GitLab

Consider the following scenario:
Voice1.lang = "en-us" (all lower case)
Voice2.lang = "en-US" (country code is upper case)
Utterance.lang = "en-us" (all lower case)
The user has set Voice2 as their preferred voice for English.

The old code would choose voice1 as the best voice for the utterance
because voice1.lang == utterance.lang. voice2.lang != utterance.lang
because of the difference in casing. This was leading to issues in
ChromeVox's voice switching behavior, where eSpeak was sometimes being
used, even if the user had set Chrome OS US English as their preferred
voice.

The tangible effect of this change is that ChromeVox voice switching
will respect the user's preferred voices.

voices.
--gtest_filter="TtsControllerTest.TestGetMatchingVoice"'. Also manually
verify the bug no longer repros after the fix.

Fixed: 1121592
AX-Relnotes: ChromeVox voice switching now respects the user's preferred
Test: 'out/cros/content_unittests
Change-Id: I0588b50fbbc4ee7f0b71741b2bb312c9d6d5b14c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2617511
Commit-Queue: Akihiro Ota <akihiroota@chromium.org>
Reviewed-by: David Tseng <dtseng@chromium.org>
Cr-Commit-Position: refs/heads/master@{#842270}