router. This feature was added for consumption by the pyramid_rpc package, which uses view configuration and lookup outside the context of a router in exactly this way. pyramid_rpc has been broken under 1.1 since 1.1b1; adding it allows us to make it work again.

  • Change all scaffolding templates that point to to use /projects/pyramid/current rather than /projects/pyramid/dev.


  • Remove compat code that served only the purpose of providing backwards compatibility with Python 2.4.

  • Add a deprecation warning for non-API function pyramid.renderers.renderer_from_name which has seen use in the wild.

  • Add a clone method to pyramid.renderers.RendererHelper for use by the pyramid.view.view_config decorator.


  • Fixed two typos in wiki2 (SQLA + URL Dispatch) tutorial.

  • Reordered chapters in narrative section for better new user friendliness.

  • Added more indexing markers to sections in documentation.

1.1b4 (2011-07-18)


  • Added a section entitled "Writing a Script" to the "Command-Line Pyramid" chapter.

Backwards Incompatibilities

  • We added the pyramid.scripting.make_request API too hastily in 1.1b3. It has been removed. Sorry for any inconvenience. Use the pyramid.request.Request.blank API instead.


  • The paster pshell, paster pviews, and paster proutes commands each now under the hood uses pyramid.paster.bootstrap, which makes it possible to supply an .ini file without naming the "right" section in the file that points at the actual Pyramid application. Instead, you can generally just run paster {pshell|proutes|pviews} development.ini and it will do mostly the right thing.

Bug Fixes

  • Omit custom environ variables when rendering a custom exception template in pyramid.httpexceptions.WSGIHTTPException._set_default_attrs; stringifying thse may trigger code that should not be executed; see

1.1b3 (2011-07-15)


  • Fix corner case to ease semifunctional testing of views: create a new rendererinfo to clear out old registry on a rescan. See

  • New API class: pyramid.static.static_view. This supersedes the deprecated pyramid.view.static class. pyramid.static.static_view by default serves up documents as the result of the request's path_info, attribute rather than it's subpath attribute (the inverse was true of pyramid.view.static, and still is). pyramid.static.static_view exposes a use_subpath flag for use when you want the static view to behave like the older deprecated version.

  • A new API function pyramid.paster.bootstrap has been added to make writing scripts that bootstrap a Pyramid environment easier, e.g.:

    from pyramid.paster import bootstrap
    info = bootstrap('/path/to/my/development.ini')
    request = info['request']
    print request.route_url('myroute')
  • A new API function pyramid.scripting.prepare has been added. It is a lower-level analogue of pyramid.paster.boostrap that accepts a request and a registry instead of a config file argument, and is used for the same purpose:

    from pyramid.scripting import prepare
    info = prepare(registry=myregistry)
    request = info['request']
    print request.route_url('myroute')
  • A new API function pyramid.scripting.make_request has been added. The resulting request will have a registry attribute. It is meant to be used in conjunction with pyramid.scripting.prepare and/or pyramid.paster.bootstrap (both of which accept a request as an argument):

    from pyramid.scripting import make_request
    request = make_request('/')
  • New API attribute pyramid.config.global_registries is an iterable object that contains references to every Pyramid registry loaded into the current process via pyramid.config.Configurator.make_app. It also has a last attribute containing the last registry loaded. This is used by the scripting machinery, and is available for introspection.


  • The pyramid.view.static class has been deprecated in favor of the newer pyramid.static.static_view class. A deprecation warning is raised when it is used. You should replace it with a reference to pyramid.static.static_view with the use_subpath=True argument.

Bug Fixes

  • Without a mo-file loaded for the combination of domain/locale, pyramid.i18n.Localizer.pluralize run using that domain/locale combination raised an inscrutable "translations object has no attr 'plural'" error. Now, instead it "works" (it uses a germanic pluralization by default). It's nonsensical to try to pluralize something without translations for that locale/domain available, but this behavior matches the behavior of pyramid.i18n.Localizer.translate so it's at least consistent; see

1.1b2 (2011-07-13)


  • New environment setting PYRAMID_PREVENT_HTTP_CACHE and new configuration file value prevent_http_cache. These are synomymous and allow you to prevent HTTP cache headers from being set by Pyramid's http_cache machinery globally in a process. see the "Influencing HTTP Caching" section of the "View Configuration" narrative chapter and the detailed documentation for this setting in the "Environment Variables and Configuration Settings" narrative chapter.

Behavior Changes

  • Previously, If a BeforeRender event subscriber added a value via the __setitem__ or update methods of the event object with a key that already existed in the renderer globals dictionary, a KeyError was raised. With the deprecation of the "add_renderer_globals" feature of the configurator, there was no way to override an existing value in the renderer globals dictionary that already existed. Now, the event object will overwrite an older value that is already in the globals dictionary when its __setitem__ or update is called (as well as the new setdefault method), just like a plain old dictionary. As a result, for maximum interoperability with other third-party subscribers, if you write an event subscriber meant to be used as a BeforeRender subscriber, your subscriber code will now need to (using .get or __contains__ of the event object) ensure no value already exists in the renderer globals dictionary before setting an overriding value.

Bug Fixes

  • The Configurator.add_route method allowed two routes with the same route to be added without an intermediate config.commit(). If you now receive a ConfigurationError at startup time that appears to be add_route related, you'll need to either a) ensure that all of your route names are unique or b) call config.commit() before adding a second route with the name of a previously added name or c) use a Configurator that works in autocommit mode.

  • The pyramid_routesalchemy and pyramid_alchemy scaffolds inappropriately used DBSession.rollback() instead of transaction.abort() in one place.

  • We now clear request.response before we invoke an exception view; an exception view will be working with a request.response that has not been touched by any code prior to the exception.

  • Views associated with routes with spaces in the route name may not have been looked up correctly when using Pyramid with zope.interface 3.6.4 and better. See


  • Wiki2 (SQLAlchemy + URL Dispatch) tutorial models.initialize_sql didn't match the pyramid_routesalchemy scaffold function of the same name; it didn't get synchronized when it was changed in the scaffold.

  • New documentation section in View Configuration narrative chapter: "Influencing HTTP Caching".

1.1b1 (2011-07-10)


  • It is now possible to invoke paster pshell even if the paste ini file section name pointed to in its argument is not actually a Pyramid WSGI application. The shell will work in a degraded mode, and will warn the user. See "The Interactive Shell" in the "Creating a Pyramid Project" narrative documentation section.

  • paster pshell now offers more built-in global variables by default (including app and settings). See "The Interactive Shell" in the "Creating a Pyramid Project" narrative documentation section.

  • It is now possible to add a [pshell] section to your application's .ini configuration file, which influences the global names available to a pshell session. See "Extending the Shell" in the "Creating a Pyramid Project" narrative documentation chapter.

  • The config.scan method has grown a **kw argument. kw argument represents a set of keyword arguments to pass to the Venusian Scanner object created by Pyramid. (See the Venusian documentation for more information about Scanner).

  • New request property: json_body. This property will return the JSON-decoded variant of the request body. If the request body is not well-formed JSON, this property will raise an exception.

  • A new value http_cache can be used as a view configuration parameter.

    When you supply an http_cache value to a view configuration, the Expires and Cache-Control headers of a response generated by the associated view callable are modified. The value for http_cache may be one of the following:

    • A nonzero integer. If it's a nonzero integer, it's treated as a number of seconds. This number of seconds will be used to compute the Expires header and the Cache-Control: max-age parameter of responses to requests which call this view. For example: http_cache=3600 instructs the requesting browser to 'cache this response for an hour, please'.

    • A datetime.timedelta instance. If it's a datetime.timedelta instance, it will be converted into a number of seconds, and that number of seconds will be used to compute the Expires header and the Cache-Control: max-age parameter of responses to requests which call this view. For example: http_cache=datetime.timedelta(days=1) instructs the requesting browser to 'cache this response for a day, please'.

    • Zero (0). If the value is zero, the Cache-Control and Expires headers present in all responses from this view will be composed such that client browser cache (and any intermediate caches) are instructed to never cache the response.

    • A two-tuple. If it's a two tuple (e.g. http_cache=(1, {'public':True})), the first value in the tuple may be a nonzero integer or a datetime.timedelta instance; in either case this value will be used as the number of seconds to cache the response. The second value in the tuple must be a dictionary. The values present in the dictionary will be used as input to the Cache-Control response header. For example: http_cache=(3600, {'public':True}) means 'cache for an hour, and add public to the Cache-Control header of the response'. All keys and values supported by the webob.cachecontrol.CacheControl interface may be added to the dictionary. Supplying {'public':True} is equivalent to calling response.cache_control.public = True.

    Providing a non-tuple value as http_cache is equivalent to calling response.cache_expires(value) within your view's body.

    Providing a two-tuple value as http_cache is equivalent to calling response.cache_expires(value[0], **value[1]) within your view's body.

    If you wish to avoid influencing, the Expires header, and instead wish to only influence Cache-Control headers, pass a tuple as http_cache with the first element of None, e.g.: (None, {'public':True}).

Bug Fixes

  • Framework wrappers of the original view (such as http_cached and so on) relied on being able to trust that the response they were receiving was an IResponse. It wasn't always, because the response was resolved by the router instead of early in the view wrapping process. This has been fixed.


  • Added a section in the "Webob" chapter named "Dealing With A JSON-Encoded Request Body" (usage of request.json_body).

Behavior Changes

  • The paster pshell, paster proutes, and paster pviews commands now take a single argument in the form /path/to/config.ini#sectionname rather than the previous 2-argument spelling /path/to/config.ini sectionname. #sectionname may be omitted, in which case #main is assumed.

1.1a4 (2011-07-01)

Bug Fixes

  • pyramid.testing.DummyRequest now raises deprecation warnings when attributes deprecated for pyramid.request.Request are accessed (like response_content_type). This is for the benefit of folks running unit tests which use DummyRequest instead of a "real" request, so they know things are deprecated without necessarily needing a functional test suite.

  • The directive behaved contrary to the documentation when passed more than one interface object to its constructor. For example, when the following listener was registered:

    @subscriber(IFoo, IBar)
    def expects_ifoo_events_and_ibar_events(event):
        print event

    The Events chapter docs claimed that the listener would be registered and listening for both IFoo and IBar events. Instead, it registered an "object event" subscriber which would only be called if an IObjectEvent was emitted where the object interface was IFoo and the event interface was IBar.

    The behavior now matches the documentation. If you were relying on the buggy behavior of the 1.0 subscriber directive in order to register an object event subscriber, you must now pass a sequence to indicate you'd like to register a subscriber for an object event. e.g.:

    @subscriber([IFoo, IBar])
    def expects_object_event(object, event):
        print object, event


  • Add JSONP renderer (see "JSONP renderer" in the Renderers chapter of the documentation).


  • Deprecated the set_renderer_globals_factory method of the Configurator and the renderer_globals Configurator constructor parameter.


  • The Wiki and Wiki2 tutorial "Tests" chapters each had two bugs: neither did told the user to depend on WebTest, and 2 tests failed in each as the result of changes to Pyramid itself. These issues have been fixed.

  • Move 1.0.X CHANGES.txt entries to HISTORY.txt.

1.1a3 (2011-06-26)


Bug fixes

  • Pyramid would raise an AttributeError in the Configurator when attempting to set a __text__ attribute on a custom predicate that was actually a classmethod. See .

  • Accessing or setting deprecated response_* attrs on request (e.g. response_content_type) now issues a deprecation warning at access time rather than at rendering time.

1.1a2 (2011-06-22)

Bug Fixes

  • 1.1a1 broke Akhet by not providing a backwards compatibility import shim for pyramid.paster.PyramidTemplate. Now one has been added, although a deprecation warning is emitted when Akhet imports it.

  • If multiple specs were provided in a single call to config.add_translation_dirs, the directories were inserted into the beginning of the directory list in the wrong order: they were inserted in the reverse of the order they were provided in the *specs list (items later in the list were added before ones earlier in the list). This is now fixed.

Backwards Incompatibilities

  • The pyramid Router attempted to set a value into the key environ['repoze.bfg.message'] when it caught a view-related exception for backwards compatibility with applications written for repoze.bfg during error handling. It did this by using code that looked like so:

    # "why" is an exception object
        msg = why[0]
        msg = ''
    environ['repoze.bfg.message'] = msg

    Use of the value environ['repoze.bfg.message'] was docs-deprecated in Pyramid 1.0. Our standing policy is to not remove features after a deprecation for two full major releases, so this code was originally slated to be removed in Pyramid 1.2. However, computing the repoze.bfg.message value was the source of at least one bug found in the wild (, and there isn't a foolproof way to both preserve backwards compatibility and to fix the bug. Therefore, the code which sets the value has been removed in this release. Code in exception views which relies on this value's presence in the environment should now use the exception attribute of the request (e.g. request.exception[0]) to retrieve the message instead of relying on request.environ['repoze.bfg.message'].

1.1a1 (2011-06-20)


  • The term "template" used to refer to both "paster templates" and "rendered templates" (templates created by a rendering engine. i.e. Mako, Chameleon, Jinja, etc.). "Paster templates" will now be referred to as "scaffolds", whereas the name for "rendered templates" will remain as "templates."

  • The wiki (ZODB+Traversal) tutorial was updated slightly.

  • The wiki2 (SQLA+URL Dispatch) tutorial was updated slightly.

  • Make pyramid.interfaces.IAuthenticationPolicy and pyramid.interfaces.IAuthorizationPolicy public interfaces, and refer to them within the pyramid.authentication and pyramid.authorization API docs.

  • Render the function definitions for each exposed interface in pyramid.interfaces.

  • Add missing docs reference to pyramid.config.Configurator.set_view_mapper and refer to it within Hooks chapter section named "Using a View Mapper".

  • Added section to the "Environment Variables and .ini File Settings" chapter in the narrative documentation section entitled "Adding a Custom Setting".

  • Added documentation for a "multidict" (e.g. the API of request.POST) as interface API documentation.

  • Added a section to the "URL Dispatch" narrative chapter regarding the new "static" route feature.

  • Added "What's New in Pyramid 1.1" to HTML rendering of documentation.

  • Added API docs for pyramid.authentication.SessionAuthenticationPolicy.

  • Added API docs for pyramid.httpexceptions.exception_response.

  • Added "HTTP Exceptions" section to Views narrative chapter including a description of pyramid.httpexceptions.exception_response.


  • Add support for language fallbacks: when trying to translate for a specific territory (such as en_GB) fall back to translations for the language (ie en). This brings the translation behaviour in line with GNU gettext and fixes partially translated texts when using C extensions.

  • New authentication policy: pyramid.authentication.SessionAuthenticationPolicy, which uses a session to store credentials.

  • Accessing the response attribute of a pyramid.request.Request object (e.g. request.response within a view) now produces a new pyramid.response.Response object. This feature is meant to be used mainly when a view configured with a renderer needs to set response attributes: all renderers will use the Response object implied by request.response as the response object returned to the router.

    request.response can also be used by code in a view that does not use a renderer, however the response object that is produced by request.response must be returned when a renderer is not in play (it is not a "global" response).

  • Integers and longs passed as elements to pyramid.url.resource_url or pyramid.request.Request.resource_url e.g. resource_url(context, request, 1, 2) (1 and 2 are the elements) will now be converted implicitly to strings in the result. Previously passing integers or longs as elements would cause a TypeError.

  • pyramid_alchemy paster template now uses query.get rather than query.filter_by to take better advantage of identity map caching.

  • pyramid_alchemy paster template now has unit tests.

  • Added pyramid.i18n.make_localizer API (broken out from get_localizer guts).

  • An exception raised by a NewRequest event subscriber can now be caught by an exception view.

  • It is now possible to get information about why Pyramid raised a Forbidden exception from within an exception view. The ACLDenied object returned by the permits method of each stock authorization policy (pyramid.interfaces.IAuthorizationPolicy.permits) is now attached to the Forbidden exception as its result attribute. Therefore, if you've created a Forbidden exception view, you can see the ACE, ACL, permission, and principals involved in the request as eg. context.result.permission, context.result.acl, etc within the logic of the Forbidden exception view.

  • Don't explicitly prevent the timeout from being lower than the reissue_time when setting up an AuthTktAuthenticationPolicy (previously such a configuration would raise a ValueError, now it's allowed, although typically nonsensical). Allowing the nonsensical configuration made the code more understandable and required fewer tests.

  • A new paster command named paster pviews was added. This command prints a summary of potentially matching views for a given path. See the section entitled "Displaying Matching Views for a Given URL" in the "View Configuration" chapter of the narrative documentation for more information.

  • The add_route method of the Configurator now accepts a static argument. If this argument is True, the added route will never be considered for matching when a request is handled. Instead, it will only be useful for URL generation via route_url and route_path. See the section entitled "Static Routes" in the URL Dispatch narrative chapter for more information.

  • A default exception view for the context pyramid.interfaces.IExceptionResponse is now registered by default. This means that an instance of any exception response class imported from pyramid.httpexceptions (such as HTTPFound) can now be raised from within view code; when raised, this exception view will render the exception to a response.

  • A function named pyramid.httpexceptions.exception_response is a shortcut that can be used to create HTTP exception response objects using an HTTP integer status code.

  • The Configurator now accepts an additional keyword argument named exceptionresponse_view. By default, this argument is populated with a default exception view function that will be used when a response is raised as an exception. When None is passed for this value, an exception view for responses will not be registered. Passing None returns the behavior of raising an HTTP exception to that of Pyramid 1.0 (the exception will propagate to middleware and to the WSGI server).

  • The pyramid.request.Request class now has a ResponseClass interface which points at pyramid.response.Response.

  • The pyramid.response.Response class now has a RequestClass interface which points at pyramid.request.Request.

  • It is now possible to return an arbitrary object from a Pyramid view callable even if a renderer is not used, as long as a suitable adapter to pyramid.interfaces.IResponse is registered for the type of the returned object by using the new pyramid.config.Configurator.add_response_adapter API. See the section in the Hooks chapter of the documentation entitled "Changing How Pyramid Treats View Responses".

  • The Pyramid router will now, by default, call the __call__ method of WebOb response objects when returning a WSGI response. This means that, among other things, the conditional_response feature of WebOb response objects will now behave properly.

  • New method named pyramid.request.Request.is_response. This method should be used instead of the pyramid.view.is_response function, which has been deprecated.

Bug Fixes

  • URL pattern markers used in URL dispatch are permitted to specify a custom regex. For example, the pattern /{foo:\d+} means to match /12345 (foo==12345 in the match dictionary) but not /abc. However, custom regexes in a pattern marker which used squiggly brackets did not work. For example, /{foo:\d{4}} would fail to match /1234 and /{foo:\d{1,2}} would fail to match /1 or /11. One level of inner squiggly brackets is now recognized so that the prior two patterns given as examples now work. See also

  • Don't send port numbers along with domain information in cookies set by AuthTktCookieHelper (see

  • pyramid.url.route_path (and the shortcut pyramid.request.Request.route_url method) now include the WSGI SCRIPT_NAME at the front of the path if it is not empty (see

  • pyramid.testing.DummyRequest now has a script_name attribute (the empty string).

  • Don't quote :@&+$, symbols in *elements passed to pyramid.url.route_url or pyramid.url.resource_url (see

  • Include SCRIPT_NAME in redirects issued by pyramid.view.append_slash_notfound_view (see

  • Static views registered with config.add_static_view which also included a permission keyword argument would not work as expected, because add_static_view also registered a route factory internally. Because a route factory was registered internally, the context checked by the Pyramid permission machinery never had an ACL. add_static_view no longer registers a route with a factory, so the default root factory will be used.

  • config.add_static_view now passes extra keyword arguments it receives to config.add_route (calling add_static_view is mostly logically equivalent to adding a view of the type pyramid.static.static_view hooked up to a route with a subpath). This makes it possible to pass e.g., factory= to add_static_view to protect a particular static view with a custom ACL.

  • testing.DummyRequest used the wrong registry (the global registry) as self.registry if a dummy request was created before testing.setUp was executed (testing.setUp pushes a local registry onto the threadlocal stack). Fixed by implementing registry as a property for DummyRequest instead of eagerly assigning an attribute. See also

  • When visiting a URL that represented a static view which resolved to a subdirectory, the index.html of that subdirectory would not be served properly. Instead, a redirect to /subdir would be issued. This has been fixed, and now visiting a subdirectory that contains an index.html within a static view returns the index.html properly. See also

  • Redirects issued by a static view did not take into account any existing SCRIPT_NAME (such as one set by a url mapping composite). Now they do.

  • The pyramid.wsgi.wsgiapp2 decorator did not take into account the SCRIPT_NAME in the origin request.

  • The pyramid.wsgi.wsgiapp2 decorator effectively only worked when it decorated a view found via traversal; it ignored the PATH_INFO that was part of a url-dispatch-matched view.


  • Deprecated all assignments to request.response_* attributes (for example request.response_content_type = 'foo' is now deprecated). Assignments and mutations of assignable request attributes that were considered by the framework for response influence are now deprecated: response_content_type, response_headerlist, response_status, response_charset, and response_cache_for. Instead of assigning these to the request object for later detection by the rendering machinery, users should use the appropriate API of the Response object created by accessing request.response (e.g. code which does request.response_content_type = 'abc' should be changed to request.response.content_type = 'abc').

  • Passing view-related parameters to pyramid.config.Configurator.add_route is now deprecated. Previously, a view was permitted to be connected to a route using a set of view* parameters passed to the add_route method of the Configurator. This was a shorthand which replaced the need to perform a subsequent call to add_view. For example, it was valid (and often recommended) to do:

    config.add_route('home', '/', view='mypackage.views.myview',

    Passing view* arguments to add_route is now deprecated in favor of connecting a view to a predefined route via Configurator.add_view using the route's route_name parameter. As a result, the above example should now be spelled:

    config.add_route('home', '/')
    config.add_view('mypackage.views.myview', route_name='home')

    This deprecation was done to reduce confusion observed in IRC, as well as to (eventually) reduce documentation burden (see also A deprecation warning is now issued when any view-related parameter is passed to Configurator.add_route.

  • Passing an environ dictionary to the __call__ method of a "traverser" (e.g. an object that implements pyramid.interfaces.ITraverser such as an instance of pyramid.traversal.ResourceTreeTraverser) as its request argument now causes a deprecation warning to be emitted. Consumer code should pass a request object instead. The fact that passing an environ dict is permitted has been documentation-deprecated since repoze.bfg 1.1, and this capability will be removed entirely in a future version.

  • The following (undocumented, dictionary-like) methods of the pyramid.request.Request object have been deprecated: __contains__, __delitem__, __getitem__, __iter__, __setitem__, get, has_key, items, iteritems, itervalues, keys, pop, popitem, setdefault, update, and values. Usage of any of these methods will cause a deprecation warning to be emitted. These methods were added for internal compatibility in repoze.bfg 1.1 (code that currently expects a request object expected an environ object in BFG 1.0 and before). In a future version, these methods will be removed entirely.

  • Deprecated pyramid.view.is_response function in favor of (newly-added) pyramid.request.Request.is_response method. Determining if an object is truly a valid response object now requires access to the registry, which is only easily available as a request attribute. The pyramid.view.is_response function will still work until it is removed, but now may return an incorrect answer under some (very uncommon) circumstances.

Behavior Changes

  • The default Mako renderer is now configured to escape all HTML in expression tags. This is intended to help prevent XSS attacks caused by rendering unsanitized input from users. To revert this behavior in user's templates, they need to filter the expression through the 'n' filter. For example, ${ myhtml | n }. See

  • A custom request factory is now required to return a request object that has a response attribute (or "reified"/lazy property) if they the request is meant to be used in a view that uses a renderer. This response attribute should be an instance of the class pyramid.response.Response.

  • The JSON and string renderer factories now assign to request.response.content_type rather than request.response_content_type.

  • Each built-in renderer factory now determines whether it should change the content type of the response by comparing the response's content type against the response's default content type; if the content type is the default content type (usually text/html), the renderer changes the content type (to application/json or text/plain for JSON and string renderers respectively).

  • The pyramid.wsgi.wsgiapp2 now uses a slightly different method of figuring out how to "fix" SCRIPT_NAME and PATH_INFO for the downstream application. As a result, those values may differ slightly from the perspective of the downstream application (for example, SCRIPT_NAME will now never possess a trailing slash).

  • Previously, pyramid.request.Request inherited from webob.request.Request and implemented __getattr__, __setattr__ and __delattr__ itself in order to overidde "adhoc attr" WebOb behavior where attributes of the request are stored in the environ. Now, pyramid.request.Request object inherits from (the more recent) webob.request.BaseRequest instead of webob.request.Request, which provides the same behavior. pyramid.request.Request no longer implements its own __getattr__, __setattr__ or __delattr__ as a result.

  • pyramid.response.Response is now a subclass of webob.response.Response (in order to directly implement the pyramid.interfaces.IResponse interface).

  • The "exception response" objects importable from pyramid.httpexceptions (e.g. HTTPNotFound) are no longer just import aliases for classes that actually live in webob.exc. Instead, we've defined our own exception classes within the module that mirror and emulate the webob.exc exception response objects almost entirely. See the "Design Defense" doc section named "Pyramid Uses its Own HTTP Exception Classes" for more information.

Backwards Incompatibilities

  • Pyramid no longer supports Python 2.4. Python 2.5 or better is required to run Pyramid 1.1+.

  • The Pyramid router now, by default, expects response objects returned from view callables to implement the pyramid.interfaces.IResponse interface. Unlike the Pyramid 1.0 version of this interface, objects which implement IResponse now must define a __call__ method that accepts environ and start_response, and which returns an app_iter iterable, among other things. Previously, it was possible to return any object which had the three WebOb app_iter, headerlist, and status attributes as a response, so this is a backwards incompatibility. It is possible to get backwards compatibility back by registering an adapter to IResponse from the type of object you're now returning from view callables. See the section in the Hooks chapter of the documentation entitled "Changing How Pyramid Treats View Responses".

  • The pyramid.interfaces.IResponse interface is now much more extensive. Previously it defined only app_iter, status and headerlist; now it is basically intended to directly mirror the webob.Response API, which has many methods and attributes.

  • The pyramid.httpexceptions classes named HTTPFound, HTTPMultipleChoices, HTTPMovedPermanently, HTTPSeeOther, HTTPUseProxy, and HTTPTemporaryRedirect now accept location as their first positional argument rather than detail. This means that you can do, e.g. return pyramid.httpexceptions.HTTPFound('http://foo') rather than return pyramid.httpexceptions.HTTPFound(location='http//foo') (the latter will of course continue to work).


  • Pyramid now depends on WebOb >= 1.0.2 as tests depend on the bugfix in that release: "Fix handling of WSGI environs with missing SCRIPT_NAME". (Note that in reality, everyone should probably be using 1.0.4 or better though, as WebOb 1.0.2 and 1.0.3 were effectively brownbag releases.)

1.0 (2011-01-30)


  • Fixed bug in ZODB Wiki tutorial (missing dependency on docutils in "models" step within

  • Removed API documentation for pyramid.testing APIs named registerDummySecurityPolicy, registerResources, registerModels, registerEventListener, registerTemplateRenderer, registerDummyRenderer, registerView, registerUtility, registerAdapter, registerSubscriber, registerRoute, and registerSettings.

  • Moved "Using ZODB With ZEO" and "Using repoze.catalog Within Pyramid" tutorials out of core documentation and into the Pyramid Tutorials site (

  • Changed "Cleaning up After a Request" section in the URL Dispatch chapter to use request.add_finished_callback instead of jamming an object with a __del__ into the WSGI environment.

  • Remove duplication of add_route API documentation from URL Dispatch narrative chapter.

  • Remove duplication of API and narrative documentation in pyramid.view.view_config API docs by pointing to pyramid.config.add_view documentation and narrative chapter documentation.

  • Removed some API documentation duplicated in narrative portions of documentation

  • Removed "Overall Flow of Authentication" from SQLAlchemy + URL Dispatch wiki tutorial due to print space concerns (moved to Pyramid Tutorials site).

Bug Fixes

  • Deprecated-since-BFG-1.2 APIs from pyramid.testing now properly emit deprecation warnings.

  • Added egg:repoze.retry#retry middleware to the WSGI pipeline in ZODB templates (retry ZODB conflict errors which occur in normal operations).

  • Removed duplicate implementations of is_response. Two competing implementations existed: one in pyramid.config and one in pyramid.view. Now the one defined in pyramid.view is used internally by pyramid.config and continues to be advertised as an API.

1.0b3 (2011-01-28)

Bug Fixes

  • Use &copy; instead of copyright symbol in paster templates / tutorial templates for the benefit of folks who cutnpaste and save to a non-UTF8 format.

  • pyramid.view.append_slash_notfound_view now preserves GET query parameters across redirects.


  • Beef up documentation related to set_default_permission: explicitly mention that default permissions also protect exception views.

  • Paster templates and tutorials now use spaces instead of tabs in their HTML templates.

1.0b2 (2011-01-24)

Bug Fixes

  • The production.ini generated by all paster templates now have an effective logging level of WARN, which prevents e.g. SQLAlchemy statement logging and other inappropriate output.

  • The production.ini of the pyramid_routesalchemy and pyramid_alchemy paster templates did not have a sqlalchemy logger section, preventing paster serve production.ini from working.

  • The pyramid_routesalchemy and pyramid_alchemy paster templates used the {{package}} variable in a place where it should have used the {{project}} variable, causing applications created with uppercase letters e.g. paster create -t pyramid_routesalchemy Dibbus to fail to start when paster serve development.ini was used against the result. See

  • The render_view method of pyramid.renderers.RendererHelper passed an incorrect value into the renderer for renderer_info. It now passes an instance of RendererHelper instead of a dictionary, which is consistent with other usages. See

  • A bug existed in the pyramid.authentication.AuthTktCookieHelper which would break any usage of an AuthTktAuthenticationPolicy when one was configured to reissue its tokens (reissue_time < timeout / max_age). Symptom: ValueError: ('Invalid token %r', ''). See

1.0b1 (2011-01-21)


  • The AuthTktAuthenticationPolicy now accepts a tokens parameter via The value must be a sequence of strings. Tokens are placed into the auth_tkt "tokens" field and returned in the auth_tkt cookie.

  • Add wild_domain argument to AuthTktAuthenticationPolicy, which defaults to True. If it is set to False, the feature of the policy which sets a cookie with a wildcard domain will be turned off.

  • Add a file to each paster template. See

Bug Fixes

  • testing.setUp now adds a settings attribute to the registry (both when it's passed a registry without any settings and when it creates one).

  • The testing.setUp function now takes a settings argument, which should be a dictionary. Its values will subsequently be available on the returned config object as config.registry.settings.


  • Added "What's New in Pyramid 1.0" chapter to HTML rendering of documentation.

  • Merged caseman-master narrative editing branch, many wording fixes and extensions.

  • Fix deprecated example showing chameleon_zpt API call in testing narrative chapter.

  • Added "Adding Methods to the Configurator via add_directive" section to Advanced Configuration narrative chapter.

  • Add docs for add_finished_callback, add_response_callback, route_path, route_url, and static_url methods to pyramid.request.Request API docs.

  • Add (minimal) documentation about using I18N within Mako templates to "Internationalization and Localization" narrative chapter.

  • Move content of "Forms" chapter back to "Views" chapter; I can't think of a better place to put it.

  • Slightly improved interface docs for IAuthorizationPolicy.

  • Minimally explain usage of custom regular expressions in URL dispatch replacement markers within URL Dispatch chapter.


  • Using the pyramid.view.bfg_view alias for pyramid.view.view_config (a backwards compatibility shim) now issues a deprecation warning.

Backwards Incompatibilities

  • Using testing.setUp now registers an ISettings utility as a side effect. Some test code which queries for this utility after testing.setUp via queryAdapter will expect a return value of None. This code will need to be changed.

  • When a pyramid.exceptions.Forbidden error is raised, its status code now 403 Forbidden. It was previously 401 Unauthorized, for backwards compatibility purposes with repoze.bfg. This change will cause problems for users of Pyramid with repoze.who, which intercepts 401 Unauthorized by default, but allows 403 Forbidden to pass through. Those deployments will need to configure repoze.who to also react to 403 Forbidden.

  • The default value for the cookie_on_exception parameter to pyramid.session.UnencryptedCookieSessionFactory is now True. This means that when view code causes an exception to be raised, and the session has been mutated, a cookie will be sent back in the response. Previously its default value was False.

Paster Templates

  • The pyramid_zodb, pyramid_routesalchemy and pyramid_alchemy paster templates now use a default "commit veto" hook when configuring the repoze.tm2 transaction manager in development.ini. This prevents a transaction from being committed when the response status code is within the 400 or 500 ranges. See also

1.0a10 (2011-01-18)

Bug Fixes

  • URL dispatch now properly handles a .* or * appearing in a regex match when used inside brackets. Resolves issue #90.

Backwards Incompatibilities

  • The add_handler method of a Configurator has been removed from the Pyramid core. Handlers are now a feature of the pyramid_handlers package, which can be downloaded from PyPI. Documentation for the package should be available via, which describes how to add a configuration statement to your main block to reobtain this method. You will also need to add an install_requires dependency upon pyramid_handlers to your file.

  • The load_zcml method of a Configurator has been removed from the Pyramid core. Loading ZCML is now a feature of the pyramid_zcml package, which can be downloaded from PyPI. Documentation for the package should be available via, which describes how to add a configuration statement to your main block to reobtain this method. You will also need to add an install_requires dependency upon pyramid_zcml to your file.

  • The pyramid.includes subpackage has been removed. ZCML files which use include the package pyramid.includes (e.g. <include package="pyramid.includes"/>) now must include the pyramid_zcml package instead (e.g. <include package="pyramid_zcml"/>).

  • The pyramid.view.action decorator has been removed from the Pyramid core. Handlers are now a feature of the pyramid_handlers package. It should now be imported from pyramid_handlers e.g. from pyramid_handlers import action.

  • The handler ZCML directive has been removed. It is now a feature of the pyramid_handlers package.

  • The pylons_minimal, pylons_basic and pylons_sqla paster templates were removed. Use pyramid_sqla (available from PyPI) as a generic replacement for Pylons-esque development.

  • The make_app function has been removed from the pyramid.router module. It continues life within the pyramid_zcml package. This leaves the pyramid.router module without any API functions.

  • The configure_zcml setting within the deployment settings (within **settings passed to a Pyramid main function) has ceased to have any meaning.


  • pyramid.testing.setUp and pyramid.testing.tearDown have been undeprecated. They are now the canonical setup and teardown APIs for test configuration, replacing "direct" creation of a Configurator. This is a change designed to provide a facade that will protect against any future Configurator deprecations.

  • Add charset attribute to pyramid.testing.DummyRequest (unconditionally UTF-8).

  • Add add_directive method to configurator, which allows framework extenders to add methods to the configurator (ala ZCML directives).

  • When Configurator.include is passed a module as an argument, it defaults to attempting to find and use a callable named includeme within that module. This makes it possible to use config.include('some.module') rather than config.include('some.module.somefunc') as long as the include function within some.module is named includeme.

  • The bfg2pyramid script now converts ZCML include tags that have repoze.bfg.includes as a package attribute to the value pyramid_zcml. For example, <include package="repoze.bfg.includes"> will be converted to <include package="pyramid_zcml">.

Paster Templates

  • All paster templates now use pyramid.testing.setUp and pyramid.testing.tearDown rather than creating a Configurator "by hand" within their module, as per decision in features above.

  • The starter_zcml paster template has been moved to the pyramid_zcml package.


  • The wiki and wiki2 tutorials now use pyramid.testing.setUp and pyramid.testing.tearDown rather than creating a Configurator "by hand", as per decision in features above.

  • The "Testing" narrative chapter now explains pyramid.testing.setUp and pyramid.testing.tearDown instead of Configurator creation and Configurator.begin() and Configurator.end().

  • Document the request.override_renderer attribute within the narrative "Renderers" chapter in a section named "Overriding A Renderer at Runtime".

  • The "Declarative Configuration" narrative chapter has been removed (it was moved to the pyramid_zcml package).

  • Most references to ZCML in narrative chapters have been removed or redirected to pyramid_zcml locations.


  • Deprecation warnings related to import of the following API functions were added: pyramid.traversal.find_model, pyramid.traversal.model_path, pyramid.traversal.model_path_tuple, pyramid.url.model_url. The instructions emitted by the deprecation warnings instruct the developer to change these method spellings to their resource equivalents. This is a consequence of the mass concept rename of "model" to "resource" performed in 1.0a7.

1.0a9 (2011-01-08)

Bug Fixes

  • The proutes command tried too hard to resolve the view for printing, resulting in exceptions when an exceptional root factory was encountered. Instead of trying to resolve the view, if it cannot, it will now just print <unknown>.

  • The self argument was included in new methods of the ISession interface signature, causing pyramid_beaker tests to fail.

  • Readd pyramid.traversal.model_path_tuple as an alias for pyramid.traversal.resource_path_tuple for backwards compatibility.


  • Add a new API pyramid.url.current_route_url, which computes a URL based on the "current" route (if any) and its matchdict values.

  • config.add_view now accepts a decorator keyword argument, a callable which will decorate the view callable before it is added to the registry.

  • If a handler class provides an __action_decorator__ attribute (usually a classmethod or staticmethod), use that as the decorator for each view registration for that handler.

  • The pyramid.interfaces.IAuthenticationPolicy interface now specifies an unauthenticated_userid method. This method supports an important optimization required by people who are using persistent storages which do not support object caching and whom want to create a "user object" as a request attribute.

  • A new API has been added to the module named unauthenticated_userid. This API function calls the unauthenticated_userid method of the effective security policy.

  • An unauthenticated_userid method has been added to the dummy authentication policy returned by pyramid.config.Configurator.testing_securitypolicy. It returns the same thing as that the dummy authentication policy's authenticated_userid method.

  • The class pyramid.authentication.AuthTktCookieHelper is now an API. This class can be used by third-party authentication policy developers to help in the mechanics of authentication cookie-setting.

  • New constructor argument to Configurator: default_view_mapper. Useful to create systems that have alternate view calling conventions. A view mapper allows objects that are meant to be used as view callables to have an arbitrary argument list and an arbitrary result. The object passed as default_view_mapper should implement the pyramid.interfaces.IViewMapperFactory interface.

  • add a set_view_mapper API to Configurator. Has the same result as passing default_view_mapper to the Configurator constructor.

  • config.add_view now accepts a mapper keyword argument, which should either be None, a string representing a Python dotted name, or an object which is an IViewMapperFactory. This feature is not useful for "civilians", only for extension writers.

  • Allow static renderer provided during view registration to be overridden at request time via a request attribute named override_renderer, which should be the name of a previously registered renderer. Useful to provide "omnipresent" RPC using existing rendered views.

  • Instances of pyramid.testing.DummyRequest now have a session object, which is mostly a dictionary, but also implements the other session API methods for flash and CSRF.

Backwards Incompatibilities

  • Since the pyramid.interfaces.IAuthenticationPolicy interface now specifies that a policy implementation must implement an unauthenticated_userid method, all third-party custom authentication policies now must implement this method. It, however, will only be called when the global function named is invoked, so if you're not invoking that, you will not notice any issues.

  • pyramid.interfaces.ISession.get_csrf_token now mandates that an implementation should return a new token if one doesn't already exist in the session (previously it would return None). The internal sessioning implementation has been changed.


  • The (weak) "Converting a CMF Application to Pyramid" tutorial has been removed from the tutorials section. It was moved to the pyramid_tutorials Github repository.

  • The "Resource Location and View Lookup" chapter has been replaced with a variant of Rob Miller's "Much Ado About Traversal" (originally published at

  • Many minor wording tweaks and refactorings (merged Casey Duncan's docs fork, in which he is working on general editing).

  • Added (weak) description of new view mapper feature to Hooks narrative chapter.

  • Split views chapter into 2: View Callables and View Configuration.

  • Reorder Renderers and Templates chapters after View Callables but before View Configuration.

  • Merge Session Objects, Cross-Site Request Forgery, and Flash Messaging chapter into a single Sessions chapter.

  • The Wiki and Wiki2 tutorials now have much nicer CSS and graphics.


  • The "view derivation" code is now factored into a set of classes rather than a large number of standalone functions (a side effect of the view mapper refactoring).

  • The pyramid.renderer.RendererHelper class has grown a render_view method, which is used by the default view mapper (a side effect of the view mapper refactoring).

  • The object passed as renderer to the "view deriver" is now an instance of pyramid.renderers.RendererHelper rather than a dictionary (a side effect of view mapper refactoring).

  • The class used as the "page template" in pyramid.chameleon_text was removed, in preference to using a Chameleon-inbuilt version.

  • A view callable wrapper registered in the registry now contains an __original_view__ attribute which references the original view callable (or class).

  • The (non-API) method of all internal authentication policy implementations previously named _get_userid is now named unauthenticated_userid, promoted to an API method. If you were overriding this method, you'll now need to override it as unauthenticated_userid instead.

  • Remove (non-API) function of named _map_view.

1.0a8 (2010-12-27)

Bug Fixes

  • The name registry was not available in the paster pshell environment under IPython.


  • If a resource implements a __resource_url__ method, it will be called as the result of invoking the pyramid.url.resource_url function to generate a URL, overriding the default logic. See the new "Generating The URL Of A Resource" section within the Resources narrative chapter.

  • Added flash messaging, as described in the "Flash Messaging" narrative documentation chapter.

  • Added CSRF token generation, as described in the narrative chapter entitled "Preventing Cross-Site Request Forgery Attacks".

  • Prevent misunderstanding of how the view and view_permission arguments to add_route work by raising an exception during configuration if view-related arguments exist but no view argument is passed.

  • Add paster proute command which displays a summary of the routing table. See the narrative documentation section within the "URL Dispatch" chapter entitled "Displaying All Application Routes".

Paster Templates

  • The pyramid_zodb Paster template no longer employs ZCML. Instead, it is based on scanning.


  • Added "Generating The URL Of A Resource" section to the Resources narrative chapter (includes information about overriding URL generation using __resource_url__).

  • Added "Generating the Path To a Resource" section to the Resources narrative chapter.

  • Added "Finding a Resource by Path" section to the Resources narrative chapter.

  • Added "Obtaining the Lineage of a Resource" to the Resources narrative chapter.

  • Added "Determining if a Resource is In The Lineage of Another Resource" to Resources narrative chapter.

  • Added "Finding the Root Resource" to Resources narrative chapter.

  • Added "Finding a Resource With a Class or Interface in Lineage" to Resources narrative chapter.

  • Added a "Flash Messaging" narrative documentation chapter.

  • Added a narrative chapter entitled "Preventing Cross-Site Request Forgery Attacks".

  • Changed the "ZODB + Traversal Wiki Tutorial" based on changes to pyramid_zodb Paster template.

  • Added "Advanced Configuration" narrative chapter which documents how to deal with configuration conflicts, two-phase configuration, include and commit.

  • Fix API documentation rendering for pyramid.view.static

  • Add "Pyramid Provides More Than One Way to Do It" to Design Defense documentation.

  • Changed "Static Assets" narrative chapter: clarify that name represents a prefix unless it's a URL, added an example of a root-relative static view fallback for URL dispatch, added an example of creating a simple view that returns the body of a file.

  • Move ZCML usage in Hooks chapter to Declarative Configuration chapter.

  • Merge "Static Assets" chapter into the "Assets" chapter.

  • Added narrative documentation section within the "URL Dispatch" chapter entitled "Displaying All Application Routes" (for paster proutes command).

1.0a7 (2010-12-20)

Terminology Changes

  • The Pyramid concept previously known as "model" is now known as "resource". As a result:

    • The following API changes have been made:

      pyramid.url.model_url ->
      pyramid.traversal.find_model ->
      pyramid.traversal.model_path ->
      pyramid.traversal.model_path_tuple ->
      pyramid.traversal.ModelGraphTraverser ->
      pyramid.config.Configurator.testing_models ->
      pyramid.testing.registerModels ->
      pyramid.testing.DummyModel ->
    • All documentation which previously referred to "model" now refers to "resource".

    • The starter and starter_zcml paster templates now have a module instead of a module.

    • Positional argument names of various APIs have been changed from model to resource.

    Backwards compatibility shims have been left in place in all cases. They will continue to work "forever".

  • The Pyramid concept previously known as "resource" is now known as "asset". As a result:

    • The (non-API) module previously known as pyramid.resource is now known as pyramid.asset.

    • All docs that previously referred to "resource specification" now refer to "asset specification".

    • The following API changes were made:

      pyramid.config.Configurator.absolute_resource_spec ->
      pyramid.config.Configurator.override_resource ->
    • The ZCML directive previously known as resource is now known as asset.

    • The setting previously known as BFG_RELOAD_RESOURCES (envvar) or reload_resources (config file) is now known, respectively, as PYRAMID_RELOAD_ASSETS and reload_assets.

    Backwards compatibility shims have been left in place in all cases. They will continue to work "forever".

Bug Fixes

  • Make it possible to successfully run all tests via nosetests command directly (rather than indirectly via python nosetests).

  • When a configuration conflict is encountered during scanning, the conflict exception now shows the decorator information that caused the conflict.


  • Added debug_routematch configuration setting that logs matched routes (including the matchdict and predicates).

  • The name registry is now available in a pshell environment by default. It is the application registry object.


  • All environment variables which used to be prefixed with BFG_ are now prefixed with PYRAMID_ (e.g. BFG_DEBUG_NOTFOUND is now PYRAMID_DEBUG_NOTFOUND)


  • Added "Debugging Route Matching" section to the urldispatch narrative documentation chapter.

  • Added reference to PYRAMID_DEBUG_ROUTEMATCH envvar and debug_routematch config file setting to the Environment narrative docs chapter.

  • Changed "Project" chapter slightly to expand on use of paster pshell.

  • Direct Jython users to Mako rather than Jinja2 in "Install" narrative chapter.

  • Many changes to support terminological renaming of "model" to "resource" and "resource" to "asset".

  • Added an example of WebTest functional testing to the testing narrative chapter.

  • Rearranged chapter ordering by popular demand (URL dispatch first, then traversal). Put hybrid chapter after views chapter.

  • Split off "Renderers" as its own chapter from "Views" chapter in narrative documentation.

Paster Templates

  • Added debug_routematch = false to all paster templates.


  • Depend on Venusian >= 0.5 (for scanning conflict exception decoration).

1.0a6 (2010-12-15)

Bug Fixes

  • 1.0a5 introduced a bug when pyramid.config.Configurator.scan was used without a package argument (e.g. config.scan() as opposed to config.scan('packagename'). The symptoms were: lots of deprecation warnings printed to the console about imports of deprecated Pyramid functions and classes and non-detection of view callables decorated with view_config decorators. This has been fixed.

  • Tests now pass on Windows (no bugs found, but a few tests in the test suite assumed UNIX path segments in filenames).


  • If you followed it to-the-letter, the ZODB+Traversal Wiki tutorial would instruct you to run a test which would fail because the view callable generated by the pyramid_zodb tutorial used a one-arg view callable, but the test in the sample code used a two-arg call.

  • Updated ZODB+Traversal tutorial of all steps to match what's generated by pyramid_zodb.

  • Fix reference to repoze.bfg.traversalwrapper in "Models" chapter (point at pyramid_traversalwrapper instead).

1.0a5 (2010-12-14)


  • Add a handler ZCML directive. This directive does the same thing as pyramid.configuration.add_handler.

  • A new module named pyramid.config was added. It subsumes the duties of the older pyramid.configuration module.

  • The new pyramid.config.Configurator` class has API methods that the older ``pyramid.configuration.Configurator class did not: with_context (a classmethod), include, action, and commit. These methods exist for imperative application extensibility purposes.

  • The pyramid.testing.setUp function now accepts an autocommit keyword argument, which defaults to True. If it is passed False, the Config object returned by setUp will be a non-autocommitting Config object.

  • Add logging configuration to all paster templates.

  • pyramid_alchemy, pyramid_routesalchemy, and pylons_sqla paster templates now use idiomatic SQLAlchemy configuration in their respective .ini files and Python code.

  • pyramid.testing.DummyRequest now has a class variable, query_string, which defaults to the empty string.

  • Add support for json on GAE by catching NotImplementedError and importing simplejson from django.utils.

  • The Mako renderer now accepts a resource specification for mako.module_directory.

  • New boolean Mako settings variable mako.strict_undefined. See Mako Context Variables for its meaning.


  • Depend on Mako 0.3.6+ (we now require the strict_undefined feature).

Bug Fixes

  • When creating a Configurator from within a paster pshell session, you were required to pass a package argument although package is not actually required. If you didn't pass package, you would receive an error something like KeyError: '__name__' emanating from the pyramid.path.caller_module function. This has now been fixed.

  • The pyramid_routesalchemy paster template's unit tests failed (AssertionError: 'SomeProject' != 'someproject'). This is fixed.

  • Make default renderer work (renderer factory registered with no name, which is active for every view unless the view names a specific renderer).

  • The Mako renderer did not properly turn the mako.imports, mako.default_filters, and mako.imports settings into lists.

  • The Mako renderer did not properly convert the mako.error_handler setting from a dotted name to a callable.


  • Merged many wording, readability, and correctness changes to narrative documentation chapters from (up to and including "Models" narrative chapter).

  • "Sample Applications" section of docs changed to note existence of Cluegun, Shootout and Virginia sample applications, ported from their repoze.bfg origin packages.

  • SQLAlchemy+URLDispatch tutorial updated to integrate changes to pyramid_routesalchemy template.

  • Add pyramid.interfaces.ITemplateRenderer interface to Interfaces API chapter (has implementation() method, required to be used when getting at Chameleon macros).

  • Add a "Modifying Package Structure" section to the project narrative documentation chapter (explain turning a module into a package).

  • Documentation was added for the new handler ZCML directive in the ZCML section.


  • pyramid.configuration.Configurator is now deprecated. Use pyramid.config.Configurator, passing its constructor autocommit=True instead. The pyramid.configuration.Configurator alias will live for a long time, as every application uses it, but its import now issues a deprecation warning. The pyramid.config.Configurator class has the same API as pyramid.configuration.Configurator class, which it means to replace, except by default it is a non-autocommitting configurator. The now-deprecated pyramid.configuration.Configurator will autocommit every time a configuration method is called.

    The pyramid.configuration module remains, but it is deprecated. Use pyramid.config instead.

1.0a4 (2010-11-21)


  • URL Dispatch now allows for replacement markers to be located anywhere in the pattern, instead of immediately following a /.

  • URL Dispatch now uses the form {marker} to denote a replace marker in the route pattern instead of :marker. The old colon-style marker syntax is still accepted for backwards compatibility. The new format allows a regular expression for that marker location to be used instead of the default [^/]+, for example {marker:\d+} is now valid to require the marker to be digits.

  • Add a pyramid.url.route_path API, allowing folks to generate relative URLs. Calling route_path is the same as calling pyramid.url.route_url with the argument _app_url equal to the empty string.

  • Add a pyramid.request.Request.route_path API. This is a convenience method of the request which calls pyramid.url.route_url.

  • Make test suite pass on Jython (requires PasteScript trunk, presumably to be 1.7.4).

  • Make test suite pass on PyPy (Chameleon doesn't work).

  • Surrounding application configuration with config.begin() and config.end() is no longer necessary. All paster templates have been changed to no longer call these functions.

  • Fix configurator to not convert ImportError to ConfigurationError if the import that failed was unrelated to the import requested via a dotted name when resolving dotted names (such as view dotted names).


  • SQLAlchemy+URLDispatch and ZODB+Traversal tutorials have been updated to not call config.begin() or config.end().

Bug Fixes

  • Add deprecation warnings to import of pyramid.chameleon_text and pyramid.chameleon_zpt of get_renderer, get_template, render_template, and render_template_to_response.

  • Add deprecation warning for import of pyramid.zcml.zcml_configure and pyramid.zcml.file_configure.

  • The pyramid_alchemy paster template had a typo, preventing an import from working.

  • Fix apparent failures when calling pyramid.traversal.find_model(root, path) or pyramid.traversal.traverse(path) when path is (erroneously) a Unicode object. The user is meant to pass these APIs a string object, never a Unicode object. In practice, however, users indeed pass Unicode. Because the string that is passed must be ASCII encodeable, now, if they pass a Unicode object, its data is eagerly converted to an ASCII string rather than being passed along to downstream code as a convenience to the user and to prevent puzzling second-order failures from cropping up (all failures will occur within pyramid.traversal.traverse rather than later down the line as the result of calling e.g. traversal_path).

Backwards Incompatibilities

  • The pyramid.testing.zcml_configure API has been removed. It had been advertised as removed since repoze.bfg 1.2a1, but hadn't actually been.


  • The pyramid.settings.get_settings API is now deprecated. Use pyramid.threadlocals.get_current_registry().settings instead or use the settings attribute of the registry available from the request (request.registry.settings).


  • Removed zodbsessions tutorial chapter. It's still useful, but we now have a SessionFactory abstraction which competes with it, and maintaining documentation on both ways to do it is a distraction.


  • Replace Twill with WebTest in internal integration tests (avoid deprecation warnings generated by Twill).

1.0a3 (2010-11-16)


  • Added Mako TemplateLookup settings for mako.error_handler, mako.default_filters, and mako.imports.

  • Normalized all paster templates: each now uses the name main to represent the function that returns a WSGI application, each now uses WebError, each now has roughly the same shape of development.ini style.

  • Added class vars matchdict and matched_route to pyramid.request.Request. Each is set to None.

  • New API method: pyramid.settings.asbool.

  • New API methods for pyramid.request.Request: model_url, route_url, and static_url. These are simple passthroughs for their respective functions in pyramid.url.

  • The settings object which used to be available only when request.settings.get_settings was called is now available as registry.settings (e.g. request.registry.settings in view code).

Bug Fixes

  • The pylons_* paster templates erroneously used the {squiggly} routing syntax as the pattern supplied to add_route. This style of routing is not supported. They were replaced with :colon style route patterns.

  • The pylons_* paster template used the same string (your_app_secret_string) for the session.secret setting in the generated development.ini. This was a security risk if left unchanged in a project that used one of the templates to produce production applications. It now uses a randomly generated string.


  • ZODB+traversal wiki (wiki) tutorial updated due to changes to pyramid_zodb paster template.

  • SQLAlchemy+urldispach wiki (wiki2) tutorial updated due to changes to pyramid_routesalchemy paster template.

  • Documented the matchdict and matched_route attributes of the request object in the Request API documentation.


  • Obtaining the settings object via registry.{get|query}Utility(ISettings) is now deprecated. Instead, obtain the settings object via the registry.settings attribute. A backwards compatibility shim was added to the registry object to register the settings object as an ISettings utility when setattr(registry, 'settings', foo) is called, but it will be removed in a later release.

  • Obtaining the settings object via pyramid.settings.get_settings is now deprecated. Obtain it as the settings attribute of the registry now (obtain the registry via pyramid.threadlocal.get_registry or as request.registry).

Behavior Differences

  • Internal: ZCML directives no longer call get_current_registry() if there's a registry attribute on the ZCML context (kill off use of threadlocals).

  • Internal: Chameleon template renderers now accept two arguments: path and lookup. Lookup will be an instance of a lookup class which supplies (late-bound) arguments for debug, reload, and translate. Any third-party renderers which use (the non-API) function pyramid.renderers.template_renderer_factory will need to adjust their implementations to obey the new callback argument list. This change was to kill off inappropriate use of threadlocals.

1.0a2 (2010-11-09)


  • All references to events by interface (e.g. pyramid.interfaces.INewRequest) have been changed to reference their concrete classes (e.g. in documentation about making subscriptions.

  • All references to Pyramid-the-application were changed from mod-pyramid to app-Pyramid. A custom role setting was added to docs/ to allow for this. (internal)

1.0a1 (2010-11-05)

Features (delta from BFG 1.3)

  • Mako templating renderer supports resource specification format for template lookups and within Mako templates. Absolute filenames must be used in Pyramid to avoid this lookup process.

  • Add pyramid.httpexceptions module, which is a facade for the webob.exc module.

  • Direct built-in support for the Mako templating language.

  • A new configurator method exists: add_handler. This method adds a Pylons-style "view handler" (such a thing used to be called a "controller" in Pylons 1.0).

  • New argument to configurator: session_factory.

  • New method on configurator: set_session_factory

  • Using request.session now returns a (dictionary-like) session object if a session factory has been configured.

  • The request now has a new attribute: tmpl_context for benefit of Pylons users.

  • The decorator previously known as pyramid.view.bfg_view is now known most formally as pyramid.view.view_config in docs and paster templates. An import of pyramid.view.bfg_view, however, will continue to work "forever".

  • New API methods in pyramid.session: signed_serialize and signed_deserialize.

  • New interface: pyramid.interfaces.IRendererInfo. An object of this type is passed to renderer factory constructors (see "Backwards Incompatibilities").

  • New event type: pyramid.interfaces.IBeforeRender. An object of this type is sent as an event before a renderer is invoked (but after the application-level renderer globals factory added via pyramid.configurator.configuration.set_renderer_globals_factory, if any, has injected its own keys). Applications may now subscribe to the IBeforeRender event type in order to introspect the and modify the set of renderer globals before they are passed to a renderer. The event object iself has a dictionary-like interface that can be used for this purpose. For example:

    from import subscriber
    from pyramid.interfaces import IRendererGlobalsEvent
    def add_global(event):
        event['mykey'] = 'foo'

    If a subscriber attempts to add a key that already exist in the renderer globals dictionary, a KeyError is raised. This limitation is due to the fact that subscribers cannot be ordered relative to each other. The set of keys added to the renderer globals dictionary by all subscribers and app-level globals factories must be unique.

  • New class: pyramid.response.Response. This is a pure facade for webob.Response (old code need not change to use this facade, it's existence is mostly for vanity and documentation-generation purposes).

  • All preexisting paster templates (except zodb) now use "imperative" configuration (starter, routesalchemy, alchemy).

  • A new paster template named pyramid_starter_zcml exists, which uses declarative configuration.

Documentation (delta from BFG 1.3)

  • Added a pyramid.httpexceptions API documentation chapter.

  • Added a pyramid.session API documentation chapter.

  • Added a Session Objects narrative documentation chapter.

  • Added an API chapter for the pyramid.personality module.

  • Added an API chapter for the pyramid.response module.

  • All documentation which previously referred to webob.Response now uses pyramid.response.Response instead.

  • The documentation has been overhauled to use imperative configuration, moving declarative configuration (ZCML) explanations to a separate narrative chapter declarative.rst.

  • The ZODB Wiki tutorial was updated to take into account changes to the pyramid_zodb paster template.

  • The SQL Wiki tutorial was updated to take into account changes to the pyramid_routesalchemy paster template.

Backwards Incompatibilities (with BFG 1.3)

  • There is no longer an IDebugLogger registered as a named utility with the name repoze.bfg.debug.

  • The logger which used to have the name of repoze.bfg.debug now has the name pyramid.debug.

  • The deprecated API pyramid.testing.registerViewPermission has been removed.

  • The deprecated API named pyramid.testing.registerRoutesMapper has been removed.

  • The deprecated API named pyramid.request.get_request was removed.

  • The deprecated API named was removed.

  • The deprecated API named pyramid.view.view_execution_permitted was removed.

  • The deprecated API named pyramid.view.NotFound was removed.

  • The bfgshell paster command is now named pshell.

  • The Venusian "category" for all built-in Venusian decorators (e.g. subscriber and view_config/bfg_view) is now pyramid instead of bfg.

  • pyramid.renderers.rendered_response function removed; use render_pyramid.renderers.render_to_response instead.

  • Renderer factories now accept a renderer info object rather than an absolute resource specification or an absolute path. The object has the following attributes: name (the renderer= value), package (the 'current package' when the renderer configuration statement was found), type: the renderer type, registry: the current registry, and settings: the deployment settings dictionary.

    Third-party repoze.bfg renderer implementations that must be ported to Pyramid will need to account for this.

    This change was made primarily to support more flexible Mako template rendering.

  • The presence of the key repoze.bfg.message in the WSGI environment when an exception occurs is now deprecated. Instead, code which relies on this environ value should use the exception attribute of the request (e.g. request.exception[0]) to retrieve the message.

  • The values bfg_localizer and bfg_locale_name kept on the request during internationalization for caching purposes were never APIs. These however have changed to localizer and locale_name, respectively.

  • The default cookie_name value of the authtktauthenticationpolicy ZCML now defaults to auth_tkt (it used to default to repoze.bfg.auth_tkt).

  • The default cookie_name value of the pyramid.authentication.AuthTktAuthenticationPolicy constructor now defaults to auth_tkt (it used to default to repoze.bfg.auth_tkt).

  • The request_type argument to the view ZCML directive, the pyramid.configuration.Configurator.add_view method, or the pyramid.view.view_config decorator (nee bfg_view) is no longer permitted to be one of the strings GET, HEAD, PUT, POST or DELETE, and now must always be an interface. Accepting the method-strings as request_type was a backwards compatibility strategy servicing repoze.bfg 1.0 applications. Use the request_method parameter instead to specify that a view a string request-method predicate.

repoze.bfg Change History (previous name for Pyramid)

1.3b1 (2010-10-25)


  • The paster template named bfg_routesalchemy has been updated to use SQLAlchemy declarative syntax. Thanks to Ergo^.

Bug Fixes

  • When a renderer factory could not be found, a misleading error message was raised if the renderer name was not a string.


  • The ""bfgwiki2" (SQLAlchemy + url dispatch) tutorial has been updated slightly. In particular, the source packages no longer attempt to use a private index, and the recommended Python version is now 2.6. It was also updated to take into account the changes to the bfg_routesalchemy template used to set up an environment.

  • The "bfgwiki" (ZODB + traversal) tutorial has been updated slightly. In particular, the source packages no longer attempt to use a private index, and the recommended Python version is now 2.6.

1.3a15 (2010-09-30)


  • The repoze.bfg.traversal.traversal_path API now eagerly attempts to encode a Unicode path into ASCII before attempting to split it and decode its segments. This is for convenience, effectively to allow a (stored-as-Unicode-in-a-database, or retrieved-as-Unicode-from-a-request-parameter) Unicode path to be passed to find_model, which eventually internally uses the traversal_path function under the hood. In version 1.2 and prior, if the path was Unicode, that Unicode was split on slashes and each resulting segment value was Unicode. An inappropriate call to the decode() method of a resulting Unicode path segment could cause a UnicodeDecodeError to occur even if the Unicode representation of the path contained no 'high order' characters (it effectively did a "double decode"). By converting the Unicode path argument to ASCII before we attempt to decode and split, genuine errors will occur in a more obvious place while also allowing us to handle (for convenience) the case that it's a Unicode representation formed entirely from ASCII-compatible characters.

1.3a14 (2010-09-14)

Bug Fixes

  • If an exception view was registered through the legacy set_notfound_view or set_forbidden_view APIs, the context sent to the view was incorrect (could be None inappropriately).


  • Compatibility with WebOb 1.0.


  • Now requires WebOb >= 1.0.

Backwards Incompatibilities

  • Due to changes introduced WebOb 1.0, the repoze.bfg.request.make_request_ascii event subscriber no longer works, so it has been removed. This subscriber was meant to be used in a deployment so that code written before BFG 0.7.0 could run unchanged. At this point, such code will need to be rewritten to expect Unicode from request.GET, request.POST and request.params or it will need to be changed to use request.str_POST, request.str_GET and/or request.str_params instead of the non-str versions of same, as the non-str versions of the same APIs always now perform decoding to Unicode.


  • A prior changelog entry asserted that the INewResponse event was not sent to listeners if the response was not "valid" (if a view or renderer returned a response object that did not have a status/headers/app_iter). This is not true in this release, nor was it true in 1.3a13.

1.3a13 (2010-09-14)

Bug Fixes

  • The traverse route predicate could not successfully generate a traversal path.


  • In support of making it easier to configure applications which are "secure by default", a default permission feature was added. If supplied, the default permission is used as the permission string to all view registrations which don't otherwise name a permission. These APIs are in support of that:

    • A new constructor argument was added to the Configurator: default_permission.

    • A new method was added to the Configurator: set_default_permission.

    • A new ZCML directive was added: default_permission.

  • Add a new request API: request.add_finished_callback. Finished callbacks are called by the router unconditionally near the very end of request processing. See the "Using Finished Callbacks" section of the "Hooks" narrative chapter of the documentation for more information.

  • A request.matched_route attribute is now added to the request when a route has matched. Its value is the "route" object that matched (see the IRoute interface within repoze.bfg.interfaces API documentation for the API of a route object).

  • The exception attribute of the request is now set slightly earlier and in a slightly different set of scenarios, for benefit of "finished callbacks" and "response callbacks". In previous versions, the exception attribute of the request was not set at all if an exception view was not found. In this version, the request.exception attribute is set immediately when an exception is caught by the router, even if an exception view could not be found.

  • The add_route method of a Configurator now accepts a pregenerator argument. The pregenerator for the resulting route is called by route_url in order to adjust the set of arguments passed to it by the user for special purposes, such as Pylons 'subdomain' support. It will influence the URL returned by route_url. See the repoze.bfg.interfaces.IRoutePregenerator interface for more information.

Backwards Incompatibilities

  • The router no longer sets the value wsgiorg.routing_args into the environ when a route matches. The value used to be something like ((), matchdict). This functionality was only ever obliquely referred to in change logs; it was never documented as an API.

  • The exception attribute of the request now defaults to None. In prior versions, the request.exception attribute did not exist if an exception was not raised by user code during request processing; it only began existence once an exception view was found.


  • The repoze.bfg.interfaces.IWSGIApplicationCreatedEvent event interface was renamed to repoze.bfg.interfaces.IApplicationCreated. Likewise, the class was renamed to The older aliases will continue to work indefinitely.

  • The repoze.bfg.interfaces.IAfterTraversal event interface was renamed to repoze.bfg.interfaces.IContextFound. Likewise, the class was renamed to The older aliases will continue to work indefinitely.

  • References to the WSGI environment values bfg.routes.matchdict and bfg.routes.route were removed from documentation. These will stick around internally for several more releases, but it is request.matchdict and request.matched_route are now the "official" way to obtain the matchdict and the route object which resulted in the match.


  • Added documentation for the default_permission ZCML directive.

  • Added documentation for the default_permission constructor value and the set_default_permission method in the Configurator API documentation.

  • Added a new section to the "security" chapter named "Setting a Default Permission".

  • Document renderer_globals_factory and request_factory arguments to Configurator constructor.

  • Added two sections to the "Hooks" chapter of the documentation: "Using Response Callbacks" and "Using Finished Callbacks".

  • Added documentation of the request.exception attribute to the repoze.bfg.request.Request API documentation.

  • Added glossary entries for "response callback" and "finished callback".

  • The "Request Processing" narrative chapter has been updated to note finished and response callback steps.

  • New interface in interfaces API documentation: IRoutePregenerator.

  • Added a "The Matched Route" section to the URL Dispatch narrative docs chapter, detailing the matched_route attribute.

1.3a12 (2010-09-08)

Bug Fixes

  • Fix a bug in repoze.bfg.url.static_url URL generation: if two resource specifications were used to create two separate static views, but they shared a common prefix, it was possible that static_url would generate an incorrect URL.

  • Fix another bug in repoze.bfg.static_url URL generation: too many slashes in generated URL.

  • Prevent a race condition which could result in a RuntimeError when rendering a Chameleon template that has not already been rendered once. This would usually occur directly after a restart, when more than one person or thread is trying to execute the same view at the same time:


  • The argument to repoze.bfg.configuration.Configurator.add_route which was previously called path is now called pattern for better explicability. For backwards compatibility purposes, passing a keyword argument named path to add_route will still work indefinitely.

  • The path attribute to the ZCML route directive is now named pattern for better explicability. The older path attribute will continue to work indefinitely.


  • All narrative, API, and tutorial docs which referred to a route pattern as a path have now been updated to refer to them as a pattern.

  • The repoze.bfg.interfaces API documentation page is now rendered via repoze.sphinx.autointerface.

  • The URL Dispatch narrative chapter now refers to the interfaces chapter to explain the API of an IRoute object.

Paster Templates

  • The routesalchemy template has been updated to use pattern in its route declarations rather than path.


  • tests_require now includes repoze.sphinx.autointerface as a dependency.


  • Add an API to the Configurator named get_routes_mapper. This returns an object implementing the IRoutesMapper interface.

  • The repoze.bfg.urldispatch.RoutesMapper object now has a get_route method which returns a single Route object or None.

  • A new interface repoze.bfg.interfaces.IRoute was added. The repoze.bfg.urldispatch.Route object implements this interface.

  • The canonical attribute for accessing the routing pattern from a route object is now pattern rather than path.

  • Use hash() rather than id() when computing the "phash" of a custom route/view predicate in order to allow the custom predicate some control over which predicates are "equal".

  • Use response.headerlist.append instead of response.headers.add in repoze.bfg.request.add_global_response_headers in case the response is not a WebOb response.

  • The repoze.bfg.urldispatch.Route constructor (not an API) now accepts a different ordering of arguments. Previously it was (pattern, name, factory=None, predicates=()). It is now (name, pattern, factory=None, predicates=()). This is in support of consistency with configurator.add_route.

  • The repoze.bfg.urldispatch.RoutesMapper.connect method (not an API) now accepts a different ordering of arguments. Previously it was (pattern, name, factory=None, predicates=()). It is now (name, pattern, factory=None, predicates=()). This is in support of consistency with configurator.add_route.

1.3a11 (2010-09-05)

Bug Fixes

  • Process the response callbacks and the NewResponse event earlier, to enable mutations to the response to take effect.

1.3a10 (2010-09-05)


  • A new repoze.bfg.request.Request.add_response_callback API has been added. This method is documented in the new repoze.bfg.request API chapter. It can be used to influence response values before a concrete response object has been created.

  • The repoze.bfg.interfaces.INewResponse interface now includes a request attribute; as a result, a handler for INewResponse now has access to the request which caused the response.

  • Each of the follow methods of the Configurator now allow the below-named arguments to be passed as "dotted name strings" (e.g. "") rather than as actual implementation objects that must be imported:


    root_factory, authentication_policy, authorization_policy, debug_logger, locale_negotiator, request_factory, renderer_globals_factory


    subscriber, iface




    view, for_, context, request_type, containment


    view, view_for, factory, for_, view_context

















Bug Fixes

  • The route pattern registered internally for a local "static view" (either via the static ZCML directive or via the add_static_view method of the configurator) was incorrect. It was registered for e.g. static*traverse, while it should have been registered for static/*traverse. Symptom: two static views could not reliably be added to a system when they both shared the same path prefix (e.g. /static and /static2).

Backwards Incompatibilities

  • The INewResponse event is now not sent to listeners if the response returned by view code (or a renderer) is not a "real" response (e.g. if it does not have .status, .headerlist and .app_iter attribtues).


  • Add an API chapter for the repoze.bfg.request module, which includes documentation for the repoze.bfg.request.Request class (the "request object").

  • Modify the "Request and Response" narrative chapter to reference the new repoze.bfg.request API chapter. Some content was moved from this chapter into the API documentation itself.

  • Various changes to denote that Python dotted names are now allowed as input to Configurator methods.


  • The (internal) feature which made it possible to attach a global_response_headers attribute to the request (which was assumed to contain a sequence of header key/value pairs which would later be added to the response by the router), has been removed. The functionality of repoze.bfg.request.Request.add_response_callback takes its place.

  • The class's construct has changed: it now must be created with (request, response) rather than simply (response).

1.3a9 (2010-08-22)


  • The Configurator now accepts a dotted name string to a package as a package constructor argument. The package argument was previously required to be a package object (not a dotted name string).

  • The repoze.bfg.configuration.Configurator.with_package method was added. This method returns a new Configurator using the same application registry as the configurator object it is called upon. The new configurator is created afresh with its package constructor argument set to the value passed to with_package. This feature will make it easier for future BFG versions to allow dotted names as arguments in places where currently only object references are allowed (the work to allow dotted names instead of object references everywhere has not yet been done, however).

  • The new repoze.bfg.configuration.Configurator.maybe_dotted method resolves a Python dotted name string supplied as its dotted argument to a global Python object. If the value cannot be resolved, a repoze.bfg.configuration.ConfigurationError is raised. If the value supplied as dotted is not a string, the value is returned unconditionally without any resolution attempted.

  • The new repoze.bfg.configuration.Configurator.absolute_resource_spec method resolves a potentially relative "resource specification" string into an absolute version. If the value supplied as relative_spec is not a string, the value is returned unconditionally without any resolution attempted.

Backwards Incompatibilities

  • The functions in repoze.bfg.renderers named render and render_to_response introduced in 1.3a6 previously took a set of **values arguments for the values to be passed to the renderer. This was wrong, as renderers don't need to accept only dictionaries (they can accept any type of object). Now, the value sent to the renderer must be supplied as a positional argument named value. The request argument is still a keyword argument, however.

  • The functions in repoze.bfg.renderers named render and render_to_response now accept an additional keyword argument named package.

  • The get_renderer API in repoze.bfg.renderers now accepts a package argument.


  • The ZCML include directive docs were incorrect: they specified filename rather than (the correct) file as an allowable attribute.


  • The repoze.bfg.resource.resolve_resource_spec function can now accept a package object as its pname argument instead of just a package name.

  • The _renderer_factory_from_name and _renderer_from_name methods of the Configurator were removed. These were never APIs.

  • The _render, _render_to_response and _make_response functions with repoze.bfg.render (added in 1.3a6) have been removed.

  • A new helper class repoze.bfg.renderers.RendererHelper was added.

  • The _map_view function of repoze.bfg.configuration now takes only a renderer_name argument instead of both a renderer and renderer``_name argument.  It also takes a ``package argument now.

  • Use imp.get_suffixes indirection in repoze.bfg.path.package_name instead of hardcoded .py .pyc and .pyo to use for comparison when attempting to decide if a directory is a package.

  • Make tests runnable again under Jython (although they do not all pass currently).

  • The reify decorator now maintains the docstring of the function it wraps.

1.3a8 (2010-08-08)


  • New public interface: repoze.bfg.exceptions.IExceptionResponse. This interface is provided by all internal exception classes (such as repoze.bfg.exceptions.NotFound and repoze.bfg.exceptions.Forbidden), instances of which are both exception objects and can behave as WSGI response objects. This interface is made public so that exception classes which are also valid WSGI response factories can be configured to implement them or exception instances which are also or response instances can be configured to provide them.

  • New API class: repoze.bfg.view.AppendSlashNotFoundViewFactory.

    There can only be one Not Found view in any repoze.bfg application. Even if you use repoze.bfg.view.append_slash_notfound_view as the Not Found view, repoze.bfg still must generate a 404 Not Found response when it cannot redirect to a slash-appended URL; this not found response will be visible to site users.

    If you don't care what this 404 response looks like, and you only need redirections to slash-appended route URLs, you may use the repoze.bfg.view.append_slash_notfound_view object as the Not Found view. However, if you wish to use a custom notfound view callable when a URL cannot be redirected to a slash-appended URL, you may wish to use an instance of the repoze.bfg.view.AppendSlashNotFoundViewFactory class as the Not Found view, supplying the notfound view callable as the first argument to its constructor. For instance:

    from repoze.bfg.exceptions import NotFound
    from repoze.bfg.view import AppendSlashNotFoundViewFactory
    def notfound_view(context, request):
        return HTTPNotFound('It aint there, stop trying!')
    custom_append_slash = AppendSlashNotFoundViewFactory(notfound_view)
    config.add_view(custom_append_slash, context=NotFound)

    The notfound_view supplied must adhere to the two-argument view callable calling convention of (context, request) (context will be the exception object).


  • Expanded the "Cleaning Up After a Request" section of the URL Dispatch narrative chapter.

  • Expanded the "Redirecting to Slash-Appended Routes" section of the URL Dispatch narrative chapter.


  • Previously, two default view functions were registered at Configurator setup (one for repoze.bfg.exceptions.NotFound named default_notfound_view and one for repoze.bfg.exceptions.Forbidden named default_forbidden_view) to render internal exception responses. Those default view functions have been removed, replaced with a generic default view function which is registered at Configurator setup for the repoze.bfg.interfaces.IExceptionResponse interface that simply returns the exception instance; the NotFound and Forbidden classes are now still exception factories but they are also response factories which generate instances that implement the new repoze.bfg.interfaces.IExceptionResponse interface.

1.3a7 (2010-08-01)


  • The repoze.bfg.configuration.Configurator.add_route API now returns the route object that was added.

  • A decorator was added. This decorator decorates module-scope functions, which are then treated as event listeners after a scan() is performed. See the Events narrative documentation chapter and the module documentation for more information.

Bug Fixes

  • When adding a view for a route which did not yet exist ("did not yet exist" meaning, temporally, a view was added with a route name for a route which had not yet been added via add_route), the value of the custom_predicate argument to add_view was lost. Symptom: wrong view matches when using URL dispatch and custom view predicates together.

  • Pattern matches for a :segment marker in a URL dispatch route pattern now always match at least one character. See "Backwards Incompatibilities" below in this changelog.

Backwards Incompatibilities

  • A bug existed in the regular expression to do URL matching. As an example, the URL matching machinery would cause the pattern /{foo} to match the root URL / resulting in a match dictionary of {'foo':u''} or the pattern /{fud}/edit might match the URL ``//edit resulting in a match dictionary of {'fud':u''}. It was always the intent that :segment markers in the pattern would need to match at least one character, and never match the empty string. This, however, means that in certain circumstances, a routing match which your application inadvertently depended upon may no longer happen.


  • Added description of the decorator to the Events narrative chapter.

  • Added API documentation to API docs.

  • Added a section named "Zope 3 Enforces 'TTW' Authorization Checks By Default; BFG Does Not" to the "Design Defense" chapter.

1.3a6 (2010-07-25)


  • New argument to repoze.bfg.configuration.Configurator.add_route and the route ZCML directive: traverse. If you would like to cause the context to be something other than the root object when this route matches, you can spell a traversal pattern as the traverse argument. This traversal pattern will be used as the traversal path: traversal will begin at the root object implied by this route (either the global root, or the object returned by the factory associated with this route).

    The syntax of the traverse argument is the same as it is for path. For example, if the path provided is articles/:article/edit, and the traverse argument provided is /:article, when a request comes in that causes the route to match in such a way that the article match value is '1' (when the request URI is /articles/1/edit), the traversal path will be generated as /1. This means that the root object's __getitem__ will be called with the name 1 during the traversal phase. If the 1 object exists, it will become the context of the request. The Traversal narrative has more information about traversal.

    If the traversal path contains segment marker names which are not present in the path argument, a runtime error will occur. The traverse pattern should not contain segment markers that do not exist in the path.

    A similar combining of routing and traversal is available when a route is matched which contains a *traverse remainder marker in its path. The traverse argument allows you to associate route patterns with an arbitrary traversal path without using a *traverse remainder marker; instead you can use other match information.

    Note that the traverse argument is ignored when attached to a route that has a *traverse remainder marker in its path.

  • A new method of the Configurator exists: set_request_factory. If used, this method will set the factory used by the repoze.bfg router to create all request objects.

  • The Configurator constructor takes an additional argument: request_factory. If used, this argument will set the factory used by the repoze.bfg router to create all request objects.

  • The Configurator constructor takes an additional argument: request_factory. If used, this argument will set the factory used by the repoze.bfg router to create all request objects.

  • A new method of the Configurator exists: set_renderer_globals_factory. If used, this method will set the factory used by the repoze.bfg router to create renderer globals.

  • A new method of the Configurator exists: get_settings. If used, this method will return the current settings object (performs the same job as the repoze.bfg.settings.get_settings API).

  • The Configurator constructor takes an additional argument: renderer_globals_factory. If used, this argument will set the factory used by the repoze.bfg router to create renderer globals.

  • Add repoze.bfg.renderers.render, repoze.bfg.renderers.render_to_response and repoze.bfg.renderers.get_renderer functions. These are imperative APIs which will use the same rendering machinery used by view configurations with a renderer= attribute/argument to produce a rendering or renderer. Because these APIs provide a central API for all rendering, they now form the preferred way to perform imperative template rendering. Using functions named render_* from modules such as repoze.bfg.chameleon_zpt and repoze.bfg.chameleon_text is now discouraged (although not deprecated). The code the backing older templating-system-specific APIs now calls into the newer repoze.bfg.renderer code.

  • The repoze.bfg.configuration.Configurator.testing_add_template has been renamed to testing_add_renderer. A backwards compatibility alias is present using the old name.


  • The Hybrid narrative chapter now contains a description of the traverse route argument.

  • The Hooks narrative chapter now contains sections about changing the request factory and adding a renderer globals factory.

  • The API documentation includes a new module: repoze.bfg.renderers.

  • The Templates chapter was updated; all narrative that used templating-specific APIs within examples to perform rendering (such as the repoze.bfg.chameleon_zpt.render_template_to_response method) was changed to use repoze.bfg.renderers.render_* functions.

Bug Fixes

  • The header predicate (when used as either a view predicate or a route predicate) had a problem when specified with a name/regex pair. When the header did not exist in the headers dictionary, the regex match could be fed None, causing it to throw a TypeError: expected string or buffer exception. Now, the predicate returns False as intended.


  • The repoze.bfg.renderers.rendered_response function was never an official API, but may have been imported by extensions in the wild. It is officially deprecated in this release. Use repoze.bfg.renderers.render_to_response instead.

  • The following APIs are documentation deprecated (meaning they are officially deprecated in documentation but do not raise a deprecation error upon their usage, and may continue to work for an indefinite period of time):

    In the repoze.bfg.chameleon_zpt module: get_renderer, get_template, render_template, render_template_to_response. The suggested alternatives are documented within the docstrings of those methods (which are still present in the documentation).

    In the repoze.bfg.chameleon_text module: get_renderer, get_template, render_template, render_template_to_response. The suggested alternatives are documented within the docstrings of those methods (which are still present in the documentation).

    In general, to perform template-related functions, one should now use the various methods in the repoze.bfg.renderers module.

Backwards Incompatibilities

  • A new internal exception class (not an API) named repoze.bfg.exceptions.PredicateMismatch now exists. This exception is currently raised when no constituent view of a multiview can be called (due to no predicate match). Previously, in this situation, a repoze.bfg.exceptions.NotFound was raised. We provide backwards compatibility for code that expected a NotFound to be raised when no predicates match by causing repoze.bfg.exceptions.PredicateMismatch to inherit from NotFound. This will cause any exception view registered for NotFound to be called when a predicate mismatch occurs, as was the previous behavior.

    There is however, one perverse case that will expose a backwards incompatibility. If 1) you had a view that was registered as a member of a multiview 2) this view explicitly raised a NotFound exception in order to proceed to the next predicate check in the multiview, that code will now behave differently: rather than skipping to the next view match, a NotFound will be raised to the top-level exception handling machinery instead. For code to be depending upon the behavior of a view raising NotFound to proceed to the next predicate match, would be tragic, but not impossible, given that NotFound is a public interface. repoze.bfg.exceptions.PredicateMismatch is not a public API and cannot be depended upon by application code, so you should not change your view code to raise PredicateMismatch. Instead, move the logic which raised the NotFound exception in the view out into a custom view predicate.

  • If, when you run your application's unit test suite under BFG 1.3, a KeyError naming a template or a ValueError indicating that a 'renderer factory' is not registered may is raised (e.g. ValueError: No factory for renderer named '.pt' when looking up karl.views:templates/, you may need to perform some extra setup in your test code.

    The best solution is to use the repoze.bfg.configuration.Configurator.testing_add_renderer (or, alternately the deprecated repoze.bfg.testing.registerTemplateRenderer or registerDummyRenderer) API within the code comprising each individual unit test suite to register a "dummy" renderer for each of the templates and renderers used by code under test. For example:

    config = Configurator()

    This will register a basic dummy renderer for this particular missing template. The testing_add_renderer API actually returns the renderer, but if you don't care about how the render is used, you don't care about having a reference to it either.

    A more rough way to solve the issue exists. It causes the "real" template implementations to be used while the system is under test, which is suboptimal, because tests will run slower, and unit tests won't actually be unit tests, but it is easier. Always ensure you call the setup_registry() method of the Configurator . Eg:

    reg = MyRegistry()
    config = Configurator(registry=reg)

    Calling setup_registry only has an effect if you're passing in a registry argument to the Configurator constructor. setup_registry is called by the course of normal operations anyway if you do not pass in a registry.

    If your test suite isn't using a Configurator yet, and is still using the older repoze.bfg.testing APIs name setUp or cleanUp, these will register the renderers on your behalf.

    A variant on the symptom for this theme exists: you may already be dutifully registering a dummy template or renderer for a template used by the code you're testing using testing_register_renderer or registerTemplateRenderer, but (perhaps unbeknownst to you) the code under test expects to be able to use a "real" template renderer implementation to retrieve or render another template that you forgot was being rendered as a side effect of calling the code you're testing. This happened to work because it found the real template while the system was under test previously, and now it cannot. The solution is the same.

    It may also help reduce confusion to use a resource specification to specify the template path in the test suite and code rather than a relative path in either. A resource specification is unambiguous, while a relative path needs to be relative to "here", where "here" isn't always well-defined ("here" in a test suite may or may not be the same as "here" in the code under test).

1.3a5 (2010-07-14)


  • New internal exception: repoze.bfg.exceptions.URLDecodeError. This URL is a subclass of the built-in Python exception named UnicodeDecodeError.

  • When decoding a URL segment to Unicode fails, the exception raised is now repoze.bfg.exceptions.URLDecodeError instead of UnicodeDecodeError. This makes it possible to register an exception view invoked specifically when repoze.bfg cannot decode a URL.

Bug Fixes

  • Fix regression in repoze.bfg.configuration.Configurator.add_static_view. Before 1.3a4, view names that contained a slash were supported as route prefixes. 1.3a4 broke this by trying to treat them as full URLs.


  • The repoze.bfg.exceptions.URLDecodeError exception was added to the exceptions chapter of the API documentation.

Backwards Incompatibilities

  • in previous releases, when a URL could not be decoded from UTF-8 during traversal, a TypeError was raised. Now the error which is raised is a repoze.bfg.exceptions.URLDecodeError.

1.3a4 (2010-07-03)


  • Undocumented hook: make get_app and get_root of the repoze.bfg.paster.BFGShellCommand hookable in cases where endware may interfere with the default versions.

  • In earlier versions, a custom route predicate associated with a url dispatch route (each of the predicate functions fed to the custom_predicates argument of repoze.bfg.configuration.Configurator.add_route) has always required a 2-positional argument signature, e.g. (context, request). Before this release, the context argument was always None.

    As of this release, the first argument passed to a predicate is now a dictionary conventionally named info consisting of route, and match. match is a dictionary: it represents the arguments matched in the URL by the route. route is an object representing the route which was matched.

    This is useful when predicates need access to the route match. For example:

    def any_of(segment_name, *args):
        def predicate(info, request):
            if info['match'][segment_name] in args:
                return True
        return predicate
    num_one_two_or_three = any_of('num, 'one', 'two', 'three')
    add_route('num', '/:num', custom_predicates=(num_one_two_or_three,))

    The route object is an object that has two useful attributes: name and path. The name attribute is the route name. The path attribute is the route pattern. An example of using the route in a set of route predicates:

    def twenty_ten(info, request):
        if info['route'].name in ('ymd', 'ym', 'y'):
            return info['match']['year'] == '2010'
    add_route('y', '/:year', custom_predicates=(twenty_ten,))
    add_route('ym', '/:year/:month', custom_predicates=(twenty_ten,))
    add_route('ymd', '/:year/:month:/day', custom_predicates=(twenty_ten,))
  • The repoze.bfg.url.route_url API has changed. If a keyword _app_url is present in the arguments passed to route_url, this value will be used as the protocol/hostname/port/leading path prefix of the generated URL. For example, using an _app_url of would cause the URL to be returned from this function if the expansion of the route pattern associated with the route_name expanded to /fleeb/flub.

  • It is now possible to use a URL as the name argument fed to repoze.bfg.configuration.Configurator.add_static_view. When the name argument is a URL, the repoze.bfg.url.static_url API will generate join this URL (as a prefix) to a path including the static file name. This makes it more possible to put static media on a separate webserver for production, while keeping static media package-internal and served by the development webserver during development.


  • The authorization chapter of the ZODB Wiki Tutorial (docs/tutorials/bfgwiki) was changed to demonstrate authorization via a group rather than via a direct username (thanks to Alex Marandon).

  • The authorization chapter of the SQLAlchemy Wiki Tutorial (docs/tutorials/bfgwiki2) was changed to demonstrate authorization via a group rather than via a direct username.

  • Redirect requests for tutorial sources to and respectively.

  • A section named Custom Route Predicates was added to the URL Dispatch narrative chapter.

  • The Static Resources chapter has been updated to mention using static_url to generate URLs to external webservers.


  • Removed repoze.bfg.static.StaticURLFactory in favor of a new abstraction revolving around the (still-internal) repoze.bfg.static.StaticURLInfo helper class.

1.3a3 (2010-05-01)

Paster Templates

  • The bfg_alchemy and bfg_routesalchemy templates no longer register a handle_teardown event listener which calls DBSession.remove. This was found by Chris Withers to be unnecessary.


  • The "bfgwiki2" (URL dispatch wiki) tutorial code and documentation was changed to remove the handle_teardown event listener which calls DBSession.remove.

  • Any mention of the handle_teardown event listener as used by the paster templates was removed from the URL Dispatch narrative chapter.

  • A section entitled Detecting Available Languages was added to the i18n narrative docs chapter.

1.3a2 (2010-04-28)


  • A locale negotiator no longer needs to be registered explicitly. The default locale negotiator at repoze.bfg.i18n.default_locale_negotiator is now used unconditionally as... um, the default locale negotiator.

  • The default locale negotiator has become more complex.

    • First, the negotiator looks for the _LOCALE_ attribute of the request object (possibly set by a view or an event listener).

    • Then it looks for the request.params['_LOCALE_'] value.

    • Then it looks for the request.cookies['_LOCALE_'] value.

Backwards Incompatibilities

  • The default locale negotiator now looks for the parameter named _LOCALE_ rather than a parameter named locale in request.params.

Behavior Changes

  • A locale negotiator may now return None, signifying that the default locale should be used.


  • Documentation concerning locale negotiation in the Internationalizationa and Localization chapter was updated.

  • Expanded portion of i18n narrative chapter docs which discuss working with gettext files.

1.3a1 (2010-04-26)


  • Added "exception views". When you use an exception (anything that inherits from the Python Exception builtin) as view context argument, e.g.:

    from repoze.bfg.view import bfg_view
    from repoze.bfg.exceptions import NotFound
    from webob.exc import HTTPNotFound
    def notfound_view(request):
        return HTTPNotFound()

    For the above example, when the repoze.bfg.exceptions.NotFound exception is raised by any view or any root factory, the notfound_view view callable will be invoked and its response returned.

    Other normal view predicates can also be used in combination with an exception view registration:

    from repoze.bfg.view import bfg_view
    from repoze.bfg.exceptions import NotFound
    from webob.exc import HTTPNotFound
    @bfg_view(context=NotFound, route_name='home')
    def notfound_view(request):
        return HTTPNotFound()

    The above exception view names the route_name of home, meaning that it will only be called when the route matched has a name of home. You can therefore have more than one exception view for any given exception in the system: the "most specific" one will be called when the set of request circumstances which match the view registration. The only predicate that cannot be not be used successfully is name. The name used to look up an exception view is always the empty string.

    Existing (pre-1.3) normal views registered against objects inheriting from Exception will continue to work. Exception views used for user-defined exceptions and system exceptions used as contexts will also work.

    The feature can be used with any view registration mechanism (@bfg_view decorator, ZCML, or imperative config.add_view styles).

    This feature was kindly contributed by Andrey Popp.

  • Use "Venusian" ( to perform bfg_view decorator scanning rather than relying on a BFG-internal decorator scanner. (Truth be told, Venusian is really just a generalization of the BFG-internal decorator scanner).

  • Internationalization and localization features as documented in the narrative documentation chapter entitled Internationalization and Localization.

  • A new deployment setting named default_locale_name was added. If this string is present as a Paster .ini file option, it will be considered the default locale name. The default locale name is used during locale-related operations such as language translation.

  • It is now possible to turn on Chameleon template "debugging mode" for all Chameleon BFG templates by setting a BFG-related Paster .ini file setting named debug_templates. The exceptions raised by Chameleon templates when a rendering fails are sometimes less than helpful. debug_templates allows you to configure your application development environment so that exceptions generated by Chameleon during template compilation and execution will contain more helpful debugging information. This mode is on by default in all new projects.

  • Add a new method of the Configurator named derive_view which can be used to generate a BFG view callable from a user-supplied function, instance, or class. This useful for external framework and plugin authors wishing to wrap callables supplied by their users which follow the same calling conventions and response conventions as objects that can be supplied directly to BFG as a view callable. See the derive_view method in the repoze.bfg.configuration.Configurator docs.


  • Add a translationdir ZCML directive to support localization.

  • Add a localenegotiator ZCML directive to support localization.


  • The exception views feature replaces the need for the set_notfound_view and set_forbidden_view methods of the Configurator as well as the notfound and forbidden ZCML directives. Those methods and directives will continue to work for the foreseeable future, but they are deprecated in the documentation.


  • A new install-time dependency on the venusian distribution was added.

  • A new install-time dependency on the translationstring distribution was added.

  • Chameleon 1.2.3 or better is now required (internationalization and per-template debug settings).


  • View registrations and lookups are now done with three "requires" arguments instead of two to accommodate orthogonality of exception views.

  • The repoze.bfg.interfaces.IForbiddenView and repoze.bfg.interfaces.INotFoundView interfaces were removed; they weren't APIs and they became vestigial with the addition of exception views.

  • Remove and import alias repoze.bfg.compat.walk_packages. These were only required by internal scanning machinery; Venusian replaced the internal scanning machinery, so these are no longer required.


  • Exception view documentation was added to the Hooks narrative chapter.

  • A new narrative chapter entitled Internationalization and Localization was added.

  • The "Environment Variables and ini File Settings" chapter was changed: documentation about the default_locale_name setting was added.

  • A new API chapter for the repoze.bfg.i18n module was added.

  • Documentation for the new translationdir and localenegotiator ZCML directives were added.

  • A section was added to the Templates chapter entitled "Nicer Exceptions in Templates" describing the result of setting debug_templates = true.

Paster Templates

  • All paster templates now create a setup.cfg which includes commands related to nose testing and Babel message catalog extraction/compilation.

  • A default_locale_name = en setting was added to each existing paster template.

  • A debug_templates = true setting was added to each existing paster template.


  • The Edgewall (BSD) license was added to the LICENSES.txt file, as some code in the repoze.bfg.i18n derives from Babel source.

1.2 (2010-02-10)

  • No changes from 1.2b6.

1.2b6 (2010-02-06)

Backwards Incompatibilities

  • Remove magical feature of repoze.bfg.url.model_url which prepended a fully-expanded urldispatch route URL before a the model's path if it was noticed that the request had matched a route. This feature was ill-conceived, and didn't work in all scenarios.

Bug Fixes

  • More correct conversion of provided renderer values to resource specification values (internal).

1.2b5 (2010-02-04)

Bug Fixes

  • 1.2b4 introduced a bug whereby views added via a route configuration that named a view callable and also a view_attr became broken. Symptom: MyViewClass is not callable or the __call__ of a class was being called instead of the method named via view_attr.

  • Fix a bug whereby a renderer argument to the @bfg_view decorator that provided a package-relative template filename might not have been resolved properly. Symptom: inappropriate Missing template resource errors.

1.2b4 (2010-02-03)


  • Update GAE tutorial to use Chameleon instead of Jinja2 (now that it's possible).

Bug Fixes

  • Ensure that secure flag for AuthTktAuthenticationPolicy constructor does what it's documented to do (merge Daniel Holth's fancy-cookies-2 branch).


  • Add path and http_only options to AuthTktAuthenticationPolicy constructor (merge Daniel Holth's fancy-cookies-2 branch).

Backwards Incompatibilities

  • Remove view_header, view_accept, view_xhr, view_path_info, view_request_method, view_request_param, and view_containment predicate arguments from the Configurator.add_route argument list. These arguments were speculative. If you need the features exposed by these arguments, add a view associated with a route using the route_name argument to the add_view method instead.

  • Remove view_header, view_accept, view_xhr, view_path_info, view_request_method, view_request_param, and view_containment predicate arguments from the route ZCML directive attribute set. These attributes were speculative. If you need the features exposed by these attributes, add a view associated with a route using the route_name attribute of the view ZCML directive instead.


  • Remove dependency on sourcecodegen (not depended upon by Chameleon 1.1.1+).

1.2b3 (2010-01-24)

Bug Fixes

  • When "hybrid mode" (both traversal and urldispatch) is in use, default to finding route-related views even if a non-route-related view registration has been made with a more specific context. The default used to be to find views with a more specific context first. Use the new use_global_views argument to the route definition to get back the older behavior.


  • Add use_global_views argument to add_route method of Configurator. When this argument is true, views registered for no route will be found if no more specific view related to the route is found.

  • Add use_global_views attribute to ZCML <route> directive (see above).


  • When registering a view, register the view adapter with the "requires" interfaces as (request_type, context_type) rather than (context_type, request_type). This provides for saner lookup, because the registration will always be made with a specific request interface, but registration may not be made with a specific context interface. In general, when creating multiadapters, you want to order the requires interfaces so that the elements which are more likely to be registered using specific interfaces are ordered before those which are less likely.

1.2b2 (2010-01-21)

Bug Fixes

  • When the Configurator is passed an instance of zope.component.registry.Components as a registry constructor argument, fix the instance up to have the attributes we expect of an instance of repoze.bfg.registry.Registry when setup_registry is called. This makes it possible to use the global Zope component registry as a BFG application registry.

  • When WebOb was used, a deprecation warning was issued for the class attribute named charset within repoze.bfg.request.Request. BFG now requires WebOb >= 0.9.7, and code was added so that this deprecation warning has disappeared.

  • Fix a view lookup ordering bug whereby a view with a larger number of predicates registered first (literally first, not "earlier") for a triad would lose during view lookup to one registered with fewer.

  • Make sure views with exactly N custom predicates are always called before views with exactly N non-custom predicates given all else is equal in the view configuration.


  • Change renderings of ZCML directive documentation.

  • Add a narrative documentation chapter: "Using the Zope Component Architecture in repoze.bfg".


  • Require WebOb >= 0.9.7

1.2b1 (2010-01-18)

Bug Fixes

  • In bfg_routesalchemy, bfg_alchemy paster templates and the bfgwiki2 tutorial, clean up the SQLAlchemy connection by registering a callback instead of relying on a __del__ method of a Cleanup class added to the WSGI environment. The __del__ strategy was fragile and caused problems in the wild. Thanks to Daniel Holth for testing.


  • Read logging configuration from PasteDeploy config file loggers section (and related) when paster bfgshell is invoked.


  • Major rework in preparation for book publication.

1.2a11 (2010-01-05)

Bug Fixes

  • Make paster bfgshell and paster create -t bfg_xxx work on Jython (fix minor incompatibility with treatment of __doc__ at the class level).

  • Updated dependency on WebOb to require a version which supports features now used in tests.


  • Jython compatibility (at least when repoze.bfg.jinja2 is used as the templating engine; Chameleon does not work under Jython).

  • Show the derived abspath of template resource specifications in the traceback when a renderer template cannot be found.

  • Show the original traceback when a Chameleon template cannot be rendered due to a platform incompatibility.

1.2a10 (2010-01-04)


  • The Configurator.add_view method now accepts an argument named context. This is an alias for the older argument named for_; it is preferred over for_, but for_ will continue to be supported "forever".

  • The view ZCML directive now accepts an attribute named context. This is an alias for the older attribute named for; it is preferred over for, but for will continue to be supported "forever".

  • The Configurator.add_route method now accepts an argument named view_context. This is an alias for the older argument named view_for; it is preferred over view_for, but view_for will continue to be supported "forever".

  • The route ZCML directive now accepts an attribute named view_context. This is an alias for the older attribute named view_for; it is preferred over view_for, but view_for will continue to be supported "forever".

Documentation and Paster Templates

  • LaTeX rendering tweaks.

  • All uses of the Configurator.add_view method that used its for_ argument now use the context argument instead.

  • All uses of the Configurator.add_route method that used its view_for argument now use the view_context argument instead.

  • All uses of the view ZCML directive that used its for attribute now use the context attribute instead.

  • All uses of the route ZCML directive that used its view_for attribute now use the view_context attribute instead.

  • Add a (minimal) tutorial dealing with use of repoze.catalog in a repoze.bfg application.

Documentation Licensing

1.2a9 (2009-12-27)

Documentation Licensing

  • The documentation (the result of make <html|latex|htmlhelp> within the docs directory) in this release is now offered under the Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License as described by . This is only a licensing change for the documentation; the repoze.bfg software continues to be offered under the Repoze Public License at (BSD-like).


  • Added manual index entries to generated index.

  • Document the previously existing (but non-API) repoze.bfg.configuration.Configurator.setup_registry method as an official API of a Configurator.

  • Fix syntax errors in various documentation code blocks.

  • Created new top-level documentation section: "ZCML Directives". This section contains detailed ZCML directive information, some of which was removed from various narrative chapters.

  • The LaTeX rendering of the documentation has been improved.

  • Added a "Fore-Matter" section with author, copyright, and licensing information.

1.2a8 (2009-12-24)


  • Add a **kw arg to the Configurator.add_settings API.

  • Add hook_zca and unhook_zca methods to the Configurator API.

  • The repoze.bfg.testing.setUp method now returns a Configurator instance which can be used to do further configuration during unit tests.

Bug Fixes

  • The json renderer failed to set the response content type to application/json. It now does, by setting request.response_content_type unless this attribute is already set.

  • The string renderer failed to set the response content type to text/plain. It now does, by setting request.response_content_type unless this attribute is already set.


  • General documentation improvements by using better Sphinx roles such as "class", "func", "meth", and so on. This means that there are many more hyperlinks pointing to API documentation for API definitions in all narrative, tutorial, and API documentation elements.

  • Added a description of imperative configuration in various places which only described ZCML configuration.

  • A syntactical refreshing of various tutorials.

  • Added the repoze.bfg.authentication, repoze.bfg.authorization, and repoze.bfg.interfaces modules to API documentation.


  • The repoze.bfg.testing.registerRoutesMapper API (added in an early 1.2 alpha) was deprecated. Its import now generates a deprecation warning.

1.2a7 (2009-12-20)


  • Add four new testing-related APIs to the repoze.bfg.configuration.Configurator class: testing_securitypolicy, testing_models, testing_add_subscriber, and testing_add_template. These were added in order to provide more direct access to the functionality of the repoze.bfg.testing APIs named registerDummySecurityPolicy, registerModels, registerEventListener, and registerTemplateRenderer when a configurator is used. The testing APIs named are nominally deprecated (although they will likely remain around "forever", as they are in heavy use in the wild).

  • Add a new API to the repoze.bfg.configuration.Configurator class: add_settings. This API can be used to add "settings" (information returned within via the repoze.bfg.settings.get_settings API) after the configurator has been initially set up. This is most useful for testing purposes.

  • Add a custom_predicates argument to the Configurator add_view method, the bfg_view decorator and the attribute list of the ZCML view directive. If custom_predicates is specified, it must be a sequence of predicate callables (a predicate callable accepts two arguments: context and request and returns True or False). The associated view callable will only be invoked if all custom predicates return True. Use one or more custom predicates when no existing predefined predicate is useful. Predefined and custom predicates can be mixed freely.

  • Add a custom_predicates argument to the Configurator add_route and the attribute list of the ZCML route directive. If custom_predicates is specified, it must be a sequence of predicate callables (a predicate callable accepts two arguments: context and request and returns True or False). The associated route will match will only be invoked if all custom predicates return True, else route matching continues. Note that the value context will always be None when passed to a custom route predicate. Use one or more custom predicates when no existing predefined predicate is useful. Predefined and custom predicates can be mixed freely.


  • Remove the repoze.bfg.testing.registerTraverser function. This function was never an API.


  • Doc-deprecated most helper functions in the repoze.bfg.testing module. These helper functions likely won't be removed any time soon, nor will they generate a warning any time soon, due to their heavy use in the wild, but equivalent behavior exists in methods of a Configurator.

1.2a6 (2009-12-18)


  • The Configurator object now has two new methods: begin and end. The begin method is meant to be called before any "configuration" begins (e.g. before add_view, et. al are called). The end method is meant to be called after all "configuration" is complete.

    Previously, before there was imperative configuration at all (1.1 and prior), configuration begin and end was invariably implied by the process of loading a ZCML file. When a ZCML load happened, the threadlocal data structure containing the request and registry was modified before the load, and torn down after the load, making sure that all framework code that needed get_current_registry for the duration of the ZCML load was satisfied.

    Some API methods called during imperative configuration, (such as Configurator.add_view when a renderer is involved) end up for historical reasons calling get_current_registry. However, in 1.2a5 and below, the Configurator supplied no functionality that allowed people to make sure that get_current_registry returned the registry implied by the configurator being used. begin now serves this purpose. Inversely, end pops the thread local stack, undoing the actions of begin.

    We make this boundary explicit to reduce the potential for confusion when the configurator is used in different circumstances (e.g. in unit tests and app code vs. just in initial app setup).

    Existing code written for 1.2a1-1.2a5 which does not call begin or end continues to work in the same manner it did before. It is however suggested that this code be changed to call begin and end to reduce the potential for confusion in the future.

  • All paster templates which generate an application skeleton now make use of the new begin and end methods of the Configurator they use in their respective copies of and


  • All documentation that makes use of a Configurator object to do application setup and test setup now makes use of the new begin and end methods of the configurator.

Bug Fixes

  • When a repoze.bfg.exceptions.NotFound or repoze.bfg.exceptions.Forbidden class (as opposed to instance) was raised as an exception within a root factory (or route root factory), the exception would not be caught properly by the repoze.bfg. Router and it would propagate to up the call stack, as opposed to rendering the not found view or the forbidden view as would have been expected.

  • When Chameleon page or text templates used as renderers were added imperatively (via Configurator.add_view or some derivative), they too-eagerly attempted to look up the reload_templates setting via get_settings, meaning they were always registered in non-auto-reload-mode (the default). Each now waits until its respective template attribute is accessed to look up the value.

  • When a route with the same name as a previously registered route was added, the old route was not removed from the mapper's routelist. Symptom: the old registered route would be used (and possibly matched) during route lookup when it should not have had a chance to ever be used.

1.2a5 (2009-12-10)


  • When the repoze.bfg.exceptions.NotFound or repoze.bfg.exceptions.Forbidden error is raised from within a custom root factory or the factory of a route, the appropriate response is now sent to the requesting user agent (the result of the notfound view or the forbidden view, respectively). When these errors are raised from within a root factory, the context passed to the notfound or forbidden view will be None. Also, the request will not be decorated with view_name, subpath, context, etc. as would normally be the case if traversal had been allowed to take place.


  • The exception class representing the error raised by various methods of a Configurator is now importable as repoze.bfg.exceptions.ConfigurationError.


  • General documentation freshening which takes imperative configuration into account in more places and uses glossary references more liberally.

  • Remove explanation of changing the request type in a new request event subscriber, as other predicates are now usually an easier way to get this done.

  • Added "Thread Locals" narrative chapter to documentation, and added a API chapter documenting the repoze.bfg.threadlocals module.

  • Added a "Special Exceptions" section to the "Views" narrative documentation chapter explaining the effect of raising repoze.bfg.exceptions.NotFound and repoze.bfg.exceptions.Forbidden from within view code.


  • A new dependency on the twill package was added to the tests_require argument (Twill will only be downloaded when repoze.bfg test or nosetests is invoked).

1.2a4 (2009-12-07)


  • repoze.bfg.testing.DummyModel now accepts a new constructor keyword argument: __provides__. If this constructor argument is provided, it should be an interface or a tuple of interfaces. The resulting model will then provide these interfaces (they will be attached to the constructed model via zope.interface.alsoProvides).

Bug Fixes

  • Operation on GAE was broken, presumably because the repoze.bfg.configuration module began to attempt to import the repoze.bfg.chameleon_zpt and repoze.bfg.chameleon_text modules, and these cannot be used on non-CPython platforms. It now tolerates startup time import failures for these modules, and only raise an import error when a template from one of these packages is actually used.

1.2a3 (2009-12-02)

Bug Fixes

  • The repoze.bfg.url.route_url function inappropriately passed along _query and/or _anchor arguments to the mapper.generate function, resulting in blowups.

  • When two views were registered with differering for interfaces or classes, and the for of first view registered was a superclass of the second, the repoze.bfg view machinery would incorrectly associate the two views with the same "multiview". Multiviews are meant to be collections of views that have exactly the same for/request/viewname values, without taking inheritance into account. Symptom: wrong view callable found even when you had correctly specified a for_ interface/class during view configuration for one or both view configurations.

Backwards Incompatibilities

  • The repoze.bfg.templating module has been removed; it had been deprecated in 1.1 and never actually had any APIs in it.

1.2a2 (2009-11-29)

Bug Fixes

  • The long description of this package (as shown on PyPI) was not valid reStructuredText, and so was not renderable.

  • Trying to use an HTTP method name string such as GET as a request_type predicate argument caused a startup time failure when it was encountered in imperative configuration or in a decorator (symptom: Type Error: Required specification must be a specification). This now works again, although request_method is now the preferred predicate argument for associating a view configuration with an HTTP request method.


  • Fixed "Startup" narrative documentation chapter; it was explaining "the old way" an application constructor worked.

1.2a1 (2009-11-28)


  • An imperative configuration mode.

    A repoze.bfg application can now begin its life as a single Python file. Later, the application might evolve into a set of Python files in a package. Even later, it might start making use of other configuration features, such as ZCML. But neither the use of a package nor the use of non-imperative configuration is required to create a simple repoze.bfg application any longer.

    Imperative configuration makes repoze.bfg competitive with "microframeworks" such as Bottle and Tornado. repoze.bfg has a good deal of functionality that most microframeworks lack, so this is hopefully a "best of both worlds" feature.

    The simplest possible repoze.bfg application is now:

    from webob import Response
    from wsgiref import simple_server
    from repoze.bfg.configuration import Configurator
    def hello_world(request):
        return Response('Hello world!')
    if __name__ == '__main__':
        config = Configurator()
        app = config.make_wsgi_app()
        simple_server.make_server('', 8080, app).serve_forever()
  • A new class now exists: repoze.bfg.configuration.Configurator. This class forms the basis for sharing machinery between "imperatively" configured applications and traditional declaratively-configured applications.

  • The repoze.bfg.testing.setUp function now accepts three extra optional keyword arguments: registry, request and hook_zca.

    If the registry argument is not None, the argument will be treated as the registry that is set as the "current registry" (it will be returned by repoze.bfg.threadlocal.get_current_registry) for the duration of the test. If the registry argument is None (the default), a new registry is created and used for the duration of the test.

    The value of the request argument is used as the "current request" (it will be returned by repoze.bfg.threadlocal.get_current_request) for the duration of the test; it defaults to None.

    If hook_zca is True (the default), the zope.component.getSiteManager function will be hooked with a function that returns the value of registry (or the default-created registry if registry is None) instead of the registry returned by zope.component.getGlobalSiteManager, causing the Zope Component Architecture API (getSiteManager, getAdapter, getUtility, and so on) to use the testing registry instead of the global ZCA registry.

  • The repoze.bfg.testing.tearDown function now accepts an unhook_zca argument. If this argument is True (the default), zope.component.getSiteManager.reset() will be called. This will cause the result of the zope.component.getSiteManager function to be the global ZCA registry (the result of zope.component.getGlobalSiteManager) once again.

  • The module in various repoze.bfg paster templates now use a repoze.bfg.configuration.Configurator class instead of the (now-legacy) repoze.bfg.router.make_app function to produce a WSGI application.


  • The documentation now uses the "request-only" view calling convention in most examples (as opposed to the context, request convention). This is a documentation-only change; the context, request convention is also supported and documented, and will be "forever".

  • repoze.bfg.configuration API documentation has been added.

  • A narrative documentation chapter entitled "Creating Your First repoze.bfg Application" has been added. This chapter details usage of the new repoze.bfg.configuration.Configurator class, and demonstrates a simplified "imperative-mode" configuration; doing repoze.bfg application configuration imperatively was previously much more difficult.

  • A narrative documentation chapter entitled "Configuration, Decorations and Code Scanning" explaining ZCML- vs. imperative- vs. decorator-based configuration equivalence.

  • The "ZCML Hooks" chapter has been renamed to "Hooks"; it documents how to override hooks now via imperative configuration and ZCML.

  • The explanation about how to supply an alternate "response factory" has been removed from the "Hooks" chapter. This feature may be removed in a later release (it still works now, it's just not documented).

  • Add a section entitled "Test Set Up and Tear Down" to the unittesting chapter.

Bug Fixes

  • The ACL authorization policy debugging output when debug_authorization console debugging output was turned on wasn't as clear as it could have been when a view execution was denied due to an authorization failure resulting from the set of principals passed never having matched any ACE in any ACL in the lineage. Now in this case, we report <default deny> as the ACE value and either the root ACL or <No ACL found on any object in model lineage> if no ACL was found.

  • When two views were registered with the same accept argument, but were otherwise registered with the same arguments, if a request entered the application which had an Accept header that accepted either of the media types defined by the set of views registered with predicates that otherwise matched, a more or less "random" one view would "win". Now, we try harder to use the view callable associated with the view configuration that has the most specific accept argument. Thanks to Alberto Valverde for an initial patch.


  • The routes mapper is no longer a root factory wrapper. It is now consulted directly by the router.

  • The repoze.bfg.registry.make_registry callable has been removed.

  • The repoze.bfg.view.map_view callable has been removed.

  • The repoze.bfg.view.owrap_view callable has been removed.

  • The repoze.bfg.view.predicate_wrap callable has been removed.

  • The repoze.bfg.view.secure_view callable has been removed.

  • The repoze.bfg.view.authdebug_view callable has been removed.

  • The repoze.bfg.view.renderer_from_name callable has been removed. Use repoze.bfg.configuration.Configurator.renderer_from_name instead (still not an API, however).

  • The repoze.bfg.view.derive_view callable has been removed. Use repoze.bfg.configuration.Configurator.derive_view instead (still not an API, however).

  • The repoze.bfg.settings.get_options callable has been removed. Its job has been subsumed by the repoze.bfg.settings.Settings class constructor.

  • The repoze.bfg.view.requestonly function has been moved to repoze.bfg.configuration.requestonly.

  • The repoze.bfg.view.rendered_response function has been moved to repoze.bfg.configuration.rendered_response.

  • The repoze.bfg.view.decorate_view function has been moved to repoze.bfg.configuration.decorate_view.

  • The repoze.bfg.view.MultiView class has been moved to repoze.bfg.configuration.MultiView.

  • The repoze.bfg.zcml.Uncacheable class has been removed.

  • The repoze.bfg.resource.resource_spec function has been removed.

  • All ZCML directives which deal with attributes which are paths now use the path method of the ZCML context to resolve a relative name to an absolute one (imperative configuration requirement).

  • The repoze.bfg.scripting.get_root API now uses a 'real' WebOb request rather than a FakeRequest when it sets up the request as a threadlocal.

  • The repoze.bfg.traversal.traverse API now uses a 'real' WebOb request rather than a FakeRequest when it calls the traverser.

  • The repoze.bfg.request.FakeRequest class has been removed.

  • Most uses of the ZCA threadlocal API (the getSiteManager, getUtility, getAdapter, getMultiAdapter threadlocal API) have been removed from the core. Instead, when a threadlocal is necessary, the core uses the repoze.bfg.threadlocal.get_current_registry API to obtain the registry.

  • The internal ILogger utility named repoze.bfg.debug is now just an IDebugLogger unnamed utility. A named utility with the old name is registered for b/w compat.

  • The repoze.bfg.interfaces.ITemplateRendererFactory interface was removed; it has become unused.

  • Instead of depending on the martian package to do code scanning, we now just use our own scanning routines.

  • We now no longer have a dependency on repoze.zcml package; instead, the repoze.bfg package includes implementations of the adapter, subscriber and utility directives.

  • Relating to the following functions:












    Each of these functions now expects to be called with a request object that has a registry attribute which represents the current repoze.bfg registry. They fall back to obtaining the registry from the threadlocal API.

Backwards Incompatibilities

  • Unit tests which use zope.testing.cleanup.cleanUp for the purpose of isolating tests from one another may now begin to fail due to lack of isolation between tests.

    Here's why: In repoze.bfg 1.1 and prior, the registry returned by repoze.bfg.threadlocal.get_current_registry when no other registry had been pushed on to the threadlocal stack was the zope.component.globalregistry.base global registry (aka the result of zope.component.getGlobalSiteManager()). In repoze.bfg 1.2+, however, the registry returned in this situation is the new module-scope repoze.bfg.registry.global_registry object. The zope.testing.cleanup.cleanUp function clears the zope.component.globalregistry.base global registry unconditionally. However, it does not know about the repoze.bfg.registry.global_registry object, so it does not clear it.

    If you use the zope.testing.cleanup.cleanUp function in the setUp of test cases in your unit test suite instead of using the (more correct as of 1.1) repoze.bfg.testing.setUp, you will need to replace all calls to zope.testing.cleanup.cleanUp with a call to repoze.bfg.testing.setUp.

    If replacing all calls to zope.testing.cleanup.cleanUp with a call to repoze.bfg.testing.setUp is infeasible, you can put this bit of code somewhere that is executed exactly once (not for each test in a test suite; in the ```` of your package or your package's tests subpackage would be a reasonable place):

    import zope.testing.cleanup
    from repoze.bfg.testing import setUp
  • When there is no "current registry" in the repoze.bfg.threadlocal.manager threadlocal data structure (this is the case when there is no "current request" or we're not in the midst of a r.b.testing.setUp-bounded unit test), the .get method of the manager returns a data structure containing a global registry. In previous releases, this function returned the global Zope "base" registry: the result of zope.component.getGlobalSiteManager, which is an instance of the zope.component.registry.Component class. In this release, however, the global registry returns a globally importable instance of the repoze.bfg.registry.Registry class. This registry instance can always be imported as repoze.bfg.registry.global_registry.

    Effectively, this means that when you call repoze.bfg.threadlocal.get_current_registry when no request or setUp bounded unit test is in effect, you will always get back the global registry that lives in repoze.bfg.registry.global_registry. It also means that repoze.bfg APIs that call get_current_registry will use this registry.

    This change was made because repoze.bfg now expects the registry it uses to have a slightly different API than a bare instance of zope.component.registry.Components.

  • View registration no longer registers a repoze.bfg.interfaces.IViewPermission adapter (it is no longer checked by the framework; since 1.1, views have been responsible for providing their own security).

  • The repoze.bfg.router.make_app callable no longer accepts the authentication_policy nor the authorization_policy arguments. This feature was deprecated in version 1.0 and has been removed.

  • Obscure: the machinery which configured views with a request_type and a route_name would ignore the request interface implied by route_name registering a view only for the interface implied by request_type. In the unlikely event that you were trying to use these two features together, the symptom would have been that views that named a request_type but which were also associated with routes were not found when the route matched. Now if a view is configured with both a request_type and a route_name, an error is raised.

  • The route ZCML directive now no longer accepts the request_type or view_request_type attributes. These attributes didn't actually work in any useful way (see entry above this one).

  • Because the repoze.bfg package now includes implementations of the adapter, subscriber and utility ZCML directives, it is now an error to have <include package="repoze.zcml" file="meta.zcml"/> in the ZCML of a repoze.bfg application. A ZCML conflict error will be raised if your ZCML does so. This shouldn't be an issue for "normal" installations; it has always been the responsibility of the repoze.bfg.includes ZCML to include this file in the past; it now just doesn't.

  • The repoze.bfg.testing.zcml_configure API was removed. Use the Configurator.load_zcml API instead.


  • The repoze.bfg.router.make_app function is now nominally deprecated. Its import and usage does not throw a warning, nor will it probably ever disappear. However, using a repoze.bfg.configuration.Configurator class is now the preferred way to generate a WSGI application.

    Note that make_app calls zope.component.getSiteManager.sethook( repoze.bfg.threadlocal.get_current_registry) on the caller's behalf, hooking ZCA global API lookups, for backwards compatibility purposes. If you disuse make_app, your calling code will need to perform this call itself, at least if your application uses the ZCA global API (getSiteManager, getAdapter, etc).


  • A dependency on the martian package has been removed (its functionality is replaced internally).

  • A dependency on the repoze.zcml package has been removed (its functionality is replaced internally).

1.1.1 (2009-11-21)

Bug Fixes

  • "Hybrid mode" applications (applications which explicitly used traversal after url dispatch via <route> paths containing the *traverse element) were broken in 1.1-final and all 1.1 alpha and beta releases. Views registered without a route_name route shadowed views registered with a route_name inappropriately.

1.1 (2009-11-15)


  • Remove dead IRouteRequirement interface from repoze.bfg.zcml module.


  • Improve the "Extending an Existing Application" narrative chapter.

  • Add more sections to the "Defending Design" chapter.

1.1b4 (2009-11-12)

Bug Fixes

  • Use alsoProvides in the urldispatch module to attach an interface to the request rather than directlyProvides to avoid disturbing interfaces set in a NewRequest event handler.


  • Move 1.0.1 and previous changelog to HISTORY.txt.

  • Add examples to repoze.bfg.url.model_url docstring.

  • Add "Defending BFG Design" chapter to frontpage docs.


  • Remove and its import from all paster templates, samples, and tutorials for distribute compatibility. The documentation already explains how to install virtualenv (which will include some setuptools package), so these files, imports and usages were superfluous.


  • The options kw arg to the repoze.bfg.router.make_app function is deprecated. In its place is the keyword argument settings. The options keyword continues to work, and a deprecation warning is not emitted when it is detected. However, the paster templates, code samples, and documentation now make reference to settings rather than options. This change/deprecation was mainly made for purposes of clarity and symmetry with the get_settings() API and discussions of "settings" in various places in the docs: we want to use the same name to refer to the same thing everywhere.

1.1b3 (2009-11-06)


  • repoze.bfg.testing.registerRoutesMapper testing facility added. This testing function registers a routes "mapper" object in the registry, for tests which require its presence. This function is documented in the repoze.bfg.testing API documentation.

Bug Fixes

  • Compound statements that used an assignment entered into in an interactive IPython session invoked via paster bfgshell no longer fail to mutate the shell namespace correctly. For example, this set of statements used to fail:

    In [2]: def bar(x): return x
    In [3]: list(bar(x) for x in 'abc')
    Out[3]: NameError: 'bar'

    In this release, the bar function is found and the correct output is now sent to the console. Thanks to Daniel Holth for the patch.

  • The bfgshell command did not function properly; it was still expecting to be able to call the root factory with a bare environ rather than a request object.

Backwards Incompatibilities

  • The repoze.bfg.scripting.get_root function now expects a request object as its second argument rather than an environ.

1.1b2 (2009-11-02)

Bug Fixes

  • Prevent PyPI installation failure due to easy_install trying way too hard to guess the best version of Paste. When easy_install pulls from PyPI it reads links off various pages to determine "more up to date" versions. It incorrectly picks up a link for an ancient version of a package named "Paste-Deploy-0.1" (note the dash) when trying to find the "Paste" distribution and somehow believes it's the latest version of "Paste". It also somehow "helpfully" decides to check out a version of this package from SVN. We pin the Paste dependency version to a version greater than 1.7 to work around this easy_install bug.


  • Fix "Hybrid" narrative chapter: stop claiming that <view> statements that mention a route_name need to come afer (in XML order) the <route> statement which creates the route. This hasn't been true since 1.1a1.

  • "What's New in repoze.bfg 1.1" document added to narrative documentation.


  • Add a new event type: Events of this type will be sent after traversal is completed, but before any view code is invoked. Like, This event will have a single attribute: request representing the current request. Unlike the request attribute of however, during an AfterTraversal event, the request object will possess attributes set by the traverser, most notably context, which will be the context used when a view is found and invoked. The interface can be used to subscribe to the event. For example:

    <subscriber for="repoze.bfg.interfaces.IAfterTraversal"

    Like any framework event, a subscriber function should expect one parameter: event.


  • Rather than depending on chameleon.core and chameleon.zpt distributions individually, depend on Malthe's repackaged Chameleon distribution (which includes both chameleon.core and chameleon.zpt).

1.1b1 (2009-11-01)

Bug Fixes

  • The routes root factory called route factories and the default route factory with an environ rather than a request. One of the symptoms of this bug: applications generated using the bfg_zodb paster template in 1.1a9 did not work properly.

  • Reinstate renderer alias for view_renderer in the <route> ZCML directive (in-the-wild 1.1a bw compat).

  • bfg_routesalchemy paster template: change <route> declarations: rename renderer attribute to view_renderer.

  • Header values returned by the authtktauthenticationpolicy remember and forget methods would be of type unicode. This violated the WSGI spec, causing a TypeError to be raised when these headers were used under mod_wsgi.

  • If a BFG app that had a route matching the root URL was mounted under a path in modwsgi, ala WSGIScriptAlias /myapp /Users/chrism/projects/modwsgi/env/bfg.wsgi, the home route (a route with the path of '/' or '') would not match when the path /myapp was visited (only when the path /myapp/ was visited). This is now fixed: if the urldispatch root factory notes that the PATH_INFO is empty, it converts it to a single slash before trying to do matching.


  • In <route> declarations in tutorial ZCML, rename renderer attribute to view_renderer (fwd compat).

  • Fix various tutorials broken by 1.1a9 <route> directive changes.


  • Deal with a potential circref in the traversal module.

1.1a9 (2009-10-31)

Bug Fixes

  • An incorrect ZCML conflict would be encountered when the request_param predicate attribute was used on the ZCML view directive if any two otherwise same-predicated views had the combination of a predicate value with an = sign and one without (e.g. a vs. a=123).


  • In previous versions of BFG, the "root factory" (the get_root callable passed to make_app or a function pointed to by the factory attribute of a route) was called with a "bare" WSGI environment. In this version, and going forward, it will be called with a request object. The request object passed to the factory implements dictionary-like methods in such a way that existing root factory code which expects to be passed an environ will continue to work.

  • The __call__ of a plugin "traverser" implementation (registered as an adapter for ITraverser or ITraverserFactory) will now receive a request as the single argument to its __call__ method. In previous versions it was passed a WSGI environ object. The request object passed to the factory implements dictionary-like methods in such a way that existing traverser code which expects to be passed an environ will continue to work.

  • The ZCML route directive's attributes xhr, request_method, path_info, request_param, header and accept are now route predicates rather than view predicates. If one or more of these predicates is specified in the route configuration, all of the predicates must return true for the route to match a request. If one or more of the route predicates associated with a route returns False when checked during a request, the route match fails, and the next match in the routelist is tried. This differs from the previous behavior, where no route predicates existed and all predicates were considered view predicates, because in that scenario, the next route was not tried.


  • Various changes were made to narrative and API documentation supporting the change from passing a request rather than an environ to root factories and traversers.


  • The request implements dictionary-like methods that mutate and query the WSGI environ. This is only for the purpose of backwards compatibility with root factories which expect an environ rather than a request.

  • The repoze.bfg.request.create_route_request_factory function, which returned a request factory was removed in favor of a repoze.bfg.request.route_request_interface function, which returns an interface.

  • The repoze.bfg.request.Request class, which is a subclass of webob.Request now defines its own __setattr__, __getattr__ and __delattr__ methods, which override the default WebOb behavior. The default WebOb behavior stores attributes of the request in self.environ['webob.adhoc_attrs'], and retrieves them from that dictionary during a __getattr__. This behavior was undesirable for speed and "expectation" reasons. Now attributes of the request are stored in request.__dict__ (as you otherwise might expect from an object that did not override these methods).

  • The router no longer calls repoze.bfg.traversal._traverse and does its work "inline" (speed).

  • Reverse the order in which the router calls the request factory and the root factory. The request factory is now called first; the resulting request is passed to the root factory.

  • The repoze.bfg.request.request_factory function has been removed. Its functionality is no longer required.

  • The "routes root factory" that wraps the default root factory when there are routes mentioned in the configuration now attaches an interface to the request via zope.interface.directlyProvides. This replaces logic in the (now-gone) repoze.bfg.request.request_factory function.

  • The route and view ZCML directives now register an interface as a named utility (retrieved from repoze.bfg.request.route_request_interface) rather than a request factory (the previous return value of the now-missing repoze.bfg.request.create_route_request_factory.

  • The repoze.bfg.functional module was renamed to repoze.bfg.compat.

Backwards Incompatibilities

  • Explicitly revert the feature introduced in 1.1a8: where the name root is available as an attribute of the request before a NewRequest event is emitted. This makes some potential future features impossible, or at least awkward (such as grouping traversal and view lookup into a single adapter lookup).

  • The containment, attr and renderer attributes of the route ZCML directive were removed.

1.1a8 (2009-10-27)


  • Add path_info view configuration predicate.

  • paster bfgshell now supports IPython if it's available for import. Thanks to Daniel Holth for the initial patch.

  • Add repoze.bfg.testing.registerSettings API, which is documented in the "repoze.bfg.testing" API chapter. This allows for registration of "settings" values obtained via repoze.bfg.settings.get_settings() for use in unit tests.

  • The name root is available as an attribute of the request slightly earlier now (before a NewRequest event is emitted). root is the result of the application "root factory".

  • Added max_age parameter to authtktauthenticationpolicy ZCML directive. If this value is set, it must be an integer representing the number of seconds which the auth tkt cookie will survive. Mainly, its existence allows the auth_tkt cookie to survive across browser sessions.

Bug Fixes

  • Fix bug encountered during "scan" (when <scan ..> directive is used in ZCML) introduced in 1.1a7. Symptom: AttributeError: object has no attribute __provides__ raised at startup time.

  • The reissue_time argument to the authtktauthenticationpolicy ZCML directive now actually works. When it is set to an integer value, an authticket set-cookie header is appended to the response whenever a request requires authentication and 'now' minus the authticket's timestamp is greater than reissue_time seconds.


  • Add a chapter titled "Request and Response" to the narrative documentation, content cribbed from the WebOb documentation.

  • Call out predicate attributes of ZCML directive within "Views" chapter.

  • Fix route_url documentation (_query argument documented as query and _anchor argument documented as anchor).

Backwards Incompatibilities

  • The authtkt authentication policy remember method now no longer honors token or userdata keyword arguments.


  • Change how bfg_view decorator works when used as a class method decorator. In 1.1a7, the``scan``directive actually tried to grope every class in scanned package at startup time, calling dir against each found class, and subsequently invoking getattr against each thing found by dir to see if it was a method. This led to some strange symptoms (e.g. AttributeError: object has no attribute __provides__), and was generally just a bad idea. Now, instead of groping classes for methods at startup time, we just cause the bfg_view decorator itself to populate the method's class' __dict__ when it is used as a method decorator. This also requires a nasty _getframe thing but it's slightly less nasty than the startup time groping behavior. This is essentially a reversion back to 1.1a6 "grokking" behavior plus some special magic for using the bfg_view decorator as method decorator inside the bfg_view class itself.

  • The router now checks for a global_response_headers attribute of the request object before returning a response. If this value exists, it is presumed to be a sequence of two-tuples, representing a set of headers to append to the 'normal' response headers. This feature is internal, rather than exposed externally, because it's unclear whether it will stay around in the long term. It was added to support the reissue_time feature of the authtkt authentication policy.

  • The interface ITraverserFactory is now just an alias for ITraverser.

1.1a7 (2009-10-18)


  • More than one @bfg_view decorator may now be stacked on top of any number of others. Each invocation of the decorator registers a single view configuration. For instance, the following combination of decorators and a function will register two view configurations for the same view callable:

    from repoze.bfg.view import bfg_view
    def edit(context, request):

    This makes it possible to associate more than one view configuration with a single callable without requiring any ZCML.

  • The @bfg_view decorator can now be used against a class method:

    from webob import Response
    from repoze.bfg.view import bfg_view
    class MyView(object):
        def __init__(self, context, request):
            self.context = context
            self.request = request
        def amethod(self):
            return Response('hello from %s!' % self.context)

    When the bfg_view decorator is used against a class method, a view is registered for the class (it's a "class view" where the "attr" happens to be the name of the method it is attached to), so the class it's defined within must have a suitable constructor: one that accepts context, request or just request.


  • Added Changing the Traverser and Changing How :mod:`repoze.bfg.url.model_url` Generates a URL to the "Hooks" narrative chapter of the docs.


  • Remove and imports of it within In the new world, and as per virtualenv setup instructions, people will already have either setuptools or distribute.

1.1a6 (2009-10-15)


  • Add xhr, accept, and header view configuration predicates to ZCML view declaration, ZCML route declaration, and bfg_view decorator. See the Views narrative documentation chapter for more information about these predicates.

  • Add setUp and tearDown functions to the repoze.bfg.testing module. Using setUp in a test setup and tearDown in a test teardown is now the recommended way to do component registry setup and teardown. Previously, it was recommended that a single function named repoze.bfg.testing.cleanUp be called in both the test setup and tear down. repoze.bfg.testing.cleanUp still exists (and will exist "forever" due to its widespread use); it is now just an alias for repoze.bfg.testing.setUp and is nominally deprecated.

  • The BFG component registry is now available in view and event subscriber code as an attribute of the request ie. request.registry. This fact is currently undocumented except for this note, because BFG developers never need to interact with the registry directly anywhere else.

  • The BFG component registry now inherits from dict, meaning that it can optionally be used as a simple dictionary. Component registrations performed against it via e.g. registerUtility, registerAdapter, and similar API methods are kept in a completely separate namespace than its dict members, so using the its component API methods won't effect the keys and values in the dictionary namespace. Likewise, though the component registry "happens to be" a dictionary, use of mutating dictionary methods such as __setitem__ will have no influence on any component registrations made against it. In other words, the registry object you obtain via e.g. repoze.bfg.threadlocal.get_current_registry or request.registry happens to be both a component registry and a dictionary, but using its component-registry API won't impact data added to it via its dictionary API and vice versa. This is a forward compatibility move based on the goals of "marco".

  • Expose and document repoze.bfg.testing.zcml_configure API. This function populates a component registry from a ZCML file for testing purposes. It is documented in the "Unit and Integration Testing" chapter.


  • Virtual hosting narrative docs chapter updated with info about mod_wsgi.

  • Point all index URLs at the literal 1.1 index (this alpha cycle may go on a while).

  • Various tutorial test modules updated to use repoze.bfg.testing.setUp and repoze.bfg.testing.tearDown methods in order to encourage this as best practice going forward.

  • Added "Creating Integration Tests" section to unit testing narrative documentation chapter. As a result, the name of the unittesting chapter is now "Unit and Integration Testing".

Backwards Incompatibilities

  • Importing getSiteManager and get_registry from repoze.bfg.registry is no longer supported. These imports were deprecated in repoze.bfg 1.0. Import of getSiteManager should be done as from zope.component import getSiteManager. Import of get_registry should be done as from repoze.bfg.threadlocal import get_current_registry. This was done to prevent a circular import dependency.

  • Code bases which alternately invoke both zope.testing.cleanup.cleanUp and repoze.bfg.testing.cleanUp (treating them equivalently, using them interchangeably) in the setUp/tearDown of unit tests will begin to experience test failures due to lack of test isolation. The "right" mechanism is repoze.bfg.testing.cleanUp (or the combination of repoze.bfg.testing.setUp and repoze.bfg.testing.tearDown). but a good number of legacy codebases will use zope.testing.cleanup.cleanUp instead. We support zope.testing.cleanup.cleanUp but not in combination with repoze.bfg.testing.cleanUp in the same codebase. You should use one or the other test cleanup function in a single codebase, but not both.


  • Created new repoze.bfg.configuration module which assumes responsibilities previously held by the repoze.bfg.registry and repoze.bfg.router modules (avoid a circular import dependency).

  • The result of the zope.component.getSiteManager function in unit tests set up with repoze.bfg.testing.cleanUp or repoze.bfg.testing.setUp will be an instance of repoze.bfg.registry.Registry instead of the global zope.component.globalregistry.base registry. This also means that the threadlocal ZCA API functions such as getAdapter and getUtility as well as internal BFG machinery (such as model_url and route_url) will consult this registry within unit tests. This is a forward compatibility move based on the goals of "marco".

  • Removed repoze.bfg.testing.addCleanUp function and associated module-scope globals. This was never an API.

1.1a5 (2009-10-10)


  • Change "Traversal + ZODB" and "URL Dispatch + SQLAlchemy" Wiki tutorials to make use of the new-to-1.1 "renderer" feature (return dictionaries from all views).

  • Add tests to the "URL Dispatch + SQLAlchemy" tutorial after the "view" step.

  • Added a diagram of model graph traversal to the "Traversal" narrative chapter of the documentation.

  • An exceptions API chapter was added, documenting the new repoze.bfg.exceptions module.

  • Describe "request-only" view calling conventions inside the urldispatch narrative chapter, where it's most helpful.

  • Add a diagram which explains the operation of the BFG router to the "Router" narrative chapter.


  • Add a new repoze.bfg.testing API: registerRoute, for registering routes to satisfy calls to e.g. repoze.bfg.url.route_url in unit tests.

  • The notfound and forbidden ZCML directives now accept the following additional attributes: attr, renderer, and wrapper. These have the same meaning as they do in the context of a ZCML view directive.

  • For behavior like Django's APPEND_SLASH=True, use the repoze.bfg.view.append_slash_notfound_view view as the Not Found view in your application. When this view is the Not Found view (indicating that no view was found), and any routes have been defined in the configuration of your application, if the value of PATH_INFO does not already end in a slash, and if the value of PATH_INFO plus a slash matches any route's path, do an HTTP redirect to the slash-appended PATH_INFO. Note that this will lose POST data information (turning it into a GET), so you shouldn't rely on this to redirect POST requests.

  • Speed up repoze.bfg.location.lineage slightly.

  • Speed up repoze.bfg.encode.urlencode (nee' repoze.bfg.url.urlencode) slightly.

  • Speed up repoze.bfg.traversal.model_path.

  • Speed up repoze.bfg.traversal.model_path_tuple slightly.

  • Speed up repoze.bfg.traversal.traverse slightly.

  • Speed up repoze.bfg.url.model_url slightly.

  • Speed up repoze.bfg.url.route_url slightly.

  • Sped up repoze.bfg.traversal.ModelGraphTraverser:__call__ slightly.

  • Minor speedup of repoze.bfg.router.Router.__call__.

  • New repoze.bfg.exceptions module was created to house exceptions that were previously sprinkled through various modules.


  • Move repoze.bfg.traversal._url_quote into repoze.bfg.encode as url_quote.


  • The import of repoze.bfg.view.NotFound is deprecated in favor of repoze.bfg.exceptions.NotFound. The old location still functions, but emits a deprecation warning.

  • The import of is deprecated in favor of repoze.bfg.exceptions.Forbidden. The old location still functions but emits a deprecation warning. The rename from Unauthorized to Forbidden brings parity to the name of the exception and the system view it invokes when raised.

Backwards Incompatibilities

  • We previously had a Unicode-aware wrapper for the urllib.urlencode function named repoze.bfg.url.urlencode which delegated to the stdlib function, but which marshalled all unicode values to utf-8 strings before calling the stdlib version. A newer replacement now lives in repoze.bfg.encode The replacement does not delegate to the stdlib.

    The replacement diverges from the stdlib implementation and the previous repoze.bfg.url url implementation inasmuch as its doseq argument is now a decoy: it always behaves in the doseq=True way (which is the only sane behavior) for speed purposes.

    The old import location (repoze.bfg.url.urlencode) still functions and has not been deprecated.

  • In 0.8a7, the return value expected from an object implementing ITraverserFactory was changed from a sequence of values to a dictionary containing the keys context, view_name, subpath, traversed, virtual_root, virtual_root_path, and root. Until now, old-style traversers which returned a sequence have continued to work but have generated a deprecation warning. In this release, traversers which return a sequence instead of a dictionary will no longer work.

1.1a4 (2009-09-23)

Bug Fixes

  • On 64-bit Linux systems, views that were members of a multiview (orderings of views with predicates) were not evaluated in the proper order. Symptom: in a configuration that had two views with the same name but one with a request_method=POST predicate and one without, the one without the predicate would be called unconditionally (even if the request was a POST request). Thanks much to Sebastien Douche for providing the buildbots that pointed this out.


  • Added a tutorial which explains how to use repoze.session (ZODB-based sessions) in a ZODB-based repoze.bfg app.

  • Added a tutorial which explains how to add ZEO to a ZODB-based repoze.bfg application.

  • Added a tutorial which explains how to run a repoze.bfg application under mod_wsgi. See "Running a repoze.bfg Application under mod_wsgi" in the tutorials section of the documentation.


  • Add a repoze.bfg.url.static_url API which is capable of generating URLs to static resources defined by the <static> ZCML directive. See the "Views" narrative chapter's section titled "Generating Static Resource URLs" for more information.

  • Add a string renderer. This renderer converts a non-Response return value of any view callble into a string. It is documented in the "Views" narrative chapter.

  • Give the route ZCML directive the view_attr and view_renderer parameters (bring up to speed with 1.1a3 features). These can also be spelled as attr and renderer.

Backwards Incompatibilities

  • An object implementing the IRenderer interface (and ITemplateRenderer`, which is a subclass of ``IRenderer) must now accept an extra system argument in its __call__ method implementation. Values computed by the system (as opposed to by the view) are passed by the system in the system parameter, which will always be a dictionary. Keys in the dictionary include: view (the view object that returned the value), renderer_name (the template name or simple name of the renderer), context (the context object passed to the view), and request (the request object passed to the view). Previously only ITemplateRenderers received system arguments as elements inside the main value dictionary.


  • The way bfg_view declarations are scanned for has been modified. This should have no external effects.

  • Speed: do not register an ITraverserFactory in configure.zcml; instead rely on queryAdapter and a manual default to ModelGraphTraverser.

  • Speed: do not register an IContextURL in configure.zcml; instead rely on queryAdapter and a manual default to TraversalContextURL.

  • General speed microimprovements for helloworld benchmark: replace try/excepts with statements which use 'in' keyword.

1.1a3 (2009-09-16)


  • The "Views" narrative chapter in the documentation has been updated extensively to discuss "renderers".


  • A renderer attribute has been added to view configurations, replacing the previous (1.1a2) version's template attribute. A "renderer" is an object which accepts the return value of a view and converts it to a string. This includes, but is not limited to, templating systems.

  • A new interface named IRenderer was added. The existing interface, ITemplateRenderer now derives from this new interface. This interface is internal.

  • A new interface named IRendererFactory was added. An existing interface named ITemplateRendererFactory now derives from this interface. This interface is internal.

  • The view attribute of the view ZCML directive is no longer required if the ZCML directive also has a renderer attribute. This is useful when the renderer is a template renderer and no names need be passed to the template at render time.

  • A new zcml directive renderer has been added. It is documented in the "Views" narrative chapter of the documentation.

  • A ZCML view directive (and the associated bfg_view decorator) can now accept a "wrapper" value. If a "wrapper" value is supplied, it is the value of a separate view's name attribute. When a view with a wrapper attribute is rendered, the "inner" view is first rendered normally. Its body is then attached to the request as "wrapped_body", and then a wrapper view name is looked up and rendered (using repoze.bfg.render_view_to_response), passed the request and the context. The wrapper view is assumed to do something sensible with request.wrapped_body, usually inserting its structure into some other rendered template. This feature makes it possible to specify (potentially nested) "owrap" relationships between views using only ZCML or decorators (as opposed always using ZPT METAL and analogues to wrap view renderings in outer wrappers).


  • When used under Python < 2.6, BFG now has an installation time dependency on the simplejson package.


  • The repoze.bfg.testing.registerDummyRenderer API has been deprecated in favor of repoze.bfg.testing.registerTemplateRenderer. A deprecation warning is not issued at import time for the former name; it will exist "forever"; its existence has been removed from the documentation, however.

  • The repoze.bfg.templating.renderer_from_cache function has been moved to repoze.bfg.renderer.template_renderer_factory. This was never an API, but code in the wild was spotted that used it. A deprecation warning is issued at import time for the former.

Backwards Incompatibilities

  • The ITemplateRenderer interface has been changed. Previously its __call__ method accepted **kw. It now accepts a single positional parameter named kw (REVISED: it accepts two positional parameters as of 1.1a4: value and system). This is mostly an internal change, but it was exposed in APIs in one place: if you've used the repoze.bfg.testing.registerDummyRenderer API in your tests with a custom "renderer" argument with your own renderer implementation, you will need to change that renderer implementation to accept kw instead of **kw in its __call__ method (REVISED: make it accept value and system positional arguments as of 1.1a4).

  • The ITemplateRendererFactory interface has been changed. Previously its __call__ method accepted an auto_reload keyword parameter. Now its __call__ method accepts no keyword parameters. Renderers are now themselves responsible for determining details of auto-reload. This is purely an internal change. This interface was never external.

  • The template_renderer ZCML directive introduced in 1.1a2 has been removed. It has been replaced by the renderer directive.

  • The previous release (1.1a2) added a view configuration attribute named template. In this release, the attribute has been renamed to renderer. This signifies that the attribute is more generic: it can now be not just a template name but any renderer name (ala json).

  • In the previous release (1.1a2), the Chameleon text template renderer was used if the system didn't associate the template view configuration value with a filename with a "known" extension. In this release, you must use a renderer attribute which is a path that ends with a .txt extension (e.g. templates/foo.txt) to use the Chameleon text renderer.

1.1a2 (2009-09-14)


  • A ZCML view directive (and the associated bfg_view decorator) can now accept an "attr" value. If an "attr" value is supplied, it is considered a method named of the view object to be called when the response is required. This is typically only good for views that are classes or instances (not so useful for functions, as functions typically have no methods other than __call__).

  • A ZCML view directive (and the associated bfg_view decorator) can now accept a "template" value. If a "template" value is supplied, and the view callable returns a dictionary, the associated template is rendered with the dictionary as keyword arguments. See the section named "Views That Have a template" in the "Views" narrative documentation chapter for more information.

1.1a1 (2009-09-06)

Bug Fixes

  • "tests" module removed from the bfg_alchemy paster template; these tests didn't work.

  • Bugfix: the discriminator for the ZCML "route" directive was incorrect. It was possible to register two routes that collided without the system spitting out a ConfigurationConflictError at startup time.


  • Feature addition: view predicates. These are exposed as the request_method, request_param, and containment attributes of a ZCML view declaration, or the respective arguments to a @bfg_view decorator. View predicates can be used to register a view for a more precise set of environment parameters than was previously possible. For example, you can register two views with the same name with different request_param attributes. If the request.params dict contains 'foo' (request_param="foo"), one view might be called; if it contains 'bar' (request_param="bar"), another view might be called. request_param can also name a key/value pair ala foo=123. This will match only when the foo key is in the request.params dict and it has the value '123'. This particular example makes it possible to write separate view functions for different form submissions. The other predicates, containment and request_method work similarly. containment is a view predicate that will match only when the context's graph lineage has an object possessing a particular class or interface, for example. request_method is a view predicate that will match when the HTTP REQUEST_METHOD equals some string (eg. 'POST').

  • The @bfg_view decorator now accepts three additional arguments: request_method, request_param, and containment. request_method is used when you'd like the view to match only a request with a particular HTTP REQUEST_METHOD; a string naming the REQUEST_METHOD can also be supplied as request_type for backwards compatibility. request_param is used when you'd like a view to match only a request that contains a particular request.params key (with or without a value). containment is used when you'd like to match a request that has a context that has some class or interface in its graph lineage. These are collectively known as "view predicates".

  • The route ZCML directive now honors view_request_method, view_request_param and view_containment attributes, which pass along these values to the associated view if any is provided. Additionally, the request_type attribute can now be spelled as view_request_type, and permission can be spelled as view_permission. Any attribute which starts with view_ can now be spelled without the view_ prefix, so view_for can be spelled as for now, etc. Both forms are documented in the urldispatch narrative documentation chapter.

  • The request_param ZCML view directive attribute (and its bfg_view decorator cousin) can now specify both a key and a value. For example, request_param="foo=123" means that the foo key must have a value of 123 for the view to "match".

  • Allow repoze.bfg.traversal.find_interface API to use a class object as the argument to compare against the model passed in. This means you can now do find_interface(model, SomeClass) and the first object which is found in the lineage which has SomeClass as its class (or the first object found which has SomeClass as any of its superclasses) will be returned.

  • Added static ZCML directive which registers a route for a view that serves up files in a directory. See the "Views" narrative documentation chapter's "Serving Static Resources Using a ZCML Directive" section for more information.

  • The repoze.bfg.view.static class now accepts a string as its first argument ("root_dir") that represents a package-relative name e.g. somepackage:foo/bar/static. This is now the preferred mechanism for spelling package-relative static paths using this class. A package_name keyword argument has been left around for backwards compatibility. If it is supplied, it will be honored.

  • The API repoze.bfg.testing.registerView now takes a permission argument. Use this instead of using repoze.bfg.testing.registerViewPermission.

  • The ordering of route declarations vs. the ordering of view declarations that use a "route_name" in ZCML no longer matters. Previously it had been impossible to use a route_name from a route that had not yet been defined in ZCML (order-wise) within a "view" declaration.

  • The repoze.bfg router now catches both and repoze.bfg.view.NotFound exceptions while rendering a view. When the router catches an Unauthorized, it returns the registered forbidden view. When the router catches a NotFound, it returns the registered notfound view.


  • Change urldispatch internals: Route object is now constructed using a path, a name, and a factory instead of a name, a matcher, a generator, and a factory.

  • Move (non-API) default_view, default_forbidden_view, and default_notfound_view functions into the repoze.bfg.view module (moved from repoze.bfg.router).

  • Removed ViewPermissionFactory from View permission checking is now done by registering and looking up an ISecuredView.

  • The static ZCML directive now uses a custom root factory when constructing a route.

  • The interface IRequestFactories was removed from the repoze.bfg.interfaces module. This interface was never an API.

  • The function named named_request_factories and the data structure named DEFAULT_REQUEST_FACTORIES have been removed from the repoze.bfg.request module. These were never APIs.

  • The IViewPermissionFactory interface has been removed. This was never an API.


  • Request-only-convention examples in the "Views" narrative documentation were broken.

  • Fixed documentation bugs related to forget and remember in security API docs.

  • Fixed documentation for repoze.bfg.view.static (in narrative Views chapter).


  • The API repoze.bfg.testing.registerViewPermission has been deprecated.

Backwards Incompatibilities

  • The interfaces IPOSTRequest, IGETRequest, IPUTRequest, IDELETERequest, and IHEADRequest have been removed from the repoze.bfg.interfaces module. These were not documented as APIs post-1.0. Instead of using one of these, use a request_method ZCML attribute or request_method bfg_view decorator parameter containing an HTTP method name (one of GET, POST, HEAD, PUT, DELETE) instead of one of these interfaces if you were using one explicitly. Passing a string in the set (GET, HEAD, PUT, POST, DELETE) as a request_type argument will work too. Rationale: instead of relying on interfaces attached to the request object, BFG now uses a "view predicate" to determine the request type.

  • Views registered without the help of the ZCML view directive are now responsible for performing their own authorization checking.

  • The registry_manager backwards compatibility alias importable from "repoze.bfg.registry", deprecated since repoze.bfg 0.9 has been removed. If you are tring to use the registry manager within a debug script of your own, use a combination of the "repoze.bfg.paster.get_app" and "repoze.bfg.scripting.get_root" APIs instead.

  • The INotFoundAppFactory interface has been removed; it has been deprecated since repoze.bfg 0.9. If you have something like the following in your configure.zcml:

    <utility provides="repoze.bfg.interfaces.INotFoundAppFactory"

    Replace it with something like:


    See "Changing the Not Found View" in the "Hooks" chapter of the documentation for more information.

  • The IUnauthorizedAppFactory interface has been removed; it has been deprecated since repoze.bfg 0.9. If you have something like the following in your configure.zcml:

    <utility provides="repoze.bfg.interfaces.IUnauthorizedAppFactory"

    Replace it with something like:


    See "Changing the Forbidden View" in the "Hooks" chapter of the documentation for more information.

  • ISecurityPolicy-based security policies, deprecated since repoze.bfg 0.9, have been removed. If you have something like this in your configure.zcml, it will no longer work:


    If ZCML like the above exists in your application, you will receive an error at startup time. Instead of the above, you'll need something like:


    This is just an example. See the "Security" chapter of the repoze.bfg documentation for more information about configuring security policies.

  • Custom ZCML directives which register an authentication or authorization policy (ala "authtktauthenticationpolicy" or "aclauthorizationpolicy") should register the policy "eagerly" in the ZCML directive instead of from within a ZCML action. If an authentication or authorization policy is not found in the component registry by the view machinery during deferred ZCML processing, view security will not work as expected.

1.0.1 (2009-07-22)

  • Added support for has_resource, resource_isdir, and resource_listdir to the resource "OverrideProvider"; this fixes a bug with a symptom that a file could not be overridden in a resource directory unless a file with the same name existed in the original directory being overridden.

  • Fixed documentation bug showing invalid test for values from the matchdict: they are stored as attributes of the Article, rather than subitems.

  • Fixed documentation bug showing wrong environment key for the matchdict produced by the matching route.

  • Added a workaround for a bug in Python 2.6, 2.6.1, and 2.6.2 having to do with a recursion error in the mimetypes module when trying to serve static files from Paste's FileApp: Symptom: File "/usr/lib/python2.6/", line 244, in guess_type return guess_type(url, strict) RuntimeError: maximum recursion depth exceeded. Thanks to Armin Ronacher for identifying the symptom and pointing out a fix.

  • Minor edits to tutorials for accuracy based on feedback.

  • Declared Paste and PasteDeploy dependencies.

1.0 (2009-07-05)

  • Retested and added some content to GAE tutorial.

  • Edited "Extending" narrative docs chapter.

  • Added "Deleting the Database" section to the "Defining Models" chapter of the traversal wiki tutorial.

  • Spell checking of narratives and tutorials.

1.0b2 (2009-07-03)

  • remoteuserauthenticationpolicy ZCML directive didn't work without an environ_key directive (didn't match docs).

  • Fix configure_zcml filespec check on Windows. Previously if an absolute filesystem path including a drive letter was passed as filename (or as configure_zcml in the options dict) to repoze.bfg.router.make_app, it would be treated as a package:resource_name specification.

  • Fix inaccuracies and import errors in bfgwiki (traversal+ZODB) and bfgwiki2 (urldispatch+SA) tutorials.

  • Use bfgsite index for all tutorial setup.cfg files.

  • Full documentation grammar/style/spelling audit.

1.0b1 (2009-07-02)


  • Allow a Paste config file (configure_zcml) value or an environment variable (BFG_CONFIGURE_ZCML) to name a ZCML file (optionally package-relative) that will be used to bootstrap the application. Previously, the integrator could not influence which ZCML file was used to do the boostrapping (only the original application developer could do so).


  • Added a "Resources" chapter to the narrative documentation which explains how to override resources within one package from another package.

  • Added an "Extending" chapter to the narrative documentation which explains how to extend or modify an existing BFG application using another Python package and ZCML.

1.0a9 (2009-07-01)


  • Make it possible to pass strings in the form "package_name:relative/path" to APIs like render_template, render_template_to_response, and get_template. Sometimes the package in which a caller lives is a direct namespace package, so the module which is returned is semi-useless for navigating from. In this way, the caller can control the horizontal and vertical of where things get looked up from.

1.0a8 (2009-07-01)


  • Deprecate the authentication_policy and authorization_policy arguments to repoze.bfg.router.make_app. Instead, developers should use the various authentication policy ZCML directives (repozewho1authenticationpolicy, remoteuserauthenticationpolicy and authtktauthenticationpolicy) and the aclauthorizationpolicy` authorization policy directive as described in the changes to the "Security" narrative documentation chapter and the wiki tutorials.


  • Add three new ZCML directives which configure authentication policies:

    • repozewho1authenticationpolicy

    • remoteuserauthenticationpolicy

    • authtktauthenticationpolicy

  • Add a new ZCML directive which configures an ACL authorization policy named aclauthorizationpolicy.

Bug Fixes

  • Bug fix: when a repoze.bfg.resource.PackageOverrides class was instantiated, and the package it was overriding already had a __loader__ attribute, it would fail at startup time, even if the __loader__ attribute was another PackageOverrides instance. We now replace any __loader__ that is also a PackageOverrides instance. Symptom: ConfigurationExecutionError: <type 'exceptions.TypeError'>: Package <module 'karl.views' from '/Users/chrism/projects/osi/bfgenv/src/karl/karl/views/__init__.pyc'> already has a __loader__ (probably a module in a zipped egg).

1.0a7 (2009-06-30)


  • Add a reload_resources configuration file setting (aka the BFG_RELOAD_RESOURCES environment variable). When this is set to true, the server never needs to be restarted when moving files between directory resource overrides (esp. for templates currently).

  • Add a reload_all configuration file setting (aka the BFG_RELOAD_ALL environment variable) that implies both reload_resources and reload_templates.

  • The static helper view class now uses a PackageURLParser in order to allow for the overriding of static resources (CSS / logo files, etc) using the resource ZCML directive. The PackageURLParser class was added to a (new) static module in BFG; it is a subclass of the StaticURLParser class in paste.urlparser.

  • The repoze.bfg.templating.renderer_from_cache function now checks for the reload_resources setting; if it's true, it does not register a template renderer (it won't use the registry as a template renderer cache).


  • Add pkg_resources to the glossary.

  • Update the "Environment" docs to note the existence of reload_resources and reload_all.

  • Updated the bfg_alchemy paster template to include two views: the view on the root shows a list of links to records; the view on a record shows the details for that object.


  • Use a colon instead of a tab as the separator between package name and relpath to form the "spec" when register a ITemplateRenderer.

  • Register a repoze.bfg.resource.OverrideProvider as a pkg_resources provider only for modules which are known to have overrides, instead of globally, when a <resource> directive is used (performance).

1.0a6 (2009-06-29)

Bug Fixes

  • Use caller_package function instead of caller_module function within templating to avoid needing to name the caller module in resource overrides (actually match docs).

  • Make it possible to override templates stored directly in a module with templates in a subdirectory of the same module, stored directly within another module, or stored in a subdirectory of another module (actually match docs).

1.0a5 (2009-06-28)


  • A new ZCML directive exists named "resource". This ZCML directive allows you to override Chameleon templates within a package (both directories full of templates and individual template files) with other templates in the same package or within another package. This allows you to "fake out" a view's use of a template, causing it to retrieve a different template than the one actually named by a relative path to a call like render_template_to_response('templates/'). For example, you can override a template file by doing:


    The string passed to "to_override" and "override_with" is named a "specification". The colon separator in a specification separates the package name from a package-relative directory name. The colon and the following relative path are optional. If they are not specified, the override attempts to resolve every lookup into a package from the directory of another package. For example:


    Individual subdirectories within a package can also be overridden:


    If you wish to override a directory with another directory, you must make sure to attach the slash to the end of both the to_override specification and the override_with specification. If you fail to attach a slash to the end of a specification that points a directory, you will get unexpected results. You cannot override a directory specification with a file specification, and vice versa (a startup error will occur if you try).

    You cannot override a resource with itself (a startup error will occur if you try).

    Only individual package resources may be overridden. Overrides will not traverse through subpackages within an overridden package. This means that if you want to override resources for both some.package:templates, and some.package.views:templates, you will need to register two overrides.

    The package name in a specification may start with a dot, meaning that the package is relative to the package in which the ZCML file resides. For example:


    Overrides for the same to_overrides specification can be named multiple times within ZCML. Each override_with path will be consulted in the order defined within ZCML, forming an override search path.

    Resource overrides can actually override resources other than templates. Any software which uses the pkg_resources get_resource_filename, get_resource_stream or get_resource_string APIs will obtain an overridden file when an override is used. However, the only built-in facility which uses the pkg_resources API within BFG is the templating stuff, so we only call out template overrides here.

  • Use the pkg_resources API to locate template filenames instead of dead-reckoning using the os.path module.

  • The repoze.bfg.templating module now uses pkg_resources to locate and register template files instead of using an absolute path name.

1.0a4 (2009-06-25)


  • Cause :segment matches in route paths to put a Unicode-decoded and URL-dequoted value in the matchdict for the value matched. Previously a non-decoded non-URL-dequoted string was placed in the matchdict as the value.

  • Cause *remainder matches in route paths to put a tuple in the matchdict dictionary in order to be able to present Unicode-decoded and URL-dequoted values for the traversal path. Previously a non-decoded non-URL-dequoted string was placed in the matchdict as the value.

  • Add optional max_age keyword value to the remember method of repoze.bfg.authentication.AuthTktAuthenticationPolicy; if this value is passed to remember, the generated cookie will have a corresponding Max-Age value.


  • Add information to the URL Dispatch narrative documentation about path pattern matching syntax.

Bug Fixes

  • Make route_url URL-quote segment replacements during generation. Remainder segments are not quoted.

1.0a3 (2009-06-24)

Implementation Changes

  • repoze.bfg no longer relies on the Routes package to interpret URL paths. All known existing path patterns will continue to work with the reimplemented logic, which lives in repoze.bfg.urldispatch. <route> ZCML directives which use certain attributes (uncommon ones) may not work (see "Backwards Incompatibilities" below).

Bug Fixes

  • model_url when passed a request that was generated as a result of a route match would fail in a call to route.generate.

  • BFG-on-GAE didn't work due to a corner case bug in the fallback Python implementation of threading.local (symptom: "Initialization arguments are not supported"). Thanks to Michael Bernstein for the bug report.


  • Added a "corner case" explanation to the "Hybrid Apps" chapter explaining what to do when "the wrong" view is matched.

  • Use repoze.bfg.url.route_url API in tutorials rather than Routes url_for API.


  • Added the repoze.bfg.url.route_url API. This API allows you to generate URLs based on <route> declarations. See the URL Dispatch narrative chapter and the "repoze.bfg.url" module API documentation for more information.

Backwards Incompatibilities

  • As a result of disusing Routes, using the Routes url_for API inside a BFG application (as was suggested by previous iterations of tutorials) will no longer work. Use the repoze.bfg.url.route_url method instead.

  • The following attributes on the <route> ZCML directive no longer work: encoding, static, filter, condition_method, condition_subdomain, condition_function, explicit, or subdomains. These were all Routes features.

  • The <route> ZCML directive no longer supports the <requirement> subdirective. This was a Routes feature.

1.0a2 (2009-06-23)

Bug Fixes

  • The bfg_routesalchemy paster template app tests failed due to a mismatch between test and view signatures.


  • Add a view_for attribute to the route ZCML directive. This attribute should refer to an interface or a class (ala the for attribute of the view ZCML directive).


  • Conditional documentation in installation section ("how to install a Python interpreter").

Backwards Incompatibilities

  • The callback argument of the repoze.bfg.authentication authentication policies named RepozeWho1AuthenticationPolicy, RemoteUserAuthenticationPolicy, and AuthTktAuthenticationPolicy now must accept two positional arguments: the original argument accepted by each (userid or identity) plus a second argument, which will be the current request. Apologies, this is required to service finding groups when there is no "global" database connection.

1.0a1 (2009-06-22)


  • A new ZCML directive was added named notfound. This ZCML directive can be used to name a view that should be invoked when the request can't otherwise be resolved to a view callable. For example:

  • A new ZCML directive was added named forbidden. This ZCML directive can be used to name a view that should be invoked when a view callable for a request is found, but cannot be invoked due to an authorization failure. For example:

  • Allow views to be optionally defined as callables that accept only a request object, instead of both a context and a request (which still works, and always will). The following types work as views in this style:

    • functions that accept a single argument request, e.g.:

      def aview(request):
    • new and old-style classes that have an __init__ method that accepts self, request, e.g.:

      def View(object):
          __init__(self, request):
    • Arbitrary callables that have a __call__ method that accepts self, request, e.g.:

      def AView(object):
          def __call__(self, request):
      view = AView()

    This likely should have been the calling convention all along, as the request has context as an attribute already, and with views called as a result of URL dispatch, having the context in the arguments is not very useful. C'est la vie.

  • Cache the absolute path in the caller's package globals within repoze.bfg.path to get rid of repeated (expensive) calls to os.path.abspath.

  • Add reissue_time and timeout parameters to repoze.bfg.authentication.AuthTktAuthenticationPolicy constructor. If these are passed, cookies will be reset every so often (cadged from the same change to repoze.who lately).

  • The matchdict related to the matching of a Routes route is available on the request as the matchdict attribute: request.matchdict. If no route matched, this attribute will be None.

  • Make 404 responses slightly cheaper by showing environ["PATH_INFO"] on the notfound result page rather than the fullly computed URL.

  • Move LRU cache implementation into a separate package (repoze.lru).

  • The concepts of traversal and URL dispatch have been unified. It is now possible to use the same sort of factory as both a traversal "root factory" and what used to be referred to as a urldispatch "context factory".

  • When the root factory argument (as a first argument) passed to repoze.bfg.router.make_app is None, a default root factory is used. This is in support of using routes as "root finders"; it supplants the idea that there is a default IRoutesContextFactory.

  • The view` ZCML statement and the repoze.bfg.view.bfg_view decorator now accept an extra argument: route_name. If a route_name is specified, it must match the name of a previously defined route statement. When it is specified, the view will only be called when that route matches during a request.

  • It is now possible to perform traversal after a route has matched. Use the pattern *traverse in a <route> path attribute within ZCML, and the path remainder which it matches will be used as a traversal path.

  • When any route defined matches, the WSGI environment will now contain a key bfg.routes.route (the Route object which matched), and a key bfg.routes.matchdict (the result of calling route.match).


  • Utility registrations against repoze.bfg.interfaces.INotFoundView and repoze.bfg.interfaces.IForbiddenView are now deprecated. Use the notfound and forbidden ZCML directives instead (see the "Hooks" chapter for more information). Such registrations will continue to work, but the notfound and forbidden directives do "extra work" to ensure that the callable named by the directive can be called by the router even if it's a class or request-argument-only view.


  • The IRoutesContext, IRoutesContextFactory, and IContextNotFound interfaces were removed from repoze.bfg.interfaces. These were never APIs.

  • The repoze.bfg.urldispatch.RoutesContextNotFound, repoze.bfg.urldispatch.RoutesModelTraverser and repoze.bfg.urldispatch.RoutesContextURL classes were removed. These were also never APIs.

Backwards Incompatibilities

  • Moved the repoze.bfg.push module, which implemented the pushpage decorator, into a separate distribution, repoze.bfg.pushpage. Applications which used this decorator should continue to work after adding that distribution to their installation requirements.

  • Changing the default request factory via an IRequestFactory utility registration (as used to be documented in the "Hooks" chapter's "Changing the request factory" section) is no longer supported. The dance to manufacture a request is complicated as a result of unifying traversal and url dispatch, making it highly unlikely for anyone to be able to override it properly. For those who just want to decorate or modify a request, use a NewRequestEvent subscriber (see the Events chapter in the documentation).

  • The repoze.bfg.IRequestFactory interface was removed. See the bullet above for why.

  • Routes "context factories" (spelled as the factory argument to a route statement in ZCML) must now expect the WSGI environ as a single argument rather than a set of keyword arguments. They can obtain the match dictionary by asking for environ['bfg.routes.matchdict']. This is the same set of keywords that used to be passed to urldispatch "context factories" in BFG 0.9 and below.

  • Using the @zope.component.adapter decorator on a bfg view function no longer works. Use the @repoze.bfg.view.bfg_view decorator instead to mark a function (or a class) as a view.

  • The name under which the matching route object is found in the environ was changed from bfg.route to bfg.routes.route.

  • Finding the root is now done before manufacturing a request object (and sending a new request event) within the router (it used to be performed afterwards).

  • Adding *path_info to a route no longer changes the PATH_INFO for a request that matches using URL dispatch. This feature was only there to service the repoze.bfg.wsgi.wsgiapp2 decorator and it did it wrong; use *subpath instead now.

  • The values of subpath, traversed, and virtual_root_path attached to the request object are always now tuples instead of lists (performance).

Bug Fixes

  • The bfg_alchemy Paster template named "" in its pipeline rather than "repoze.tm2", causing the startup to fail.

  • Move BBB logic for registering an IAuthenticationPolicy/IForbiddenView/INotFoundView based on older concepts from the router module's make_app function into the repoze.bfg.zcml.zcml_configure callable, to service compatibility with scripts that use "zope.configuration.xmlconfig" (replace with repoze.bfg.zml.zcml_configure as necessary to get BBB logic)


  • Add interface docs related to how to create authentication policies and authorization policies to the "Security" narrative chapter.

  • Added a (fairly sad) "Combining Traversal and URL Dispatch" chapter to the narrative documentation. This explains the usage of *traverse and *subpath in routes URL patters.

  • A "router" chapter explaining the request/response lifecycle at a high level was added.

  • Replaced all mentions and explanations of a routes "context factory" with equivalent explanations of a "root factory" (context factories have been disused).

  • Updated Routes bfgwiki2 tutorial to reflect the fact that context factories are now no longer used.

0.9.1 (2009-06-02)


  • Add API named repoze.bfg.settings.get_settings which retrieves a derivation of values passed as the options value of repoze.bfg.router.make_app. This API should be preferred instead of using getUtility(ISettings). I added a new repoze.bfg.settings API document as well.

Bug Fixes

  • Restored missing entry point declaration for bfg_alchemy paster template, which was accidentally removed in 0.9.


  • Fix a reference to wsgiapp in the wsgiapp2 API documentation within the repoze.bfg.wsgi module.

API Removals

  • The repoze.bfg.location.locate API was removed: it didn't do enough to be very helpful and had a misleading name.

0.9 (2009-06-01)

Bug Fixes

  • It was not possible to register a custom IRoutesContextFactory for use as a default context factory as documented in the "Hooks" chapter.


  • The request_type argument of ZCML view declarations and bfg_view decorators can now be one of the strings GET, POST, PUT, DELETE, or HEAD instead of a reference to the respective interface type imported from repoze.bfg.interfaces.

  • The route ZCML directive now accepts request_type as an alias for its condition_method argument for symmetry with the view directive.

  • The bfg_routesalchemy paster template now provides a unit test and actually uses the database during a view rendering.


  • Remove repoze.bfg.threadlocal.setManager. It was only used in unit tests.

  • Remove repoze.bfg.wsgi.HTTPException, repoze.bfg.wsgi.NotFound, and repoze.bfg.wsgi.Unauthorized. These classes were disused with the introduction of the IUnauthorizedView and INotFoundView machinery.


  • Add description to narrative templating chapter about how to use Chameleon text templates.

  • Changed Views narrative chapter to use method strings rather than interface types, and moved advanced interface type usage to Events narrative chapter.

  • Added a Routes+SQLAlchemy wiki tutorial.

0.9a8 (2009-05-31)


  • It is now possible to register a custom repoze.bfg.interfaces.INotFoundView for a given application. This feature replaces the repoze.bfg.interfaces.INotFoundAppFactory feature previously described in the Hooks chapter. The INotFoundView will be called when the framework detects that a view lookup done as a result of a request fails; it should accept a context object and a request object; it should return an IResponse object (a webob response, basically). See the Hooks narrative chapter of the BFG docs for more info.

  • The error presented when a view invoked by the router returns a non-response object now includes the view's name for troubleshooting purposes.

Bug Fixes

  • A "new response" event is emitted for forbidden and notfound views.


  • The repoze.bfg.interfaces.INotFoundAppFactory interface has been deprecated in favor of using the new repoze.bfg.interfaces.INotFoundView mechanism.


  • Renamed repoze.bfg.interfaces.IForbiddenResponseFactory to repoze.bfg.interfaces.IForbiddenView.

0.9a7 (2009-05-30)


  • Remove "context" argument from effective_principals and authenticated_userid function APIs in, effectively a doing reversion to 0.8 and before behavior. Both functions now again accept only the request parameter.

0.9a6 (2009-05-29)


  • Changed "BFG Wiki" tutorial to use AuthTktAuthenticationPolicy rather than repoze.who.


  • Add an AuthTktAuthenticationPolicy. This policy retrieves credentials from an auth_tkt cookie managed by the application itself (instead of relying on an upstream data source for authentication data). See the Security API chapter of the documentation for more info.

  • Allow RemoteUserAuthenticationPolicy and RepozeWho1AuthenticationPolicy to accept various constructor arguments. See the Security API chapter of the documentation for more info.

0.9a5 (2009-05-28)


  • Add a get_app API functions to the paster module. This obtains a WSGI application from a config file given a config file name and a section name. See the repoze.bfg.paster API docs for more information.

  • Add a new module named scripting. It contains a get_root API function, which, provided a Router instance, returns a traversal root object and a "closer". See the repoze.bfg.scripting API docs for more info.

0.9a4 (2009-05-27)

Bug Fixes

  • Try checking for an "old style" security policy after we parse ZCML (thinko).

0.9a3 (2009-05-27)


  • Allow IAuthenticationPolicy and IAuthorizationPolicy to be overridden via ZCML registrations (do ZCML parsing after registering these in


  • Added "BFG Wiki" tutorial to documentation; it describes step-by-step how to create a traversal-based ZODB application with authentication.


  • Added deprecations for imports of ACLSecurityPolicy, InheritingACLSecurityPolicy, RemoteUserACLSecurityPolicy, RemoteUserInheritingACLSecurityPolicy, WhoACLSecurityPolicy, and WhoInheritingACLSecurityPolicy from the module; for the meantime (for backwards compatibility purposes) these live in the repoze.bfg.secpols module. Note however, that the entire concept of a "security policy" is deprecated in BFG in favor of separate authentication and authorization policies, so any use of a security policy will generate additional deprecation warnings even if you do start using repoze.bfg.secpols. repoze.bfg.secpols will disappear in a future release of repoze.bfg.

Deprecated Import Alias Removals

  • Remove repoze.bfg.template module. All imports from this package have been deprecated since 0.3.8. Instead, import get_template, render_template, and render_template_to_response from the repoze.bfg.chameleon_zpt module.

  • Remove backwards compatibility import alias for repoze.bfg.traversal.split_path (deprecated since 0.6.5). This must now be imported as repoze.bfg.traversal.traversal_path).

  • Remove backwards compatibility import alias for repoze.bfg.urldispatch.RoutesContext (deprecated since 0.6.5). This must now be imported as repoze.bfg.urldispatch.DefaultRoutesContext.

  • Removed backwards compatibility import aliases for repoze.bfg.router.get_options and repoze.bfg.router.Settings (deprecated since 0.6.2). These both must now be imported from repoze.bfg.settings.

  • Removed backwards compatibility import alias for repoze.bfg.interfaces.IRootPolicy (deprecated since 0.6.2). It must be imported as repoze.bfg.interfaces.IRootFactory now.

  • Removed backwards compatibility import alias for repoze.bfg.interfaces.ITemplate (deprecated since 0.4.4). It must be imported as repoze.bfg.interfaces.ITemplateRenderer now.

  • Removed backwards compatibility import alias for repoze.bfg.interfaces.ITemplateFactory (deprecated since 0.4.4). It must be imported as repoze.bfg.interfaces.ITemplateRendererFactory now.

  • Removed backwards compatibility import alias for repoze.bfg.chameleon_zpt.ZPTTemplateFactory (deprecated since 0.4.4). This must be imported as repoze.bfg.ZPTTemplateRenderer now.

0.9a2 (2009-05-27)


  • A paster command has been added named "bfgshell". This command can be used to get an interactive prompt with your BFG root object in the global namespace. E.g.:

    bin/paster bfgshell /path/to/myapp.ini myapp

    See the Project chapter in the BFG documentation for more information.


  • The name repoze.bfg.registry.registry_manager was never an API, but scripts in the wild were using it to set up an environment for use under a debug shell. A backwards compatibility shim has been added for this purpose, but the feature is deprecated.

0.9a1 (2009-5-27)


  • New API functions named forget and remember are available in the security module. The forget function returns headers which will cause the currently authenticated user to be logged out when set in a response. The remember function (when passed the proper arguments) will return headers which will cause a principal to be "logged in" when set in a response. See the Security API chapter of the docs for more info.

  • New keyword arguments to the repoze.bfg.router.make_app call have been added: authentication_policy and authorization_policy. These should, respectively, be an implementation of an authentication policy (an object implementing the repoze.bfg.interfaces.IAuthenticationPolicy interface) and an implementation of an authorization policy (an object implementing repoze.bfg.interfaces.IAuthorizationPolicy). Concrete implementations of authentication policies exist in repoze.bfg.authentication. Concrete implementations of authorization policies exist in repoze.bfg.authorization.

    Both authentication_policy and authorization_policy default to None.

    If authentication_policy is None, but authorization_policy is not None, then authorization_policy is ignored (the ability to do authorization depends on authentication).

    If the authentication_policy argument is not None, and the authorization_policy argument is None, the authorization policy defaults to an authorization implementation that uses ACLs (repoze.bfg.authorization.ACLAuthorizationPolicy).

    We no longer encourage configuration of "security policies" using ZCML, as previously we did for ISecurityPolicy. This is because it's not uncommon to need to configure settings for concrete authorization or authentication policies using paste .ini parameters; the app entry point for your application is the natural place to do this.

  • Two new abstractions have been added in the way of adapters used by the system: an IAuthorizationPolicy and an IAuthenticationPolicy. A combination of these (as registered by the securitypolicy ZCML directive) take the place of the ISecurityPolicy abstraction in previous releases of repoze.who. The API functions in (such as authentication_userid, effective_principals, has_permission, and so on) have been changed to try to make use of these new adapters. If you're using an older ISecurityPolicy adapter, the system will still work, but it will print deprecation warnings when such a policy is used.

  • The way the (internal) IViewPermission utilities registered via ZCML are invoked has changed. They are purely adapters now, returning a boolean result, rather than returning a callable. You shouldn't have been using these anyway. ;-)

  • New concrete implementations of IAuthenticationPolicy have been added to the repoze.bfg.authentication module: RepozeWho1AuthenticationPolicy which uses repoze.who identity to retrieve authentication data from and RemoteUserAuthenticationPolicy, which uses the REMOTE_USER value in the WSGI environment to retrieve authentication data.

  • A new concrete implementation of IAuthorizationPolicy has been added to the repoze.bfg.authorization module: ACLAuthorizationPolicy which uses ACL inheritance to do authorization.

  • It is now possible to register a custom repoze.bfg.interfaces.IForbiddenResponseFactory for a given application. This feature replaces the repoze.bfg.interfaces.IUnauthorizedAppFactory feature previously described in the Hooks chapter. The IForbiddenResponseFactory will be called when the framework detects an authorization failure; it should accept a context object and a request object; it should return an IResponse object (a webob response, basically). Read the below point for more info and see the Hooks narrative chapter of the BFG docs for more info.

Backwards Incompatibilities

  • Custom NotFound and Forbidden (nee' Unauthorized) WSGI applications (registered as a utility for INotFoundAppFactory and IUnauthorizedAppFactory) could rely on an environment key named message describing the circumstance of the response. This key has been renamed to repoze.bfg.message (as per the WSGI spec, which requires environment extensions to contain dots).


  • The repoze.bfg.interfaces.IUnauthorizedAppFactory interface has been deprecated in favor of using the new repoze.bfg.interfaces.IForbiddenResponseFactory mechanism.

  • The view_execution_permitted API should now be imported from the module instead of the repoze.bfg.view module.

  • The authenticated_userid and effective_principals APIs in used to only take a single argument (request). They now accept two arguments (context and request). Calling them with a single argument is still supported but issues a deprecation warning. (NOTE: this change was reverted in 0.9a7; meaning the 0.9 versions of these functions again accept request only, just like 0.8 and before).

  • Use of "old-style" security policies (those base on ISecurityPolicy) is now deprecated. See the "Security" chapter of the docs for info about activating an authorization policy and an authentication poicy.

0.8.1 (2009-05-21)


  • Class objects may now be used as view callables (both via ZCML and via use of the bfg_view decorator in Python 2.6 as a class decorator). The calling semantics when using a class as a view callable is similar to that of using a class as a Zope "browser view": the class' __init__ must accept two positional parameters (conventionally named context, and request). The resulting instance must be callable (it must have a __call__ method). When called, the instance should return a response. For example:

     from webob import Response
     class MyView(object):
         def __init__(self, context, request):
             self.context = context
             self.request = request
         def __call__(self):
             return Response('hello from %s!' % self.context)
    See the "Views" chapter in the documentation and the
    ``repoze.bfg.view`` API documentation for more information.
  • Removed the pickling of ZCML actions (the code that wrote configure.zcml.cache next to configure.zcml files in projects). The code which managed writing and reading of the cache file was a source of subtle bugs when users switched between imperative (e.g. @bfg_view) registrations and declarative registrations (e.g. the view directive in ZCML) on the same project. On a moderately-sized project (535 ZCML actions and 15 ZCML files), executing actions read from the pickle was saving us only about 200ms (2.5 sec vs 2.7 sec average). On very small projects (1 ZCML file and 4 actions), startup time was comparable, and sometimes even slower when reading from the pickle, and both ways were so fast that it really just didn't matter anyway.

0.8 (2009-05-18)


  • Added a traverse function to the repoze.bfg.traversal module. This function may be used to retrieve certain values computed during path resolution. See the Traversal API chapter of the documentation for more information about this function.


  • Internal: ITraverser callables should now return a dictionary rather than a tuple. Up until 0.7.0, all ITraversers were assumed to return a 3-tuple. In 0.7.1, ITraversers were assumed to return a 6-tuple. As (by evidence) it's likely we'll need to add further information to the return value of an ITraverser callable, 0.8 assumes that an ITraverser return a dictionary with certain elements in it. See the repoze.bfg.interfaces.ITraverser interface for the list of keys that should be present in the dictionary. ITraversers which return tuples will still work, although a deprecation warning will be issued.

Backwards Incompatibilities

  • If your code used the ITraverser interface directly (not via an API function such as find_model) via an adapter lookup, you'll need to change your code to expect a dictionary rather than a 3- or 6-tuple if your code ever gets return values from the default ModelGraphTraverser or RoutesModelTraverser adapters.

0.8a7 (2009-05-16)

Backwards Incompatibilities

  • The RoutesMapper class in repoze.bfg.urldispatch has been removed, as well as its documentation. It had been deprecated since 0.6.3. Code in repoze.bfg.urldispatch.RoutesModelTraverser which catered to it has also been removed.

  • The semantics of the route ZCML directive have been simplified. Previously, it was assumed that to use a route, you wanted to map a route to an externally registered view. The new route directive instead has a view attribute which is required, specifying the dotted path to a view callable. When a route directive is processed, a view is registered using the name attribute of the route directive as its name and the callable as its value. The view_name and provides attributes of the route directive are therefore no longer used. Effectively, if you were previously using the route directive, it means you must change a pair of ZCML directives that look like this:


    To a ZCML directive that looks like this:


    In other words, to make old code work, remove the view directives that were only there to serve the purpose of backing route directives, and move their view= attribute into the route directive itself.

    This change also necessitated that the name attribute of the route directive is now required. If you were previously using route directives without a name attribute, you'll need to add one (the name is arbitrary, but must be unique among all route and view statements).

    The provides attribute of the route directive has also been removed. This directive specified a sequence of interface types that the generated context would be decorated with. Since route views are always generated now for a single interface (repoze.bfg.IRoutesContext) as opposed to being looked up arbitrarily, there is no need to decorate any context to ensure a view is found.


  • Added API docs for the repoze.bfg.testing methods registerAdapter, registerUtiity, registerSubscriber, and cleanUp.

  • Added glossary entry for "root factory".

  • Noted existence of repoze.bfg.pagetemplate template bindings in "Available Add On Template System Bindings" in Templates chapter in narrative docs.

  • Update "Templates" narrative chapter in docs (expand to show a sample template and correct macro example).


  • Courtesty Carlos de la Guardia, added an alchemy Paster template. This paster template sets up a BFG project that uses SQAlchemy (with SQLite) and uses traversal to resolve URLs. (no Routes areused). This template can be used via paster create -t bfg_alchemy.

  • The Routes Route object used to resolve the match is now put into the environment as bfg.route when URL dispatch is used.

  • You can now change the default Routes "context factory" globally. See the "ZCML Hooks" chapter of the documentation (in the "Changing the Default Routes Context Factory" section).

0.8a6 (2009-05-11)


  • Added a routesalchemy Paster template. This paster template sets up a BFG project that uses SQAlchemy (with SQLite) and uses Routes exclusively to resolve URLs (no traversal root factory is used). This template can be used via paster create -t bfg_routesalchemy.


  • Added documentation to the URL Dispatch chapter about how to catch the root URL using a ZCML route directive.

  • Added documentation to the URL Dispatch chapter about how to perform a cleanup function at the end of a request (e.g. close the SQL connection).

Bug Fixes

  • In version 0.6.3, passing a get_root callback (a "root factory") to repoze.bfg.router.make_app became optional if any route declaration was made in ZCML. The intent was to make it possible to disuse traversal entirely, instead relying entirely on URL dispatch (Routes) to resolve all contexts. However a compound set of bugs prevented usage of a Routes-based root view (a view which responds to "/"). One bug existed in repoze.bfg.urldispatch`, another existed in Routes itself.

    To resolve this issue, the urldispatch module was fixed, and a fork of the Routes trunk was put into the "dev" index named Routes-1.11dev-chrism-home. The source for the fork exists at (broken link); its contents have been merged into the Routes trunk (what will be Routes 1.11).

0.8a5 (2009-05-08)


  • Two new security policies were added: RemoteUserInheritingACLSecurityPolicy and WhoInheritingACLSecurityPolicy. These are security policies which take into account all ACLs defined in the lineage of a context rather than stopping at the first ACL found in a lineage. See the "Security" chapter of the API documentation for more information.

  • The API and narrative documentation dealing with security was changed to introduce the new "inheriting" security policy variants.

  • Added glossary entry for "lineage".


  • The security policy previously named RepozeWhoIdentityACLSecurityPolicy now has the slightly saner name of WhoACLSecurityPolicy. A deprecation warning is emitted when this policy is imported under the "old" name; usually this is due to its use in ZCML within your application. If you're getting this deprecation warning, change your ZCML to use the new name, e.g. change:




0.8a4 (2009-05-04)


  • zope.testing is no longer a direct dependency, although our dependencies (such as zope.interface, repoze.zcml, etc) still depend on it.

  • Tested on Google App Engine. Added a tutorial to the documentation explaining how to deploy a BFG app to GAE.

Backwards Incompatibilities

  • Applications which rely on zope.testing.cleanup.cleanUp in unit tests can still use that function indefinitely. However, for maximum forward compatibility, they should import cleanUp from repoze.bfg.testing instead of from zope.testing.cleanup. The BFG paster templates and docs have been changed to use this function instead of the zope.testing.cleanup version.

0.8a3 (2009-05-03)


  • Don't require a successful import of zope.testing at BFG application runtime. This allows us to get rid of zope.testing on platforms like GAE which have file limits.

0.8a2 (2009-05-02)


  • We no longer include the configure.zcml of the chameleon.zpt package within the configure.zcml of the "repoze.bfg.includes" package. This has been a no-op for some time now.

  • The repoze.bfg.chameleon_zpt package no longer imports from chameleon.zpt at module scope, deferring the import until later within a method call. The chameleon.zpt package can't be imported on platforms like GAE.

0.8a1 (2009-05-02)

Deprecation Warning and Import Alias Removals

  • Since version 0.6.1, a deprecation warning has been emitted when the name model_url is imported from the repoze.bfg.traversal module. This import alias (and the deprecation warning) has been removed. Any import of the model_url function will now need to be done from repoze.bfg.url; any import of the name model_url from repoze.bfg.traversal will now fail. This was done to remove a dependency on zope.deferredimport.

  • Since version 0.6.5, a deprecation warning has been emitted when the name RoutesModelTraverser is imported from the repoze.bfg.traversal module. This import alias (and the deprecation warning) has been removed. Any import of the RoutesModelTraverser class will now need to be done from repoze.bfg.urldispatch; any import of the name RoutesModelTraverser from repoze.bfg.traversal will now fail. This was done to remove a dependency on zope.deferredimport.


  • This release of repoze.bfg is "C-free". This means it has no hard dependencies on any software that must be compiled from C source at installation time. In particular, repoze.bfg no longer depends on the lxml package.

    This change has introduced some backwards incompatibilities, described in the "Backwards Incompatibilities" section below.

  • This release was tested on Windows XP. It appears to work fine and all the tests pass.

Backwards Incompatibilities

Incompatibilities related to making repoze.bfg "C-free":

  • Removed the repoze.bfg.chameleon_genshi module, and thus support for Genshi-style chameleon templates. Genshi-style Chameleon templates depend upon lxml, which is implemented in C (as opposed to pure Python) and the repoze.bfg core is "C-free" as of this release. You may get Genshi-style Chameleon support back by installing the repoze.bfg.chameleon_genshi package availalable from (also available in the index at All existing code that depended on the chameleon_genshi module prior to this release of repoze.bfg should work without change after this addon is installed.

  • Removed the repoze.bfg.xslt module and thus support for XSL templates. The repoze.bfg.xslt module depended upon lxml, which is implemented in C, and the repoze.bfg core is "C-free" as of this release. You bay get XSL templating back by installing the repoze.bfg.xslt package available from (also available in the index at All existing code that depended upon the xslt module prior to this release of repoze.bfg should work without modification after this addon is installed.

  • Removed the repoze.bfg.interfaces.INodeTemplateRenderer interface and the an old b/w compat aliases from that interface to repoze.bfg.interfaces.INodeTemplate. This interface must now be imported from the repoze.bfg.xslt.interfaces package after installation of the repoze.bfg.xslt addon package described above as repoze.bfg.interfaces.INodeTemplateRenderer. This interface was never part of any public API.

Other backwards incompatibilities:

  • The render_template function in repoze.bfg.chameleon_zpt returns Unicode instead of a string. Likewise, the individual values returned by the iterable created by the render_template_to_iterable function are also each Unicode. This is actually a backwards incompatibility inherited from our new use of the combination of chameleon.core 1.0b32 (the non-lxml-depending version) and chameleon.zpt 1.0b16+ ; the chameleon.zpt PageTemplateFile implementation used to return a string, but now returns Unicode.

0.7.1 (2009-05-01)


  • The "paster create" templates have been modified to use links to the new "" and "" websites.

  • Added better documentation for virtual hosting at a URL prefix within the virtual hosting docs chapter.

  • The interface for repoze.bfg.interfaces.ITraverser and the built-in implementations that implement the interface (repoze.bfg.traversal.ModelGraphTraverser, and repoze.bfg.urldispatch.RoutesModelTraverser) now expect the __call__ method of an ITraverser to return 3 additional arguments: traversed, virtual_root, and virtual_root_path (the old contract was that the __call__ method of an ITraverser returned; three arguments, the contract new is that it returns six). traversed will be a sequence of Unicode names that were traversed (including the virtual root path, if any) or None if no traversal was performed, virtual_root will be a model object representing the virtual root (or the physical root if traversal was not performed), and virtual_root_path will be a sequence representing the virtual root path (a sequence of Unicode names) or None if traversal was not performed.

    Six arguments are now returned from BFG ITraversers. They are returned in this order: context, view_name, subpath, traversed, virtual_root, and virtual_root_path.

    Places in the BFG code which called an ITraverser continue to accept a 3-argument return value, although BFG will generate and log a warning when one is encountered.

  • The request object now has the following attributes: traversed (the sequence of names traversed or None if traversal was not performed), virtual_root (the model object representing the virtual root, including the virtual root path if any), and virtual_root_path (the seuquence of names representing the virtual root path or None if traversal was not performed).

  • A new decorator named wsgiapp2 was added to the repoze.bfg.wsgi module. This decorator performs the same function as repoze.bfg.wsgi.wsgiapp except it fixes up the SCRIPT_NAME, and PATH_INFO environment values before invoking the WSGI subapplication.

  • The repoze.bfg.testing.DummyRequest object now has default attributes for traversed, virtual_root, and virtual_root_path.

  • The RoutesModelTraverser now behaves more like the Routes "RoutesMiddleware" object when an element in the match dict is named path_info (usually when there's a pattern like http://foo/*path_info). When this is the case, the PATH_INFO environment variable is set to the value in the match dict, and the SCRIPT_NAME is appended to with the prefix of the original PATH_INFO not including the value of the new variable.

  • The notfound debug now shows the traversed path, the virtual root, and the virtual root path too.

  • Speed up / clarify 'traversal' module's 'model_path', 'model_path_tuple', and '_model_path_list' functions.

Backwards Incompatibilities

  • In previous releases, the repoze.bfg.url.model_url, repoze.bfg.traversal.model_path and repoze.bfg.traversal.model_path_tuple functions always ignored the __name__ argument of the root object in a model graph ( effectively replacing it with a leading / in the returned value) when a path or URL was generated. The code required to perform this operation was not efficient. As of this release, the root object in a model graph must have a __name__ attribute that is either None or the empty string ('') for URLs and paths to be generated properly from these APIs. If your root model object has a __name__ argument that is not one of these values, you will need to change your code for URLs and paths to be generated properly. If your model graph has a root node with a string __name__ that is not null, the value of __name__ will be prepended to every path and URL generated.

  • The repoze.bfg.location.LocationProxy class and the repoze.bfg.location.ClassAndInstanceDescr class have both been removed in order to be able to eventually shed a dependency on zope.proxy. Neither of these classes was ever an API.

  • In all previous releases, the repoze.bfg.location.locate function worked like so: if a model did not explicitly provide the repoze.bfg.interfaces.ILocation interface, locate returned a LocationProxy object representing model with its __parent__ attribute assigned to parent and a __name__ attribute assigned to __name__. In this release, the repoze.bfg.location.locate function simply jams the __name__ and __parent__ attributes on to the supplied model unconditionally, no matter if the object implements ILocation or not, and it never returns a proxy. This was done because the LocationProxy behavior has now moved into an add-on package (repoze.bfg.traversalwrapper), in order to eventually be able to shed a dependency on zope.proxy.

  • In all previous releases, by default, if traversal was used (as opposed to URL-dispatch), and the root object supplied the``repoze.bfg.interfaces.ILocation`` interface, but the children returned via its __getitem__ returned an object that did not implement the same interface, repoze.bfg provided some implicit help during traversal. This traversal feature wrapped subobjects from the root (and thereafter) that did not implement ILocation in proxies which automatically provided them with a __name__ and __parent__ attribute based on the name being traversed and the previous object traversed. This feature has now been removed from the base repoze.bfg package for purposes of eventually shedding a dependency on zope.proxy.

    In order to re-enable the wrapper behavior for older applications which cannot be changed, register the "traversalwrapper" ModelGraphTraverser as the traversal policy, rather than the default ModelGraphTraverser. To use this feature, you will need to install the repoze.bfg.traversalwrapper package (an add-on package, available at Then change your application's configure.zcml to include the following stanza:


    factory="repoze.bfg.traversalwrapper.ModelGraphTraverser" provides="repoze.bfg.interfaces.ITraverserFactory" for="*" />

    When this ITraverserFactory is used instead of the default, no object in the graph (even the root object) must supply a __name__ or __parent__ attribute. Even if subobjects returned from the root do implement the ILocation interface, these will still be wrapped in proxies that override the object's "real" __parent__ and __name__ attributes.

    See also changes to the "Models" chapter of the documentation (in the "Location-Aware Model Instances") section.

0.7.0 (2009-04-11)

Bug Fixes

  • Fix a bug in repoze.bfg.wsgi.HTTPException: the content length was returned as an int rather than as a string.

  • Add explicit dependencies on zope.deferredimport, zope.deprecation, and zope.proxy for forward compatibility reasons (zope.component will stop relying on zope.deferredimport soon and although we use it directly, it's only a transitive dependency, and ''zope.deprecation`` and zope.proxy are used directly even though they're only transitive dependencies as well).

  • Using model_url or model_path against a broken model graph (one with models that had a non-root model with a __name__ of None) caused an inscrutable error to be thrown: ( if not _must_quote[cachekey].search(s): TypeError: expected string or buffer). Now URLs and paths generated against graphs that have None names in intermediate nodes will replace the None with the empty string, and, as a result, the error won't be raised. Of course the URL or path will still be bogus.


  • Make it possible to have testing.DummyTemplateRenderer return some nondefault string representation.

  • Added a new anchor keyword argument to model_url. If anchor is present, its string representation will be used as a named anchor in the generated URL (e.g. if anchor is passed as foo and the model URL is, the generated URL will be

Backwards Incompatibilities

  • The default request charset encoding is now utf-8. As a result, the request machinery will attempt to decode values from the utf-8 encoding to Unicode automatically when they are obtained via request.params, request.GET, and request.POST. The previous behavior of BFG was to return a bytestring when a value was accessed in this manner. This change will break form handling code in apps that rely on values from those APIs being considered bytestrings. If you are manually decoding values from form submissions in your application, you'll either need to change the code that does that to expect Unicode values from request.params, request.GET and request.POST, or you'll need to explicitly reenable the previous behavior. To reenable the previous behavior, add the following to your application's configure.zcml:

    <subscriber for="repoze.bfg.interfaces.INewRequest"

    See also the documentation in the "Views" chapter of the BFG docs entitled "Using Views to Handle Form Submissions (Unicode and Character Set Issues)".


  • Add a section to the narrative Views chapter entitled "Using Views to Handle Form Submissions (Unicode and Character Set Issues)" explaining implicit decoding of form data values.

0.6.9 (2009-02-16)

Bug Fixes

  • lru cache was unstable under concurrency (big surprise!) when it tried to redelete a key in the cache that had already been deleted. Symptom: line 64 in put:del data[oldkey]:KeyError: '/some/path'. Now we just ignore the key error if we can't delete the key (it has already been deleted).

  • Empty location names in model paths when generating a URL using repoze.bfg.model_url based on a model obtained via traversal are no longer ignored in the generated URL. This means that if a non-root model object has a __name__ of '', the URL will reflect it (e.g. model_url will generate http://foo/bar//baz if an object with the __name__ of '' is a child of bar and the parent of baz). URLs generated with empty path segments are, however, still irresolveable by the model graph traverser on request ingress (the traverser strips empty path segment names).


  • Microspeedups of repoze.bfg.traversal.model_path, repoze.bfg.traversal.model_path_tuple, repoze.bfg.traversal.quote_path_segment, and repoze.bfg.url.urlencode.

  • add zip_safe = false to setup.cfg.


  • Add a note to the repoze.bfg.traversal.quote_path_segment API docs about caching of computed values.

Implementation Changes

  • Simplification of repoze.bfg.traversal.TraversalContextURL.__call__ (it now uses repoze.bfg.traversal.model_path instead of rolling its own path-generation).

0.6.8 (2009-02-05)

Backwards Incompatibilities

  • The repoze.bfg.traversal.model_path API now returns a quoted string rather than a string represented by series of unquoted elements joined via / characters. Previously it returned a string or unicode object representing the model path, with each segment name in the path joined together via / characters, e.g. /foo /bar. Now it returns a string, where each segment is a UTF-8 encoded and URL-quoted element e.g. /foo%20/bar. This change was (as discussed briefly on the repoze-dev maillist) necessary to accommodate model objects which themselves have __name__ attributes that contain the / character.

    For people that have no models that have high-order Unicode __name__ attributes or __name__ attributes with values that require URL-quoting with in their model graphs, this won't cause any issue. However, if you have code that currently expects model_path to return an unquoted string, or you have an existing application with data generated via the old method, and you're too lazy to change anything, you may wish replace the BFG-imported model_path in your code with this function (this is the code of the "old" model_path implementation):

    from repoze.bfg.location import lineage
    def i_am_too_lazy_to_move_to_the_new_model_path(model, *elements):
        rpath = []
        for location in lineage(model):
            if location.__name__:
        path = '/' + '/'.join(reversed(rpath))
        if elements:
            suffix = '/'.join(elements)
            path = '/'.join([path, suffix])
        return path
  • The repoze.bfg.traversal.find_model API no longer implicitly converts unicode representations of a full path passed to it as a Unicode object into a UTF-8 string. Callers should either use prequoted path strings returned by repoze.bfg.traversal.model_path, or tuple values returned by the result of repoze.bfg.traversal.model_path_tuple or they should use the guidelines about passing a string path argument described in the find_model API documentation.


  • Each argument contained in elements passed to repoze.bfg.traversal.model_path will now have any / characters contained within quoted to %2F in the returned string. Previously, / characters in elements were left unquoted (a bug).


  • A repoze.bfg.traversal.model_path_tuple API was added. This API is an alternative to model_path (which returns a string); model_path_tuple returns a model path as a tuple (much like Zope's getPhysicalPath).

  • A repoze.bfg.traversal.quote_path_segment API was added. This API will quote an individual path segment (string or unicode object). See the repoze.bfg.traversal API documentation for more information.

  • The repoze.bfg.traversal.find_model API now accepts "path tuples" (see the above note regarding model_path_tuple) as well as string path representations (from repoze.bfg.traversal.model_path) as a path argument.

  • Add ` renderer` argument (defaulting to None) to repoze.bfg.testing.registerDummyRenderer. This makes it possible, for instance, to register a custom renderer that raises an exception in a unit test.

Implementation Changes

  • Moved _url_quote function back to repoze.bfg.traversal from repoze.bfg.url. This is not an API.

0.6.7 (2009-01-27)


  • The repoze.bfg.url.model_url API now works against contexts derived from Routes URL dispatch (Routes.util.url_for is called under the hood).

  • "Virtual root" support for traversal-based applications has been added. Virtual root support is useful when you'd like to host some model in a repoze.bfg model graph as an application under a URL pathname that does not include the model path itself. For more information, see the (new) "Virtual Hosting" chapter in the documentation.

  • A repoze.bfg.traversal.virtual_root API has been added. When called, it returns the virtual root object (or the physical root object if no virtual root has been specified).

Implementation Changes

  • repoze.bfg.traversal.RoutesModelTraverser has been moved to repoze.bfg.urldispatch.

  • model_url URL generation is now performed via an adapter lookup based on the context and the request.

  • ZCML which registers two adapters for the IContextURL interface has been added to the configure.zcml in repoze.bfg.includes.

0.6.6 (2009-01-26)

Implementation Changes

  • There is an indirection in repoze.bfg.url.model_url now that consults a utility to generate the base model url (without extra elements or a query string). Eventually this will service virtual hosting; for now it's undocumented and should not be hooked.

0.6.5 (2009-01-26)


  • You can now override the NotFound and Unauthorized responses that repoze.bfg generates when a view cannot be found or cannot be invoked due to lack of permission. See the "ZCML Hooks" chapter in the docs for more information.

  • Added Routes ZCML directive attribute explanations in documentation.

  • Added a traversal_path API to the traversal module; see the "traversal" API chapter in the docs. This was a function previously known as split_path that was not an API but people were using it anyway. Unlike split_path, it now returns a tuple instead of a list (as its values are cached).

Behavior Changes

  • The repoze.bfg.view.render_view_to_response API will no longer raise a ValueError if an object returned by a view function it calls does not possess certain attributes (headerlist, app_iter, status). This API used to attempt to perform a check using the is_response function in repoze.bfg.view, and raised a ValueError if the is_response check failed. The responsibility is now the caller's to ensure that the return value from a view function is a "real" response.

  • WSGI environ dicts passed to repoze.bfg 's Router must now contain a REQUEST_METHOD key/value; if they do not, a KeyError will be raised (speed).

  • It is no longer permissible to pass a "nested" list of principals to repoze.bfg.ACLAuthorizer.permits (e.g. ['fred', ['larry', 'bob']]). The principals list must be fully expanded. This feature was never documented, and was never an API, so it's not a backwards incompatibility.

  • It is no longer permissible for a security ACE to contain a "nested" list of permissions (e.g. (Allow, Everyone, ['read', ['view', ['write', 'manage']]])`)`.  The list must instead be fully expanded (e.g. ``(Allow, Everyone, ['read', 'view', 'write', 'manage])). This feature was never documented, and was never an API, so it's not a backwards incompatibility.

  • The repoze.bfg.urldispatch.RoutesRootFactory now injects the wsgiorg.routing_args environment variable into the environ when a route matches. This is a tuple of ((), routing_args) where routing_args is the value that comes back from the routes mapper match (the "match dict").

  • The repoze.bfg.traversal.RoutesModelTraverser class now wants to obtain the view_name and subpath from the wsgiorgs.routing_args environment variable. It falls back to obtaining these from the context for backwards compatibility.

Implementation Changes

  • Get rid of the ACLSecurityPolicy now does what it did inline.

  • Get rid of repoze.bfg.interfaces.NoAuthorizationInformation exception: it was used only by ACLAuthorizer.

  • Use a homegrown NotFound error instead of webob.exc.HTTPNotFound (the latter is slow).

  • Use a homegrown Unauthorized error instead of webob.exc.Unauthorized (the latter is slow).

  • the repoze.bfg.lru.lru_cached decorator now uses functools.wraps in order to make documentation of LRU-cached functions possible.

  • Various speed micro-tweaks.

Bug Fixes

  • repoze.bfg.testing.DummyModel did not have a get method; it now does.

0.6.4 (2009-01-23)

Backwards Incompatibilities

  • The unicode_path_segments configuration variable and the BFG_UNICODE_PATH_SEGMENTS configuration variable have been removed. Path segments are now always passed to model __getitem__ methods as unicode. "True" has been the default for this setting since 0.5.4, but changing this configuration setting to false allowed you to go back to passing raw path element strings to model __getitem__ methods. Removal of this knob services a speed goal (we get about +80 req/s by removing the check), and it's clearer just to always expect unicode path segments in model __getitem__ methods.

Implementation Changes

  • repoze.bfg.traversal.split_path now also handles decoding path segments to unicode (for speed, because its results are cached).

  • repoze.bfg.traversal.step was made a method of the


  • Use "precooked" Request subclasses (e.g. repoze.bfg.request.GETRequest) that correspond to HTTP request methods within when constructing a request object rather than using alsoProvides to attach the proper interface to an unsubclassed webob.Request. This pattern is purely an optimization (e.g. preventing calls to alsoProvides means the difference between 590 r/s and 690 r/s on a MacBook 2GHz).

  • Tease out an extra 4% performance boost by changing the Router; instead of using imported ZCA APIs, use the same APIs directly against the registry that is an attribute of the Router.

  • The registry used by BFG is now a subclass of zope.component.registry.Components (defined as repoze.bfg.registry.Registry); it has a notify method, a registerSubscriptionAdapter and a registerHandler method. If no subscribers are registered via registerHandler or registerSubscriptionAdapter, notify is a noop for speed.

  • The Allowed and Denied classes in now are lazier about constructing the representation of a reason message for speed; repoze.bfg.view_execution_permitted takes advantage of this.

  • The is_response check was sped up by about half at the expense of making its code slightly uglier.

New Modules

  • repoze.bfg.lru implements an LRU cache class and a decorator for internal use.

0.6.3 (2009-01-19)

Bug Fixes

  • Readd root_policy attribute on Router object (as a property which returns the IRootFactory utility). It was inadvertently removed in 0.6.2. Code in the wild depended upon its presence (esp. scripts and "debug" helpers).


  • URL-dispatch has been overhauled: it is no longer necessary to manually create a RoutesMapper in your application's entry point callable in order to use URL-dispatch (aka Routes). A new route directive has been added to the available list of ZCML directives. Each route directive inserted into your application's configure.zcml establishes a Routes mapper connection. If any route declarations are made via ZCML within a particular application, the get_root callable passed in to repoze.bfg.router.make_app will automatically be wrapped in the equivalent of a RoutesMapper. Additionally, the new route directive allows the specification of a context_interfaces attribute for a route, this will be used to tag the manufactured routes context with specific interfaces when a route specifying a context_interfaces attribute is matched.

  • A new interface repoze.bfg.interfaces.IContextNotFound was added. This interface is attached to a "dummy" context generated when Routes cannot find a match and there is no "fallback" get_root callable that uses traversal.

  • The bfg_starter and bfg_zodb "paster create" templates now contain images and CSS which are displayed when the default page is displayed after initial project generation.

  • Allow the repoze.bfg.view.static helper to be passed a relative root_path name; it will be considered relative to the file in which it was called.

  • The functionality of repoze.bfg.convention has been merged into the core. Applications which make use of repoze.bfg.convention will continue to work indefinitely, but it is recommended that apps stop depending upon it. To do so, substitute imports of repoze.bfg.convention.bfg_view with imports of repoze.bfg.view.bfg_view, and change the stanza in ZCML from <convention package="."> to <scan package=".">. As a result of the merge, bfg has grown a new dependency: martian.

  • View functions which use the pushpage decorator are now pickleable (meaning their use won't prevent a configure.zcml.cache file from being written to disk).

  • Instead of invariably using webob.Request as the "request factory" (e.g. in the Router class) and webob.Response and the "response factory" (e.g. in render_template_to_response), allow both to be overridden via a ZCML utility hook. See the "Using ZCML Hooks" chapter of the documentation for more information.


  • The class repoze.bfg.urldispatch.RoutesContext has been renamed to repoze.bfg.urldispatch.DefaultRoutesContext. The class should be imported by the new name as necessary (although in reality it probably shouldn't be imported from anywhere except internally within BFG, as it's not part of the API).

Implementation Changes

  • The repoze.bfg.wsgi.wsgiapp decorator now uses webob.Request.get_response to do its work rather than relying on homegrown WSGI code.

  • The repoze.bfg.view.static helper now uses webob.Request.get_response to do its work rather than relying on homegrown WSGI code.

  • The repoze.bfg.urldispatch.RoutesModelTraverser class has been moved to repoze.bfg.traversal.RoutesModelTraverser.

  • The repoze.bfg.registry.makeRegistry function was renamed to repoze.bfg.registry.populateRegistry and now accepts a registry argument (which should be an instance of zope.component.registry.Components).

Documentation Additions

  • Updated narrative urldispatch chapter with changes required by <route..> ZCML directive.

  • Add a section on "Using BFG Security With URL Dispatch" into the urldispatch chapter of the documentation.

  • Better documentation of security policy implementations that ship with repoze.bfg.

  • Added a "Using ZPT Macros in repoze.bfg" section to the narrative templating chapter.

0.6.2 (2009-01-13)


  • Tests can be run with coverage output if you've got nose installed in the interpreter which you use to run tests. Using an interpreter with nose installed, do python nosetests within a checkout of the repoze.bfg package to see test coverage output.

  • Added a post argument to the repoze.bfg.testing:DummyRequest constructor.

  • Added __len__ and __nonzero__ to repoze.bfg.testing:DummyModel.

  • The repoze.bfg.registry.get_options callable (now renamed to repoze.bfg.setings.get_options) used to return only framework-specific keys and values in the dictionary it returned. It now returns all the keys and values in the dictionary it is passed plus any framework-specific settings culled from the environment. As a side effect, all PasteDeploy application-specific config file settings are made available as attributes of the ISettings utility from within BFG.

  • Renamed the existing BFG paster template to bfg_starter. Added another template (bfg_zodb) showing default ZODB setup using repoze.zodbconn.

  • Add a method named assert_ to the DummyTemplateRenderer. This method accepts keyword arguments. Each key/value pair in the keyword arguments causes an assertion to be made that the renderer received this key with a value equal to the asserted value.

  • Projects generated by the paster templates now use the DummyTemplateRenderer.assert_ method in their view tests.

  • Make the (internal) thread local registry manager maintain a stack of registries in order to make it possible to call one BFG application from inside another.

  • An interface specific to the HTTP verb (GET/PUT/POST/DELETE/HEAD) is attached to each request object on ingress. The HTTP-verb-related interfaces are defined in repoze.bfg.interfaces and are IGETRequest, IPOSTRequest, IPUTRequest, IDELETERequest and IHEADRequest. These interfaces can be specified as the request_type attribute of a bfg view declaration. A view naming a specific HTTP-verb-matching interface will be found only if the view is defined with a request_type that matches the HTTP verb in the incoming request. The more general IRequest interface can be used as the request_type to catch all requests (and this is indeed the default). All requests implement IRequest. The HTTP-verb-matching idea was pioneered by repoze.bfg.restrequest . That package is no longer required, but still functions fine.

Bug Fixes

  • Fix a bug where the Paste configuration's unicode_path_segments (and os.environ's BFG_UNICODE_PATH_SEGMENTS) may have been defaulting to false in some circumstances. It now always defaults to true, matching the documentation and intent.

  • The repoze.bfg.traversal.find_model API did not work properly when passed a path argument which was unicode and contained high-order bytes when the unicode_path_segments or BFG_UNICODE_PATH_SEGMENTS configuration variables were "true".

  • A new module was added: repoze.bfg.settings. This contains deployment-settings-related code.

Implementation Changes

  • The make_app callable within repoze.bfg.router now registers the root_policy argument as a utility (unnamed, using the new repoze.bfg.interfaces.IRootFactory as a provides interface) rather than passing it as the first argument to the repoze.bfg.router.Router class. As a result, the repoze.bfg.router.Router router class only accepts a single argument: registry. The repoze.bfg.router.Router class retrieves the root policy via a utility lookup now. The repoze.bfg.router.make_app API also now performs some important application registrations that were previously handled inside repoze.bfg.registry.makeRegistry.

New Modules

  • A repoze.bfg.settings module was added. It contains code related to deployment settings. Most of the code it contains was moved to it from the repoze.bfg.registry module.

Behavior Changes

  • The repoze.bfg.settings.Settings class (an instance of which is registered as a utility providing repoze.bfg.interfaces.ISettings when any application is started) now automatically calls repoze.bfg.settings.get_options on the options passed to its constructor. This means that usage of get_options within an application's make_app function is no longer required (the "raw" options dict or None may be passed).

  • Remove old cold which attempts to recover from trying to unpickle a template; Chameleon has been the templating engine for a good long time now. Running repoze.bfg against a sandbox that has pickled templates it will now just fail with an unpickling error, but can be fixed by deleting the template cache files.


  • Moved the repoze.bfg.registry.Settings class. This has been moved to repoze.bfg.settings.Settings. A deprecation warning is issued when it is imported from the older location.

  • Moved the repoze.bfg.registry.get_options function This has been moved to repoze.bfg.settings.get_options. A deprecation warning is issued when it is imported from the older location.

  • The repoze.bfg.interfaces.IRootPolicy interface was renamed within the interfaces package. It has been renamed to IRootFactory. A deprecation warning is issued when it is imported from the older location.

0.6.1 (2009-01-06)

New Modules

  • A new module repoze.bfg.url has been added. It contains the model_url API (moved from repoze.bfg.traversal) and an implementation of urlencode (like Python's urllib.urlencode) which can handle Unicode keys and values in parameters to the query argument.


  • The model_url function has been moved from repoze.bfg.traversal into repoze.bfg.url. It can still be imported from repoze.bfg.traversal but an import from repoze.bfg.traversal will emit a DeprecationWarning.


  • A static helper class was added to the repoze.bfg.views module. Instances of this class are willing to act as BFG views which return static resources using files on disk. See the repoze.bfg.view docs for more info.

  • The repoze.bfg.url.model_url API (nee' repoze.bfg.traversal.model_url) now accepts and honors a keyword argument named query. The value of this argument will be used to compose a query string, which will be attached to the generated URL before it is returned. See the API docs (in the docs directory or on the web) for more information.

0.6 (2008-12-26)

Backwards Incompatibilities

  • Rather than prepare the "stock" implementations of the ZCML directives from the zope.configuration package for use under repoze.bfg, repoze.bfg now makes available the implementations of directives from the repoze.zcml package (see As a result, the repoze.bfg package now depends on the repoze.zcml package, and no longer depends directly on the zope.component, zope.configuration, zope.interface, or zope.proxy packages.

    The primary reason for this change is to enable us to eventually reduce the number of inappropriate repoze.bfg Zope package dependencies, as well as to shed features of dependent package directives that don't make sense for repoze.bfg.

    Note that currently the set of requirements necessary to use bfg has not changed. This is due to inappropriate Zope package requirements in chameleon.zpt, which will hopefully be remedied soon. NOTE: in lemonade index a 1.0b8-repozezcml0 package exists which does away with these requirements.

  • BFG applications written prior to this release which expect the "stock" zope.component ZCML directive implementations (e.g. adapter, subscriber, or utility) to function now must either 1) include the meta.zcml file from zope.component manually (e.g. <include package="zope.component" file="meta.zcml">) and include the package as an install_requires dependency or 2) change the ZCML in their applications to use the declarations from repoze.zcml instead of the stock declarations. repoze.zcml only makes available the adapter, subscriber and utility directives.

    In short, if you've got an existing BFG application, after this update, if your application won't start due to an import error for "", the fastest way to get it working again is to add to the "install_requires" of your BFG application's, then add the following ZCML anywhere in your application's configure.zcml:

    <include package="zope.component" file="meta.zcml">

    Then develop or reinstall your application.

  • The XML namespace is now the default XML namespace in ZCML for paster-generated applications. The docs have been updated to reflect this.

  • The copies of BFG's meta.zcml and configure.zcml were removed from the root of the repoze.bfg package. In 0.3.6, a new package named repoze.bfg.includes was added, which contains the "correct" copies of these ZCML files; the ones that were removed were for backwards compatibility purposes.

  • The BFG view ZCML directive no longer calls zope.component.interface.provideInterface for the for interface. We don't support provideInterface in BFG because it mutates the global registry.


  • The minimum requirement for chameleon.core is now 1.0b13. The minimum requirement for chameleon.zpt is now 1.0b8. The minimum requirement for chameleon.genshi is now 1.0b2.

  • Updated paster template "" to one that requires setuptools 0.6c9.

  • Turn view_execution_permitted from the repoze.bfg.view module into a documented API.

  • Doc cleanups.

  • Documented how to create a view capable of serving static resources.

0.5.6 (2008-12-18)

  • Speed up traversal.model_url execution by using a custom url quoting function instead of Python's urllib.quote, by caching URL path segment quoting and encoding results, by disusing Python's urlparse.urljoin in favor of a simple string concatenation, and by using ob.__class__ is unicode rather than isinstance(ob, unicode) in one strategic place.

0.5.5 (2008-12-17)

Backwards Incompatibilities

  • In the past, during traversal, the ModelGraphTraverser (the default traverser) always passed each URL path segment to any __getitem__ method of a model object as a byte string (a str object). Now, by default the ModelGraphTraverser attempts to decode the path segment to Unicode (a unicode object) using the UTF-8 encoding before passing it to the __getitem__ method of a model object. This makes it possible for model objects to be dumber in __getitem__ when trying to resolve a subobject, as model objects themselves no longer need to try to divine whether or not to try to decode the path segment passed by the traverser.

    Note that since 0.5.4, URLs generated by repoze.bfg's model_url API will contain UTF-8 encoded path segments as necessary, so any URL generated by BFG itself will be decodeable by the traverser. If another application generates URLs to a BFG application, to be resolved successfully, it should generate the URL with UTF-8 encoded path segments to be successfully resolved. The decoder is not at all magical: if a non-UTF-8-decodeable path segment (e.g. one encoded using UTF-16 or some other insanity) is passed in the URL, BFG will raise a TypeError with a message indicating it could not decode the path segment.

    To turn on the older behavior, where path segments were not decoded to Unicode before being passed to model object __getitem__ by the traverser, and were passed as a raw byte string, set the unicode_path_segments configuration setting to a false value in your BFG application's section of the paste .ini file, for example:

    unicode_path_segments = False

    Or start the application using the BFG_UNICODE_PATH_SEGMENT envvar set to a false value:


0.5.4 (2008-12-13)

Backwards Incompatibilities

  • URL-quote "extra" element names passed in as **elements to the traversal.model_url API. If any of these names is a Unicode string, encode it to UTF-8 before URL-quoting. This is a slight backwards incompatibility that will impact you if you were already UTF-8 encoding or URL-quoting the values you passed in as elements to this API.


  • UTF-8 encode each segment in the model path used to generate a URL before url-quoting it within the traversal.model_url API. This is a bugfix, as Unicode cannot always be successfully URL-quoted.


  • Make it possible to run unit tests using a buildout-generated Python "interpreter".

  • Add request.root to router.Router in order to have easy access to the application root.

0.5.3 (2008-12-07)

  • Remove the ITestingTemplateRenderer interface. When testing.registerDummyRenderer is used, it instead registers a dummy implementation using ITemplateRenderer interface, which is checked for when the built-in templating facilities do rendering. This change also allows developers to make explicit named utility registrations in the ZCML registry against ITemplateRenderer; these will be found before any on-disk template is looked up.

0.5.2 (2008-12-05)

  • The component registration handler for views (functions or class instances) now observes component adaptation annotations (see zope.component.adaptedBy) and uses them before the fallback values for for_ and request_type. This change does not affect existing code insomuch as the code does not rely on these defaults when an annotation is set on the view (unlikely). This means that for a new-style class you can do zope.component.adapts(ISomeContext, ISomeRequest) at class scope or at module scope as a decorator to a bfg view function you can do @zope.component.adapter(ISomeContext, ISomeRequest). This differs from r.bfg.convention inasmuch as you still need to put something in ZCML for the registrations to get done; it's only the defaults that will change if these declarations exist.

  • Strip all slashes from end and beginning of path in clean_path within traversal machinery.

0.5.1 (2008-11-25)

  • Add keys, items, and values methods to testing.DummyModel.

  • Add __delitem__ method to testing.DummyModel.

0.5.0 (2008-11-18)

  • Fix ModelGraphTraverser; don't try to change the __name__ or __parent__ of an object that claims it implements ILocation during traversal even if the __name__ or __parent__ of the object traversed does not match the name used in the traversal step or the or the traversal parent . Rationale: it was insane to do so. This bug was only found due to a misconfiguration in an application that mistakenly had intermediate persistent non-ILocation objects; traversal was causing a persistent write on every request under this setup.

  • repoze.bfg.location.locate now unconditionally sets __name__ and __parent__ on objects which provide ILocation (it previously only set them conditionally if they didn't match attributes already present on the object via equality).

0.4.9 (2008-11-17)

  • Add chameleon text template API (chameleon ${name} renderings where the template does not need to be wrapped in any containing XML).

  • Change docs to explain install in terms of a virtualenv (unconditionally).

  • Make pushpage decorator compatible with repoze.bfg.convention's bfg_view decorator when they're stacked.

  • Add content_length attribute to testing.DummyRequest.

  • Change paster template to include a true unit test. Retain old test as an integration test. Update documentation.

  • Document view registrations against classes and repoze.bfg.convention in context.

  • Change the default paster template to register its single view against a class rather than an interface.

  • Document adding a request type interface to the request via a subscriber function in the events narrative documentation.

0.4.8 (2008-11-12)

Backwards Incompatibilities

  • repoze.bfg.traversal.model_url now always appends a slash to all generated URLs unless further elements are passed in as the third and following arguments. Rationale: views often use model_url without the third-and-following arguments in order to generate a URL for a model in order to point at the default view of a model. The URL that points to the default view of the root model is technically http://mysite/ as opposed to http://mysite (browsers happen to ask for '/' implicitly in the GET request). Because URLs are never automatically generated for anything except models by model_url, and because the root model is not really special, we continue this pattern. The impact of this change is minimal (at most you will have too many slashes in your URL, which BFG deals with gracefully anyway).

0.4.7 (2008-11-11)


  • Allow testing.registerEventListener to be used with Zope 3 style "object events" (subscribers accept more than a single event argument). We extend the list with the arguments, rather than append.

0.4.6 (2008-11-10)

Bug Fixes

  • The model_path and model_url traversal APIs returned the wrong value for the root object (e.g. model_path returned '' for the root object, while it should have been returning '/').

0.4.5 (2008-11-09)


  • Added a clone method and a __contains__ method to the DummyModel testing object.

  • Allow DummyModel objects to receive extra keyword arguments, which will be attached as attributes.

  • The DummyTemplateRenderer now returns self as its implementation.

0.4.4 (2008-11-08)


  • Added a repoze.bfg.testing module to attempt to make it slightly easier to write unittest-based automated tests of BFG applications. Information about this module is in the documentation.

  • The default template renderer now supports testing better by looking for ITestingTemplateRenderer using a relative pathname. This is exposed indirectly through the API named registerTemplateRenderer in repoze.bfg.testing.


  • The names repoze.bfg.interfaces.ITemplate , repoze.bfg.interfaces.ITemplateFactory and repoze.bfg.interfaces.INodeTemplate have been deprecated. These should now be imported as repoze.bfg.interfaces.ITemplateRenderer and repoze.bfg.interfaces.ITemplateRendererFactory, and INodeTemplateRenderer respectively.

  • The name repoze.bfg.chameleon_zpt.ZPTTemplateFactory is deprecated. Use repoze.bfg.chameleon_zpt.ZPTTemplateRenderer.

  • The name repoze.bfg.chameleon_genshi.GenshiTemplateFactory is deprecated. Use repoze.bfg.chameleon_genshi.GenshiTemplateRenderer.

  • The name repoze.bfg.xslt.XSLTemplateFactory is deprecated. Use repoze.bfg.xslt.XSLTemplateRenderer.

0.4.3 (2008-11-02)

Bug Fixes

  • Not passing the result of "get_options" as the second argument of make_app could cause attribute errors when attempting to look up settings against the ISettings object (internal). Fixed by giving the Settings objects defaults for debug_authorization and debug_notfound.

  • Return an instance of Allowed (rather than True) from has_permission when no security policy is in use.

  • Fix bug where default deny in authorization check would throw a TypeError (use ACLDenied instead of Denied).

0.4.2 (2008-11-02)


  • Expose a single ILogger named "repoze.bfg.debug" as a utility; this logger is registered unconditionally and is used by the authorization debug machinery. Applications may also make use of it as necessary rather than inventing their own logger, for convenience.

  • The BFG_DEBUG_AUTHORIZATION envvar and the debug_authorization config file value now only imply debugging of view-invoked security checks. Previously, information was printed for every call to has_permission as well, which made output confusing. To debug has_permission checks and other manual permission checks, use the debugger and print statements in your own code.

  • Authorization debugging info is now only present in the HTTP response body oif debug_authorization is true.

  • The format of authorization debug messages was improved.

  • A new BFG_DEBUG_NOTFOUND envvar was added and a symmetric debug_notfound config file value was added. When either is true, and a NotFound response is returned by the BFG router (because a view could not be found), debugging information is printed to stderr. When this value is set true, the body of HTTPNotFound responses will also contain the same debugging information.

  • Allowed and Denied responses from the security machinery are now specialized into two types: ACL types, and non-ACL types. The ACL-related responses are instances of and The non-ACL-related responses are and The allowed-type responses continue to evaluate equal to things that themselves evaluate equal to the True boolean, while the denied-type responses continue to evaluate equal to things that themselves evaluate equal to the False boolean. The only difference between the two types is the information attached to them for debugging purposes.

  • Added a new BFG_DEBUG_ALL envvar and a symmetric debug_all config file value. When either is true, all other debug-related flags are set true unconditionally (e.g. debug_notfound and debug_authorization).


  • Added info about debug flag changes.

  • Added a section to the security chapter named "Debugging Imperative Authorization Failures" (for e.g. has_permssion).

Bug Fixes

  • Change default paster template generator to use Paste#http server rather than PasteScript#cherrpy server. The cherrypy server has a security risk in it when REMOTE_USER is trusted by the downstream application.

0.4.1 (2008-10-28)

Bug Fixes

  • If the render_view_to_response function was called, if the view was found and called, but it returned something that did not implement IResponse, the error would pass by unflagged. This was noticed when I created a view function that essentially returned None, but received a NotFound error rather than a ValueError when the view was rendered. This was fixed.

0.4.0 (2008-10-03)


  • An "Environment and Configuration" chapter was added to the narrative portion of the documentation.


  • Ensure bfg doesn't generate warnings when running under Python 2.6.

  • The environment variable BFG_RELOAD_TEMPLATES is now available (serves the same purpose as reload_templates in the config file).

  • A new configuration file option debug_authorization was added. This turns on printing of security authorization debug statements to sys.stderr. The BFG_DEBUG_AUTHORIZATION environment variable was also added; this performs the same duty.

Bug Fixes

  • The environment variable BFG_SECURITY_DEBUG did not always work. It has been renamed to BFG_DEBUG_AUTHORIZATION and fixed.


  • A deprecation warning is now issued when old API names from the repoze.bfg.templates module are imported.

Backwards incompatibilities

  • The BFG_SECURITY_DEBUG environment variable was renamed to BFG_DEBUG_AUTHORIZATION.

0.3.9 (2008-08-27)


  • A repoze.bfg.location API module was added.

Backwards incompatibilities

  • Applications must now use the repoze.bfg.interfaces.ILocation interface rather than zope.location.interfaces.ILocation to represent that a model object is "location-aware". We've removed a dependency on zope.location for cleanliness purposes: as new versions of zope libraries are released which have improved dependency information, getting rid of our dependence on zope.location will prevent a newly installed repoze.bfg application from requiring the, egg, which not truly used at all in a "stock" repoze.bfg setup. These dependencies are still required by the stack at this time; this is purely a futureproofing move.

    The security and model documentation for previous versions of repoze.bfg recommended using the zope.location.interfaces.ILocation interface to represent that a model object is "location-aware". This documentation has been changed to reflect that this interface should now be imported from repoze.bfg.interfaces.ILocation instead.

0.3.8 (2008-08-26)


  • Documented URL dispatch better in narrative form.

Bug fixes

  • Routes URL dispatch did not have access to the WSGI environment, so conditions such as method=GET did not work.


  • Add principals_allowed_by_permission API to security module.

  • Replace support with support for chameleon.zpt. Chameleon is the new name for the package that used to be named NOTE: If you update a repoze.bfg SVN checkout that you're using for development, you will need to run " install" or " develop" again in order to obtain the proper Chameleon packages. is no longer supported by repoze.bfg. All API functions that used to render templates will work fine with the new packages, and your templates should render almost identically.

  • Add a repoze.bfg.chameleon_zpt module. This module provides Chameleon ZPT support.

  • Add a repoze.bfg.xslt module. This module provides XSLT support.

  • Add a repoze.bfg.chameleon_genshi module. This provides direct Genshi support, which did not exist previously.


  • Importing API functions directly from repoze.bfg.template is now deprecated. The get_template, render_template, render_template_to_response functions should now be imported from repoze.chameleon_zpt. The render_transform, and render_transform_to_response functions should now be imported from repoze.bfg.xslt. The repoze.bfg.template module will remain around "forever" to support backwards compatibility.

0.3.7 (2008-09-09)


  • Add compatibility with 1.0a7+ ( became a namespace package).

Bug fixes

  • repoze.bfg.traversal.find_model function did not function properly.

0.3.6 (2008-09-04)


  • Add startup process docs.

  • Allow configuration cache to be bypassed by actions which include special "uncacheable" discriminators (for actions that have variable results).

Bug Fixes

  • Move core repoze.bfg ZCML into a repoze.bfg.includes package so we can use repoze.bfg better as a namespace package. Adjust the code generator to use it. We've left around the configure.zcml in the repoze.bfg package directly so as not to break older apps.

  • When a zcml application registry cache was unpickled, and it contained a reference to an object that no longer existed (such as a view), bfg would not start properly.

0.3.5 (2008-09-01)


  • Event notification is issued after application is created and configured (IWSGIApplicationCreatedEvent).

  • New API module: repoze.bfg.view. This module contains the functions named render_view_to_response, render_view_to_iterable, render_view and is_response, which are documented in the API docs. These features aid programmatic (non-server-driven) view execution.

0.3.4 (2008-08-28)

Backwards incompatibilities

  • Make repoze.bfg a namespace package so we can allow folks to create subpackages (e.g. repoze.bfg.otherthing) within separate eggs. This is a backwards incompatible change which makes it impossible to import "make_app" and "get_options" from the repoze.bfg module directly. This change will break all existing apps generated by the paster code generator. Instead, you need to import these functions as repoze.bfg.router:make_app and repoze.bfg.registry:get_options, respectively. Sorry folks, it has to be done now or never, and definitely better now.


  • Add model_path API function to traversal module.


  • Normalize path returned by repoze.bfg.caller_path.

0.3.3 (2008-08-23)

  • Fix generated module to use project name rather than package name.

0.3.2 (2008-08-23)

  • Remove sampleapp sample application from bfg package itself.

  • Remove dependency on FormEncode (only needed by sampleapp).

  • Fix paster template generation so that case-sensitivity is preserved for project vs. package name.

  • Depend on version 1.0a1 (which requires the [lxml] extra currently).

  • Read and write a pickled ZCML actions list, stored as configure.zcml.cache next to the applications's "normal" configuration file. A given bfg app will usually start faster if it's able to read the pickle data. It fails gracefully to reading the real ZCML file if it cannot read the pickle.

0.3.1 (2008-08-20)

  • Generated application differences: make_app entry point renamed to app in order to have a different name than the bfg function of the same name, to prevent confusion.

  • Add "options" processing to bfg's make_app to support runtime options. A new API function named get_options was added to the registry module. This function is typically used in an application's app entry point. The Paste config file section for the app can now supply the reload_templates option, which, if true, will prevent the need to restart the appserver in order for or XSLT template changes to be detected.

  • Use only the module name in generated project's "test_suite" (run all tests found in the package).

  • Default port for generated apps changed from 5432 to 6543 (Postgres default port is 6543).

0.3.0 (2008-08-16)

  • Add get_template API to template module.

0.2.9 (2008-08-11)

  • 0.2.8 was "brown bag" release. It didn't work at all. Symptom: ComponentLookupError when trying to render a page.

0.2.8 (2008-08-11)

  • Add find_model and find_root traversal APIs. In the process, make ITraverser a uni-adapter (on context) rather than a multiadapter (on context and request).

0.2.7 (2008-08-05)

  • Add a request_type attribute to the available attributes of a bfg:view configure.zcml element. This attribute will have a value which is a dotted Python path, pointing at an interface. If the request object implements this interface when the view lookup is performed, the appropriate view will be called. This is meant to allow for simple "skinning" of sites based on request type. An event subscriber should attach the interface to the request on ingress to support skins.

  • Remove "template only" views. These were just confusing and were never documented.

  • Small url dispatch overhaul: the connect method of the urldispatch.RoutesMapper object now accepts a keyword parameter named context_factory. If this parameter is supplied, it must be a callable which returns an instance. This instance is used as the context for the request when a route is matched.

  • The registration of a RoutesModelTraverser no longer needs to be performed by the application; it's in the bfg ZCML now.

0.2.6 (2008-07-31)

  • Add event sends for INewRequest and INewResponse. See the events.rst chapter in the documentation's api directory.

0.2.5 (2008-07-28)

  • Add model_url API.

0.2.4 (2008-07-27)

  • Added url-based dispatch.

0.2.3 (2008-07-20)

  • Add API functions for authenticated_userid and effective_principals.

0.2.2 (2008-07-20)

  • Add authenticated_userid and effective_principals API to security policy.

0.2.1 (2008-07-20)

  • Add find_interface API.

0.2 (2008-07-19)

  • Add wsgiapp decorator.

  • The concept of "view factories" was removed in favor of always calling a view, which is a callable that returns a response directly (as opposed to returning a view). As a result, the factory attribute in the bfg:view ZCML statement has been renamed to view. Various interface names were changed also.

  • render_template and render_transform no longer return a Response object. Instead, these return strings. The old behavior can be obtained by using render_template_to_response and render_transform_to_response.

  • Added 'repoze.bfg.push:pushpage' decorator, which creates BFG views from callables which take (context, request) and return a mapping of top-level names.

  • Added ACL-based security.

  • Support for XSLT templates via a render_transform method

0.1 (2008-07-08)

  • Initial release.