Currently, there are three major issues:
1. The Unicode Script itemization code is quite outdated (I believe it hasn't been changed since Unicode 4.0 times, or even longer) and doesn't provide us with good enough results for mixed scripts cases.
2. When the text formatting is used, the script item might be broken by the format range boundaries anywhere, even in the middle of grapheme cluster, which causes the text to be shaped/rendered with artifacts and other glitches, up to making the text completely unreadable.
3. The text capitalization transform is treated like the text formatting option and may cause issue 2 as well. This also prevents implementing the context-sensitive case-changing properly.
|For Gerrit Dashboard: QTBUG-29930|
|77458,4||Improve the Unicode script itemization implementation||stable||qt/qtbase||Status: ABANDONED||+2||0|
|83186,3||Improve the Unicode script itemization implementation||release||qt/qtbase||Status: MERGED||+2||0|
|111596,4||[WIP] Improve the script itemization algorithm to match Unicode 7.0||5.5||qt/qtbase||Status: ABANDONED||-2||0|
|140393,7||Improve the script itemization algorithm to match Unicode 8.0||5.7||qt/qtbase||Status: MERGED||-2||0|