ded5ed963c3939a2668ce1152ab60efb7bbbb2bd
Characters from complex scripts without dedicated typefaces were getting baseGlyphCount counted twice: once in shapeFontRun() and then a second time during SkScalerContext::getNextContext() chaining. This was a problem for text runs beginning with full-width Latn characters - shapeFontRun() would get the base glyph count for the typeface containing those characters, then realise it had no typeface specified for the Latn script, resetting the typeface to default without resetting the base glyph count along with it. Mojibake full-width Latn characters, ho! The same issue also affected Cherokee text, which this fix also addresses. Bug: 8127795 Bug: 8034351 Change-Id: I058d81cffbbf86426cc3c5aafcc91a76f4fe6b62
cherrypick from ics-mr1 docs: source for nw app Change-Id: If50f407a0e56fa802fe9beedaa650e3a131872b2
…
am
74e8693d: am ddc0a7bd: am 11fa4460: am 9da9a7cd: am e60b6023: Merge "makefile changes for gcm javadocs" into jb-mr1-dev
Description
No description provided
Languages
Java
73.7%
Kotlin
14%
PowerBuilder
5.8%
C++
5.2%
AIDL
1%