Bug 98527

Misha can you please have a look at this?

It looks like the patch in question is insufficient. After the web process returns "pass through" in response to InjectedBundlePage::decidePolicyForResponse(), it then queries the UIProcess through InjectedBundlePage::decidePolicyForResponse(). Because WKTR has not installed a WKPagePolicyClient, it defaults to "use", and the page loads the binary data.

Added this test to TestExpectations as Failing in r130536.

Think the described problem can be fixed with the same patch with bug#95974

(In reply to comment #4) > Think the described problem can be fixed with the same patch with bug#95974 Just uploaded patch for bug#95974 that contains partial fix for this bug -- it's introducing PagePolicyClient for WTR. Should also add decidePolicyForResponse callback to fix this bug problem.

Created attachment 168173 [details] patch

*** Bug 98994 has been marked as a duplicate of this bug. ***

Comment on attachment 168173 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=168173&action=review > LayoutTests/platform/efl-wk2/http/tests/loading/text-content-type-with-binary-extension-expected.txt:10 > +Binary data not loaded as text. Maybe PASS. Maybe PASS? > Tools/WebKitTestRunner/TestController.cpp:1081 > + if (WKURLResponseHTTPStatusCode(response) == 204) { We have defines for these numbers somewhere in webkit! > Tools/WebKitTestRunner/TestController.cpp:1087 > + // If the URL Response has "Content-Disposition: attachment;" header, then > + // we should download it. keep on one line > Tools/WebKitTestRunner/TestController.cpp:1108 > + // We should ignore downloadable top-level content for subframes, with an exception for text/xml and application/xml so we can still support Acid3 test. > + // It makes the browser intentionally behave differently when it comes to text(application)/xml content in subframes vs. mainframe. the browser? this is the test system, is this behavior even tested?

Comment on attachment 168173 [details] patch View in context: https://bugs.webkit.org/attachment.cgi?id=168173&action=review >> LayoutTests/platform/efl-wk2/http/tests/loading/text-content-type-with-binary-extension-expected.txt:10 >> +Binary data not loaded as text. Maybe PASS. > > Maybe PASS? That is test's expected output :) >> Tools/WebKitTestRunner/TestController.cpp:1081 >> + if (WKURLResponseHTTPStatusCode(response) == 204) { > > We have defines for these numbers somewhere in webkit! in WebCore.. WebCore/HTTPStatusCodes.h. Are we supposed to use WebCore data types in WTR? >> Tools/WebKitTestRunner/TestController.cpp:1087 >> + // we should download it. > > keep on one line sure. >> Tools/WebKitTestRunner/TestController.cpp:1108 >> + // It makes the browser intentionally behave differently when it comes to text(application)/xml content in subframes vs. mainframe. > > the browser? this is the test system, is this behavior even tested? Checking which layout tests should cover it.. Thanks for having a look!

> > + // It makes the browser intentionally behave differently when it comes to text(application)/xml content in subframes vs. mainframe. > > the browser? this is the test system, is this behavior even tested? Furthermore, the response has been already handled by WKBundlePagePolicyClient so no need in that complicetd logic here simple ignoring should be enough.

Created attachment 168200 [details] patch v2 Simplified TestController::decidePolicyForResponse().

Comment on attachment 168200 [details] patch v2 Rejecting attachment 168200 [details] from commit-queue. Failed to run "['/mnt/git/webkit-commit-queue/Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '-..." exit_code: 2 Last 500 characters of output: Kit/chromium/third_party/yasm/source/patched-yasm --revision 154708 --non-interactive --force --accept theirs-conflict --ignore-externals' in '/mnt/git/webkit-commit-queue/Source/WebKit/chromium' 51>At revision 154708. ________ running '/usr/bin/python tools/clang/scripts/update.py --mac-only' in '/mnt/git/webkit-commit-queue/Source/WebKit/chromium' ________ running '/usr/bin/python gyp_webkit' in '/mnt/git/webkit-commit-queue/Source/WebKit/chromium' Updating webkit projects from gyp files... Full output: http://queues.webkit.org/results/14262134

Created attachment 168211 [details] to be landed

Comment on attachment 168211 [details] to be landed Clearing flags on attachment: 168211 Committed r131057: <http://trac.webkit.org/changeset/131057>

All reviewed patches have been landed. Closing bug.

(In reply to comment #14) > (From update of attachment 168211 [details]) > Clearing flags on attachment: 168211 > > Committed r131057: <http://trac.webkit.org/changeset/131057> It caused a regression on Qt-WK2 - https://bugs.webkit.org/show_bug.cgi?id=99152

http/tests/loading/text-content-type-with-binary-extension.htm is showing up as an unexpected pass on the Mac WK2 bots. I am going to remove the failure test expectation for mac-wk2.

Remove the failing expectation for mac-wk2 in http://trac.webkit.org/changeset/141831