[Webkit-unassigned] [Bug 220081] [CMake] Use Apple's ICU as fallback

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=220081

Yusuke Suzuki <ysuzuki at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #416707|review?                     |review+
              Flags|                            |

--- Comment #22 from Yusuke Suzuki <ysuzuki at apple.com> ---
Comment on attachment 416707
  --> https://bugs.webkit.org/attachment.cgi?id=416707
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=416707&action=review

>>> Source/cmake/OptionsCommon.cmake:103
>>> +option(USE_APPLE_ICU "Use Apple's internal ICU" ${APPLE})
>> 
>> Should this be in OptionsMac.cmake  instead of here?
> 
> It's not about PORT Mac but about OS(Darwin), it affects other ports. For GTK on Darwin, it usually prefers self-compiled version-specified ICU instead of the Apple one, mainly in the case of cross-platform applications keeping their dependencies unified.
> 
> An alternative to option USE_APPLE_ICU is, we search ICU normally first, if not found, then fallback to use the Apple shipped one. I can make another patch if you like this approach.

I think this is OK for the first patch. Then, later, we can put USE_APPLE_ICU to OptionsGTK.cmake / OptionsWPT.cmake etc. to explicitly configure it in each ports. For now, let's put the global status-quo option here since we would like to keep Mac and JSCOnly using AppleICU by-default (while having an option to use user-installed ICU is *AWESOME*, it actually streamlines my Intl work in JSC). Previously, to test whether the given ICU bug is fixed in very new ICU, I was using Linux box. But with this option, I can simply do it in macOS.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20210107/35a82b5d/attachment.htm>