[Libreoffice-bugs] [Bug 104597] Text runs of RTL scripts (e.g. Arabic, Hebrew, Persian) from imported PDF are reversed, PDFIProcessor::mirrorString not behaving

[bugzilla-daemon at bugs.documentfoundation.org]

https://bugs.documentfoundation.org/show_bug.cgi?id=104597

--- Comment #67 from Armin Le Grand <Armin.Le.Grand at me.com> ---
@Kevin, Stuart: It's not really a gap - a pic would say more than 1000 words,
but...
- if next char starts at the same pos (test before) it covers current -> bad
- it's about 'guessing' the width of curr char
- if curr char is tall -> 20% start value may be good
- if curr is wide -> 80% may be good
- so, together, when between that, it may be a good guess
With my current approach we check for 23% to 100% and horizontally which should
be good. Extending to check for less than 80% would be easy, too.

Much better but more complicated would be to use the width of the current char
-> it's not in the matrix, so would need an OutDev, the font, setting it and
asking for width of that char, then using that width to create better - but
still unsharp - bounds to test against. Unsharp due to we do not know if e.g.
using block text layout how big the gaps btw single chars which should be
concatenated really are :-(

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice-bugs/attachments/20210718/657e8371/attachment.htm>