Bug 148164 – [Cocoa] Punctuation near Hindi text is garbled when styled with the system font

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

[Myles C. Maxfield]

[Cocoa] Punctuation near Hindi text is garbled when styled with the system font

[Myles C. Maxfield]

Created attachment 259354 [details]
Patch

[Myles C. Maxfield]

<rdar://problem/22335832>

[Blaze Burg]

Comment on attachment 259354 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=259354&action=review

r=me

If it's not too much a bother, please update changelog to explain why hitting the empty cache affects Japanese language and hindi characters specifically, and not an arbitrary script/language combination.

> Source/WebCore/platform/graphics/cocoa/FontCocoa.mm:96
> +    // FIXME: Unify these two codepaths

Nit: trailing period

[Alexey Proskuryakov]

Comment on attachment 259354 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=259354&action=review

> LayoutTests/fast/text/hindi-system-font-punctuation.html:3
> +This test makes sure that punctuation next to Hindi characters are rendered as expected when the system language is set to Japanese.

Is the test effective? There doesn't seem to be anything Japanese about it.

[Myles C. Maxfield]

Created attachment 259356 [details]
Patch

[Blaze Burg]

Comment on attachment 259356 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=259356&action=review

> Source/WebCore/ChangeLog:13
> +        Consulting with the cache before it is populated causes it to erreneously say that no fonts are system fonts.

erroneously

[Blaze Burg]

(In reply to comment #4)
> Comment on attachment 259354 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=259354&action=review
> 
> > LayoutTests/fast/text/hindi-system-font-punctuation.html:3
> > +This test makes sure that punctuation next to Hindi characters are rendered as expected when the system language is set to Japanese.
> 
> Is the test effective? There doesn't seem to be anything Japanese about it.

Myles, could you add a comment (in the changelog) about this test's dependency on a particular system language? If you have filed a bug for adding that capability to the test harness, please link it.

[Myles C. Maxfield]

Committed r188634: <http://trac.webkit.org/changeset/188634>

[Blaze Burg]

(In reply to comment #8)
> Committed r188634: <http://trac.webkit.org/changeset/188634>

The changelog is glorious. Thank you.