Bug 5411
| Summary: | Sarissa test cases fail in Safari | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Eric Seidel (no email) <eric> |
| Component: | XML | Assignee: | Nobody <webkit-unassigned> |
| Status: | NEW | ||
| Severity: | Normal | CC: | ahmad.saleem792, amonre, ap, bfulgham, ian, manos_lists, rniwa, webdev, webkit-bug-importer |
| Priority: | P4 | Keywords: | InRadar, NeedsReduction |
| Version: | 420+ | ||
| Hardware: | Mac | ||
| OS: | OS X 10.4 | ||
| URL: | http://dev.abiss.gr/sarissa/test/testsarissa.html | ||
Eric Seidel (no email)
Sarissa test cases fail in Safari
Someone will need to take this bug and break each of the cases into individual bugs. However I wanted to
note this in Bugzilla (instead of just adding it to my todo list) as this covers some of the more abscure
(xml related) aspects of the js DOM.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Manos Batsis
hi, i'm the author of sarissa. i've done changes with the help of some mac
owners (its really hard to test without a mac ;-). CVS HEAD should work for
webkit now, with the exception of XSLT/XPath stuff. From what i hear most of it
was about
document.implementation.createDocument("", "", null)
throwing an DOM exception 5. I think it tries to create an element with the
empty string as a name or something.
Alexey Proskuryakov
Most XSLT tests are still failing as of r17902.
Alexey Proskuryakov
The XSLT tests are failing because the document to be transformed simply fails to load. Looks like the reason for this is that XMLDocument.load doesn't work in Sarissa 0.9.7.5 - it references a non-existent _SARISSA_SYNC_NON_IMPLEMENTED variable.
Now, the same problem seems to be present in Firefox (the document also fails to load) - but it somehow applies a transformation to the empty document, so the tests still pass. Also, fixing this problem only resolves 5 of 7 failures.
Alexey Proskuryakov
The remaining two failures are caused by the result having an extra #text node at the end, probably due to a linefeed inserted by libxslt after <hr/>. I'm not sure if it's a bug.
Eric Seidel (no email)
Still fail in 3.0.4
Ahmad Saleem
I am able to find earliest copy of this test suite from Wayback Archive below:
Link - https://web.archive.org/web/20070501141748/http://dev.abiss.gr/sarissa/test/testsarissa.html
Safari 15.6 on macOS 12.5 and Chrome Canary 106 are still failing following test "XSLTProcessorTestCase" but Firefox Nightly 105 do pass all 8 of them.
Just wanted to update the latest results. Thanks!
Radar WebKit Bug Importer
<rdar://problem/97759997>