Bug 124116 - Track-changes Hebrew & Arabic punctuation shifted from correct position on regular text
Summary: Track-changes Hebrew & Arabic punctuation shifted from correct position on re...
Status: RESOLVED FIXED
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Writer (show other bugs)
Version:
(earliest affected)
Inherited From OOo
Hardware: All All
: medium normal
Assignee: Jonathan Clark
URL:
Whiteboard: target:24.8.0
Keywords:
Depends on: 61444
Blocks: Track-Changes RTL-UI Kerning RTL CTL
  Show dependency treegraph
 
Reported: 2019-03-16 10:42 UTC by Eyal Rozenberg
Modified: 2024-08-19 18:41 UTC (History)
1 user (show)

See Also:
Crash report or crash signature:


Attachments
Example manfistation and non-manifestation of this issue (9.61 KB, application/vnd.oasis.opendocument.text)
2019-03-16 10:42 UTC, Eyal Rozenberg
Details
Screenshot of the Arabic lines (13.46 KB, image/png)
2019-03-18 11:40 UTC, ⁨خالد حسني⁩
Details
Screenshot of the document as rendered in LO Writer 24.2.0.2 GTK3 (46.10 KB, image/png)
2024-05-02 21:53 UTC, Eyal Rozenberg
Details
25.2 nightly: Last phrase has box instead of fat7a on final letter (110.50 KB, image/png)
2024-08-04 18:37 UTC, Eyal Rozenberg
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Eyal Rozenberg 2019-03-16 10:42:19 UTC
Created attachment 150012 [details]
Example manfistation and non-manifestation of this issue

Hebrew punctuation mostly overlays the glyph it's applied to - with lines or dots above, below or inside the glyph. This most works fine in LO; and it also works fine when you're in Track Changes mode - the it's simply given a different color and rendered the same way.

But - if you mix the two, things go wrong: If you add a Hebrew punctuation mark in Track Changes to a letter that's not in Track Changes, or the other way around: A non-Track-Changes punctuation mark to a track changes letter - the punctuation mark gets rendered _after_ the letter.

See attachment for two examples - with a Mapiq (=Dagesh) mark and with a Patah.

Note that this does _not_ seem to occur with Arabic, although I haven't checked that thoroughly.
Comment 1 ⁨خالد حسني⁩ 2019-03-17 22:08:12 UTC
The mark placement is broken in Arabic as well, and both issues are most likely another manifestation of bug 61444.
Comment 2 Eyal Rozenberg 2019-03-17 23:54:33 UTC
(In reply to Khaled Hosny (inactive) from comment #1)
> The mark placement is broken in Arabic as well, and both issues are most
> likely another manifestation of bug 61444.

Can you post an example document in which this bug manifests for Arabic text?
Comment 3 ⁨خالد حسني⁩ 2019-03-18 11:39:36 UTC
(In reply to Eyal Rozenberg from comment #2)
> (In reply to Khaled Hosny (inactive) from comment #1)
> > The mark placement is broken in Arabic as well, and both issues are most
> > likely another manifestation of bug 61444.
> 
> Can you post an example document in which this bug manifests for Arabic text?

In the example you attached, the Arabic mark position is different between the first and the next two lines (might not be as visible in all fonts).
Comment 4 ⁨خالد حسني⁩ 2019-03-18 11:40:09 UTC
Created attachment 150045 [details]
Screenshot of the Arabic lines
Comment 5 Eyal Rozenberg 2019-03-18 12:04:13 UTC
(In reply to Khaled Hosny (inactive) from comment #3)
> In the example you attached, the Arabic mark position is different between
> the first and the next two lines (might not be as visible in all fonts).

Yeah, you're right - it is shifted in position, although not as extremely as with the Hebrew. Can you CONFIRM, then?
Comment 6 Xisco Faulí 2019-03-22 11:27:16 UTC
Also reproduced in

LibreOffice 3.3.0 
OOO330m19 (Build:6)
tag libreoffice-3.3.0.4
Comment 7 ⁨خالد حسني⁩ 2019-03-22 19:22:36 UTC
(In reply to Xisco Faulí from comment #6)
> Also reproduced in
> 
> LibreOffice 3.3.0 
> OOO330m19 (Build:6)
> tag libreoffice-3.3.0.4

Yes, it is basically the same thing as bug 61444 which is inherited from OOo. I’m inclined to mark it as duplicate, as I don’t think there is way to fix it without at least fixing bug 61444 first.
Comment 8 Eyal Rozenberg 2019-03-22 20:13:14 UTC
(In reply to Khaled Hosny (inactive) from comment #7)
> Yes, it is basically the same thing as bug 61444 which is inherited from
> OOo. I’m inclined to mark it as duplicate, as I don’t think there is way to
> fix it without at least fixing bug 61444 first.

I'm marking it as dependent for now. It is not 100% clear to me that this will automagically be resolved by a resolution of bug 61444, and anyway it's significant enough to test separately after bug 61444 is resolved.
Comment 9 QA Administrators 2023-12-01 03:15:37 UTC Comment hidden (obsolete)
Comment 10 Eyal Rozenberg 2023-12-02 21:16:18 UTC
Bug still manifests (for Hebrew and for Arabic) with:

Version: 24.2.0.0.alpha1+ (X86_64) / LibreOffice Community
Build ID: 516f800f84b533db0082b1f39c19d1af40ab29c8
CPU threads: 4; OS: Linux 6.5; UI render: default; VCL: gtk3
Locale: he-IL (en_IL); UI: en-US
Comment 11 Eyal Rozenberg 2024-05-02 21:53:49 UTC
Created attachment 193943 [details]
Screenshot of the document as rendered in LO Writer 24.2.0.2 GTK3

Screenshot of the Writer document attachment 150012 [details], rendered with: 

Version: 24.2.0.2 (X86_64) / LibreOffice Community
Build ID: b1fd3a6f0759c6f806568e15c957f97194bbec8f
CPU threads: 4; OS: Linux 6.6; UI render: default; VCL: gtk3
Locale: en-IL (en_IL); UI: en-US

(with document page dimensions narrowed to reduce screenshot width)
Comment 12 Commit Notification 2024-05-22 17:21:30 UTC
Jonathan Clark committed a patch related to this issue.
It has been pushed to "master":

https://git.libreoffice.org/core/commit/ab0a4543cab77ae0c7c0a79feb8aebab71163dd7

tdf#124116 Correct Writer text shaping across formatting changes

It will be available in 24.8.0.

The patch should be included in the daily builds available at
https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More
information about daily builds can be found at:
https://wiki.documentfoundation.org/Testing_Daily_Builds

Affected users are encouraged to test the fix and report feedback.
Comment 13 Eyal Rozenberg 2024-08-04 18:36:05 UTC
(In reply to Commit Notification from comment #12)

Jonathan, when trying the testcase, I get a box character instead of the "fat7a" haraka. Shall I reopen? Open a a separate bug?
Comment 14 Eyal Rozenberg 2024-08-04 18:37:39 UTC
Created attachment 195699 [details]
25.2 nightly: Last phrase has box instead of fat7a on final letter

Version: 25.2.0.0.alpha0+ (X86_64) / LibreOffice Community
Build ID: 01e6e4303e5a9966f102e0357fe0354a2f74a1c4
CPU threads: 4; OS: Linux 6.6; UI render: default; VCL: gtk3
Locale: en-IL (en_IL); UI: en-US
Comment 15 Jonathan Clark 2024-08-05 13:23:22 UTC
(In reply to Eyal Rozenberg from comment #13)
> (In reply to Commit Notification from comment #12)
> 
> Jonathan, when trying the testcase, I get a box character instead of the
> "fat7a" haraka. Shall I reopen? Open a a separate bug?

Hi Eyal,

This is related to font fallback. If I change the last line to use e.g. Noto Sans Arabic, it is rendered correctly. There are other visible aberrations when adding a fatha to the end of the preceding Hebrew lines.

I think it would be best to open another bug to deal with the fallback issue.
Comment 16 Jonathan Clark 2024-08-19 18:41:01 UTC
*** Bug 113134 has been marked as a duplicate of this bug. ***