[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 #60 from Kevin Suo <suokunlong at 126.com> ---
(In reply to Armin Le Grand from comment #59)
Thank you for your suggested patch. I checked, but it does not work for some
chars.

For instance:

On The current master there is a test file named "testFontFeatures.pdf" in
sdext/source/pdfimport/test/testdocs. This file is used in the
CppunitTest_sdext_pdfimport (test case file is in
sdext/source/pdfimport/test/tests.cxx, the test name is testFontFeatures).

If you run 'make CppunitTest_sdext_pdfimport', you can see the output in
workdir/CppunitTest/sdext_pdfimport.test.log.

>From the log, you can see in line 384 that, for the line "Times New Roman
Normal" as shown in the PDF, the chars "Tim" are concatenated, but "e" and "s"
(as in "Times" are not; the chars "Ne" (as in "New) are concatenated, but "w"
is not.
(please note that in the log file the concatenated "Tim" are shown in three
different lines because there is another problem in the code which added
new-line characters in the output, but this does not have impact on this RTL
bug, thus can be addressed separately)

Also, from the log file you can also see from line 691 that all those Chinese
characters are not concatenated.

Could you please take a look and revise your patch?

-- 
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/dbb8da45/attachment-0001.htm>