644206 - attributes/test_text.html should be loaded as UTF-8

Status: RESOLVED FIXED

(Core :: Disability Access APIs, defect)

Description

[Boris Zbarsky [:bzbarsky]]

The Russian word in "area7" in this test is encoded in UTF-8, but the test file is apparently loaded as ISO-8859-1 or something by the test.  When I changed chrome:// to always be loaded as UTF-8, the test started failing.

Just adding a <meta charset="utf-8"> and fixing the test's expected values should fix this problem.

Comment 1

[Boris Zbarsky [:bzbarsky]]

It looks like text/test_words.html might have similar issues, by the way.

Comment 2

[Marco Zehe (:MarcoZ)]

Alex, can you take this one, since you know what the result should be? My screen reader doesn't speak Russian, and even if it did, I'm sure I wouldn't know if the values are correct :D Thanks!

Comment 3

[alexander :surkov (:asurkov)]

I can do it. Not today though. It should be enough to fix offset. The russian word 'Привет' is 6 characters, while test expects to see 12.

Comment 4

[Marco Zehe (:MarcoZ)]

Thanks, taking it.

Comment 5

[Marco Zehe (:MarcoZ)]

Attachment: Fixing UTF-8 coding, even makes two more tests pass instead of expected-fail.

Comment 6

[Marco Zehe (:MarcoZ)]

Landed on mozilla-central (Mozilla 4.2): http://hg.mozilla.org/mozilla-central/rev/c4e652d029e4