Branches

soc2009/http-wsgi-improvements

View source
soc2009/http-wsgi-improvements
  • Contributors
    1. Loading...
Author Commit Message Labels Comments Date
250+ commits behind default.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Update versions in docs for the various new HttpResponse features, and fixed an error on the argument list for HttpResponse.__init__.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Adds http.HttpResponseStreaming, with docs, tests, and support in four built-in middleware classes. Refs #7581.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Clean up charset-related code in HttpResponse. Refs #10190.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to r11423.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Update the render_to_response shortcut to pass request and content_type to HttpResponse, to support Accept-Charset. Refs #10190. Also updates docs.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Fixes for HttpResponse.charsets docs and HttpResponse._charset. Refs #10190.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Change the setting name to SENDFILE_HEADER and update docs. Refs #2131.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Remove setting the Content-Length header for HttpResponseSendFile from the handler, for compatibility, and add a content attribute. Refs #2131. Also adds a _charset class attribute to HttpResponse so the children all have it.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Add a how-to for using HttpResponseSendFile with various server arrangements. Refs #2131.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to r11318 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Fix early settings use in HttpResponseSendFile. Refs #2131.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Changes for ``HttpResponseSendFile`` support in FastCGI. Refs #2131.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvments] ``HttpResponseSendFile`` now works (again?) using the X-SendFile header under the wsgi handler. refs #2131 Tests for other headers (and mod_python) are underway.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Establish the priorities and fallbacks for HttpResponseSendFile methods. Change the setting to settings.HTTPRESPONSE_SENDFILE_METHOD. If this is set to None, we use handler methods, but otherwise the header gets set, and we do not send any content. If neither of these are available, use the FileWrapper fallback in HttpResponseSendFile. This passes the test suite, but is untested on mod_python. I am still trying to figure out how to view the headers of a response with …
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Throw an exception when HttpResponse.codec is set with a bad value. Improved coverage of encoding changes in request and response headers. Refs #10190. Passes the test suite.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Add docs that I missed from the patch and reformat HttpResponse.__str__. refs #6527 This and the previous revision on this branch appear to complete all the changes from #6527.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Expects content to be iterable throughout ``HttpResponse``, with regression tests. HttpResponse.close() is now a no-op. Thanks daonb and stugots! Passes the test suite.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Fix HttpResponseSendFile indentation issue and set its initial content to be an empty string, for compatibility with ticket refs #6527.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to r12258 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Added the codec attribute/property to HttpResponse, and added docs for recent work with charsets/codecs.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[gsoc2009/http-wsgi-improvements] Clean up imports in django.http and django.http.charsets.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] HttpResponseSendFile now uses django.core.servers.basehttp.FileWrapper inside __iter__ to provide fallback. regressiontests.sendfile uses this now, and passes. The fallback was tested using guppy and apache2 with mod_wsgi for heap issues, and it appears to be fine. We can go back and look at this again if it becomes an issue.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Refactor setting 406 status codes in HttpResponse. Accessing HttpResponse.content now evaluates the status_code for side effects.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvement] Repairs to initial regressiontests.sendfile test to fix temp file usage.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Change HttpResponse.status_code to a property, additional test coverage. Refs #10190. Improve charset coverage. Change HttpResponse.status_code to property, which checks for a 406 situation. This also required changes to all HttpResponse subclasses, so that their default status_code is set by _status_code, now. Passes regression test suite, except ones related to sendfile.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to r11176 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Fix core.handler.BaseHandler.process_request. Indentation issue.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Initial HttpResponseSendFile support, changes pulled from 03/21/09 patch on refs #2131. This does not pass the included regression tests. However, since this feature will be entirely based on these changes, which have already gone through a great number of iterations, I thought it would be sensible to start here. All of the work here is ymasuda, mizatservercave, and mrts (apologies if I missed anyone). I hope to take their work down the final stretch.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to r11127 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Modified tests/test_client_regress to expect correct behavior, which also fixes conflicts with charset handling in HttpResponse. The specific cases modified test if the test client will encode the data it posts such that when writers of test views try to use the request, the data will be in the encoding they expect from the tests cases they have written. We do not expect HttpRequest to decode the incoming content yet, so the test views have to be created beforehand so that they decode using the proper codec (which they can …
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to 11103 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] http.HttpResponse now initializes self._codec to None. Changeset 11030 did not actually pass the regression test suite. This passes all except two tests, from test_client_regress. They expect old behavior (which is that their responses will be encoded in settings.DEFAULT_CHARSET). I am not yet sure if that is the correct expectation for the test to have, so I will do investigations into that.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Moved common code from core.handlers.modpython and core.handlers.wsgi into core.handlers.base.process_request. Passes regression tests.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Added more tests for #10190, changed logic to pass them. http.charsets.determine_charset now takes the accept_charset header instead of the request. Passes the test suite, including the extensive tests on HttpResponse's detection of Accept-Charset and finding the codec from content_type. However, it does not test that the codec encodes properly.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Code and tests in support of http.charsets.determine_charset. This code determines the charset from a content-type or from the Accept-Charset request header. The code is fairly well documented, but that will be improved once the code is in closer to a final form. The codec that corresponds to the charset is also returned, but it is not currently used by HttpResponse.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
[soc2009/http-wsgi-improvements] Merged up to 11009 from trunk.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #9253 -- Modified the method used to generate constraint names so that it is consistent regardless of machine word size. NOTE: This change is backwards incompatible for some users. If you are using a 32-bit platform, you will observe no differences as a result of this change. However, users on 64-bit platforms may experience some problems using the `reset` management command. Prior to this change, 64-bit platforms would generate a 64-bit, 16 character digest in the constraint name; for example: ALTER TABLE `myapp…
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11194 -- Corrected loading of Proxy models from fixtures (and, by extension, save_base(raw=True) for Proxy models).
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #10672 -- Altered save_base to ensure that proxy models send a post_save signal.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #10785 -- Corrected a case for foreign key lookup where the related object is a custom primary key. Thanks to Alex Gaynor for the patch.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11274 -- Corrected doctests to not cause test failures due to missing newlines. Thanks Honza Kral.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11271 -- Added a translation marker for the list_editable save button. Thanks to dc for the report.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11082 -- Ensured that subqueries used in an exclude(X__in=) clause aren't pre-evaluated. Thanks to Henry Andrews for the report, and clement for the fix.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #10733 -- Added a regression test for queries with multiple references to multiple foreign keys in only() clauses. Thanks to mrts for the report.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #10572 -- Corrected the operation of the defer() and only() clauses when used on inherited models.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #9479 -- Corrected an edge case in bulk queryset deletion that could cause an infinite loop when using MySQL InnoDB.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11087 -- Fixed the `Count` annotation when used with `GeoManager`. Thanks to dgouldin for ticket and initial patch.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11200 -- Now use a `set` data structure for `GoogleMap` icons so that they aren't repeated in rendered JavaScript. Thanks to ludifan for ticket and initial patch.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed a few Python 2.3 incompatibilities that were causing test failures.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Fixed #11216 and #11218 -- Corrected a few typos, thanks buriy.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Added test for pickling of a model with an `ImageField`, refs #11103.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Updated Argentinian spanish translation, 100% complete as of r10858.
cca...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Changes to `ImageFileDescriptor` and `ImageField` to fix a few cases of setting image dimension fields. * Moved dimension field update logic out of `ImageFileDescriptor.__set__` and into its own method on `ImageField`. * New `ImageField.update_dimension_fields` method is attached to model instance's `post_init` signal so that: * Dimension fields are set when defined before the ImageField. * Dimension fields are set when the field is assigned in the model constructor (fixes #11196), but only …
ja...@bcc190cf-cafb-0310-a4f2-bffc1f526a37
Created "http-wsgi-improvements" branch for Chris Cahoon's SoC project.