Bug 61819 – nrwt: make webaudio work w/o content-length header

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Dirk Pranke]

nrwt: make webaudio work w/o content-length header

[Dirk Pranke]

Created attachment 95506 [details]
Patch

[Dirk Pranke]

This patch reworks the logic used by the webkit implementations for parsing the output from DRT. Chromium uses a completely different mechanism for audio so it isn't affected.

Testing this revealed related issues with --print trace-everything and the mock_drt implementation that weren't handling audio properly, so those have been rolled into this as well.

[Dirk Pranke]

Created attachment 95517 [details]
restore newlines in DRT output

[Tony Chang]

Comment on attachment 95517 [details]
restore newlines in DRT output

View in context: https://bugs.webkit.org/attachment.cgi?id=95517&action=review

> Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py:319
> +        self._print_baseline(filename, '.txt')
> +        self._print_baseline(filename, '.wav')
> +        self._print_baseline(filename, '.png')

Nit: Maybe make this a for loop (one less line of code)?

> Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:150
> +        if driver_output.text:
> +            self._save_baseline_data(driver_output.text, ".txt",

Nit: Should we just have _safe_baseline_data bail out if the first param is None?

[Dirk Pranke]

Comment on attachment 95517 [details]
restore newlines in DRT output

View in context: https://bugs.webkit.org/attachment.cgi?id=95517&action=review

>> Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py:319
>> +        self._print_baseline(filename, '.png')
> 
> Nit: Maybe make this a for loop (one less line of code)?

Will do.

>> Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:150
>> +            self._save_baseline_data(driver_output.text, ".txt",
> 
> Nit: Should we just have _safe_baseline_data bail out if the first param is None?

Will do. I think it's a bit of a toss-up which way is clearer.

[Dirk Pranke]

Created attachment 95658 [details]
update patch w/ feedback from Tony

[Dirk Pranke]

Committed r87847: <http://trac.webkit.org/changeset/87847>

[Ademar Reis]

Revision r87847 cherry-picked into qtwebkit-2.2 with commit c783730 <http://gitorious.org/webkit/qtwebkit/commit/c783730>