[webkit-reviews] review granted: [Bug 126702] REGRESSION (r160085): check-webkit-style: utf8' codec can't decode byte 0x89 in position 0: invalid start byte : [Attachment 221142] Patch

[bugzilla-daemon at webkit.org]

Alexey Proskuryakov <ap at webkit.org> has granted László Langó
<llango.u-szeged at partner.samsung.com>'s request for review:
Bug 126702: REGRESSION (r160085): check-webkit-style: utf8' codec can't decode
byte 0x89 in position 0: invalid start byte
https://bugs.webkit.org/show_bug.cgi?id=126702

Attachment 221142: Patch
https://bugs.webkit.org/attachment.cgi?id=221142&action=review

------- Additional Comments from Alexey Proskuryakov <ap at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=221142&action=review


Seems fine to land to stop bot failures as an emergency fix. Thanks for looking
into this!

Longer term, I think that the approach could be further improved. We could skip
known non-text files, and we could also handle decoding failures more
gracefully.

In fact, we should probably report an explicit error when there is malformed
utf-8 in actual text files. I do not know what our compilers do in this case,
especially when it's in comments.

> Tools/ChangeLog:8
> +	   Resolve regression and remove a FIXME comment that is already fixed.


The final patch doesn't remove any FIXMEs.

> Tools/Scripts/webkitpy/common/system/filesystem.py:201
> +    def open_text_file_for_reading(self, path, errors='strict'):

It's not good that we are calling a function named "open_text_file_for_reading"
on all files, not only text ones. This should be fixed.