Created attachment 131066 [details] Sample files to demonstrate bug Hello. Justification of Tamil text is broken upon upgrading to LibO 5.3. Please see attached documents demonstrating this. I am using latest download release on Kubuntu Trusty 64 bit and latest Lohit Tamil font from https://fedorahosted.org/lohit/. (Note that Lohit Tamil is one of the most common Tamil fonts found on Linux distributions.)
I can reproduce this with Lohit Tamil but not Noto Sans Tamil, so looks like something in Lohit Tamil is triggering this bug. I notice also that the text selection is off, looks like there is some disassociation between calculated glyph positions and the actual rendering.
*** Bug 105914 has been marked as a duplicate of this bug. ***
Hi Khaled. If you're going to unify a Devanagari and Tamil issue shouldn't you change the title and summary accordingly? Also I hope cursor placement is also implied in your mention of selection because that's off as well.
*** Bug 105934 has been marked as a duplicate of this bug. ***
I think this is might be related to the changes I did to how character widths are calculated in the new layout engine. I suggest people affected by this switch back to the old layout engine until this is fixed. If someone is bisecting this, please make sure the new layout engine is always enable by setting SAL_USE_COMMON_LAYOUT=1 environment variable since it was not always enabled during 5.3 cycle.
I bisected it using SAL_USE_COMMON_LAYOUT=1 and it pointed me to this commit: author Khaled Hosny <khaledhosny@eglug.org> 2016-09-11 08:25:46 (GMT) committer Khaled Hosny <khaledhosny@eglug.org> 2016-10-18 18:41:31 (GMT) commit 15f6a97d9f23124c19471b9d8dd38f14f53829b3 (patch) tree cbcb1cdf11821749d465b726ce2d4957a06fca05 parent 66104ff9856ff1c81ec8c776a0467796f85efc01 (diff) Fix applying DX adjustments in CommonSalLayout By overriding GetCharWidths() and ApplyDXArray() with a simpler and saner implementation. This fixes rendering of Awami Nastaliq, as well as subtending marks in Amiri and potentially other bugs. Breaks Kashida justification, will need to rewrite that one as well. I hope it helps
Thanks Xisco, I suspected it would be this commit.
*** Bug 106567 has been marked as a duplicate of this bug. ***
Still reproducable in LODev 6.0 Version: 6.0.0.0.alpha0+ (x64) Build ID: 311be7d70146e4963e4dfd1f39d6c71d45d241ca CPU threads: 4; OS: Windows 6.19; UI render: default; TinderBox: Win-x86_64@42, Branch:master, Time: 2017-08-24_00:29:42 Locale: zh-CN (zh_CN); Calc: group
*** Bug 109999 has been marked as a duplicate of this bug. ***
*** Bug 114510 has been marked as a duplicate of this bug. ***
*** Bug 107167 has been marked as a duplicate of this bug. ***
*** Bug 105798 has been marked as a duplicate of this bug. ***
Khaled Hosny committed a patch related to this issue. It has been pushed to "master": http://cgit.freedesktop.org/libreoffice/core/commit/?id=2815499d7c0b32fa05fcd697e7b2c2d897f78dfb tdf#105913: Fix applying DX array to ligatures It will be available in 6.1.0. The patch should be included in the daily builds available at http://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: http://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.
Created attachment 141735 [details] Sample files to understand word breaking with Zero Width Joiner Tested using malayalam fonts previously affected with this issue and found solved. But another issue related with this mentioned in https://bugs.documentfoundation.org/show_bug.cgi?id=109999 is still pending. The pending issue related with libre office writer is :- "last word in a line containing 'Zero Width Joiner' breaking with Zero Width Joiner unicode character to next line in a paragraph which contain malayalam Unicode characters" Reproducable in Version: 6.1.0.0.alpha1+ (x64) Build ID: a6a38c6de9c18fd1269fc8cfc0e070ef429c8e2f CPU threads: 4; OS: Windows 10.0; UI render: default; TinderBox: Win-x86_64@42, Branch:master, Time: 2018-04-28_01:58:12 Locale: en-IN (en_IN); Calc: group
(In reply to Ramesh K from comment #16) > Created attachment 141735 [details] > Sample files to understand word breaking with Zero Width Joiner > > Tested using malayalam fonts previously affected with this issue and found > solved. But another issue related with this mentioned in > https://bugs.documentfoundation.org/show_bug.cgi?id=109999 is still pending. > > The pending issue related with libre office writer is :- > > "last word in a line containing 'Zero Width Joiner' breaking with Zero Width > Joiner unicode character to next line in a paragraph which contain malayalam > Unicode characters" > > Reproducable in Version: 6.1.0.0.alpha1+ (x64) > Build ID: a6a38c6de9c18fd1269fc8cfc0e070ef429c8e2f > CPU threads: 4; OS: Windows 10.0; UI render: default; > TinderBox: Win-x86_64@42, Branch:master, Time: 2018-04-28_01:58:12 > Locale: en-IN (en_IN); Calc: group Please open a new issue and attach a document that shows the issue, as well as screenshot showing the expected result.
Verified in Version: 6.1.0.0.alpha1+ Build ID: 43af818557904450b13839350c65ad865b9ee9d2 CPU threads: 4; OS: Linux 4.13; UI render: default; VCL: gtk3; Locale: ca-ES (ca_ES.UTF-8); Calc: group @Khaled, should it be backported to LibreOffice 6 ?
(In reply to Xisco Faulí from comment #18) > Verified in > > Version: 6.1.0.0.alpha1+ > Build ID: 43af818557904450b13839350c65ad865b9ee9d2 > CPU threads: 4; OS: Linux 4.13; UI render: default; VCL: gtk3; > Locale: ca-ES (ca_ES.UTF-8); Calc: group > > @Khaled, should it be backported to LibreOffice 6 ? I don’t think it can be easily backported, it depends on several other changes.
*** Bug 112093 has been marked as a duplicate of this bug. ***
It seems to me that latin ligatures are still somewhat problematic. To help with debugging this issue I created a GitHub project with a sample file and some tools to quickly check against a new LOO version: https://github.com/schlomo/libreoffice-bug-rtl-justified-regression There is also the word "different" in this document and the ligature "ff" is not showing correctly on LOO after 5.2.7.2. Can this also be related to this bug? Is this a different issue?
(In reply to Schlomo Schapiro from comment #21) > It seems to me that latin ligatures are still somewhat problematic. To help > with debugging this issue I created a GitHub project with a sample file and > some tools to quickly check against a new LOO version: > https://github.com/schlomo/libreoffice-bug-rtl-justified-regression > > There is also the word "different" in this document and the ligature "ff" is > not showing correctly on LOO after 5.2.7.2. > > Can this also be related to this bug? Is this a different issue? Hi Schlomo Schapiro, Please create a follow-up bug for this as the original problem of this issue is fixed.
OK, sure. I did some checks and ligatures seem fine. Sorry for the false alarm. The Liberation Sans font seems to be quite bad with regard to ligatures, with other fonts it looks much better.
*** Bug 115854 has been marked as a duplicate of this bug. ***