Re: Unclear string to translate in inspectorproperties.hrc
Thank you, Mike & Shivam, for explanation. We should probably translate the strings in a similar cryptic way. The inspector pane is not displayed (I have the most recent build), so I'll wait for a fix. thanks Milos On 8/12/20 12:04 PM, Mike Kaganski wrote: > Hi, > > On 12.08.2020 12:52, Milos Sramek wrote: >> we got for translation 200+ new strings, which are located in file >> inspectorproperties.hrc and seem to be related to the Style Inspector >> sidebar deck. >> >> These strings are not full sentences and seem to have a certain logic >> structure, for example >> >> Fill Bitmap Rectangle Point >> Char Interoperability Grab Bag >> Char Weight Complex >> >> So, we do not know, how to translate them. Should they be translated at all? >> > Basically answered in i10n mailing list; citing here: > > >> Some background. >> >> These strings are related to the Styles Inspector - a GSoC project, >> being implemented by Shivam Kumar Singh >> (https://bugs.documentfoundation.org/show_bug.cgi?id=134554). The >> inspector is already available in Experimental mode (see comment 2 in >> the meta issue). >> >> The Inspector presents a detailed low-level view of properties applied >> at the cursor. As such, it shows the multitude of attributes applied to >> text (or other object in the future); most these attributes have their >> names in multiple dialogs throughout the UI, but this is the first time >> when all of them need to be named in a single place. And translation in >> dialogs can not be used directly, e.g. because the translation there in >> dialogs may use context absent in the Inspector. >> >> One possible improvement could be to provide a hint somehow near the IDs >> in inspectorproperties.hrc, where known/possible, to the related strings >> in the dialogs/other UI parts, to help translators understand what this >> item is about. But that's a separate task/TODO: the current work on >> Inspector does not consider dialogs; the Inspector is created based on >> the properties available using UNO API, and thus Shivam does not know >> where the items are located in existing UI. (This is to clarify that >> there's no information already used by developers, potentially useful to >> translation, that was somehow lost during development.) >> >> The names that Shivam has put there are possibly somewhat "cryptic" >> exactly because of the mentioned specifics: the names are mostly the API >> names, which were split by CamelCase to produce strings like "Char >> Weight Complex", intended to be used in context of table like >> >> Char Interoperability Grab Bag"some string" >> >> Char Weight Complex10 >> >> So the mentioned TODO would also allow to gradually find the existing >> names of the attributes, and make them consistent in the Inspector. > > > Milos Sramek wrote: >> Currently, I can see only two items in Style Inspector sidebar deck in >> Writer >> Font name >> Font size >> >> In the file inspectorproperties.hrc there is >> "Char Font Name" >> but no >> "Char Font Size" >> >> Are the strings, which i can see in the Inspector, derived somehow from >> the strings in inspectorproperties.hrc? >> > You obviously look somewhere else. The Inspector looks like this: > https://imgur.com/xu2Wvh2 > You might need to move cursor/type something for the initially empty > deck to initialize (a bug). > > Note that there are exactly "Char Font Name" and "Char Font Height", but > no "Font name" or "Font size" currently. > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- Milos Sramek, msrame...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Unclear string to translate in inspectorproperties.hrc
Hi, we got for translation 200+ new strings, which are located in file inspectorproperties.hrc and seem to be related to the Style Inspector sidebar deck. These strings are not full sentences and seem to have a certain logic structure, for example Fill Bitmap Rectangle Point Char Interoperability Grab Bag Char Weight Complex So, we do not know, how to translate them. Should they be translated at all? Currently, I can see only two items in Style Inspector sidebar deck in Writer Font name Font size In the file inspectorproperties.hrc there is "Char Font Name" but no "Char Font Size" Are the strings, which i can see in the Inspector, derived somehow from the strings in inspectorproperties.hrc? Thanks Milos On 8/4/20 7:59 AM, Shivam Kumar Singh wrote: > Hi, > > Work Done this week :- > * tdf#135178 tdf#135179 tdf#134820 SetChgLink was not correctly called > and used which led to crash on moving the cursor to different images > or charts. This issue was fixed, MERGED > * tdf#134831 The SI would expand only to 2/3rd of the sidebar though > its expected to expand to the whole length. This change was fixed and > merged. > * tdf#134560 tdf#135107 Previously the API names for the properties > were shown in the SI, which was not very easy to understand. These > properties are now human-readable and translatable wrt UI language. > MERGED > > To Do for this week :- > * Add color boxes to "Color" properties tdf#134561 > * Some ui issues still exist relating to the scroll bars. This should > be fixed within this week tdf#135344 > * Add a way to also display properties which are stored in structs in > the UNO API tdf#135406 > > All the bugs are being listed at META tdf#134554, It would be great > help if more number of people use the SI and report any bugs > experienced or confirm the already listed unconfirmed ones. > > Thank you Mike, Heiko and Tomaz for their continuous guidance :) > > sincerely, > Shivam Kumar Singh > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- Milos Sramek, msrame...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Update of sk_SK dictionary
Hi, I just thought that "refs/for/master" is just a placeholder (it looks like one) and that I should write something there and I did not know what. But Andras has pushed it already, so it's done thanks Milos On 6/10/20 11:48 AM, Christian Lohmaier wrote: > On Wed, Jun 10, 2020 at 11:42 AM Milos Sramek wrote: >> ! [remote rejected] sk-update -> sk-update (prohibited by Gerrit: not >> permitted: create) > you tried to create a new branch. you need to push to refs/for/master instead. > > like mentioned previously: >> git push origin master:refs/for/master > As in your case you did your work on a local branch "sk-update", you should > use > git push origin sk-update:refs/for/master > > ciao > Christian -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Update of sk_SK dictionary
Hi Andras, it looks like I do not have the required rights. So, maybe that it would be more efficient if you patch and push the dictionaries (the required link is http://sk-spell.sk.cx/files/sk_SK-20200510.patch.gz). I've checked the patch on an installed 6.4, its was OK. I've tried to push directly to master and also to push a branch. In both cases git complained about insufficient rights, for example with a branch sk-update: milos:~/sbox/git/dictionaries$ git push --set-upstream origin sk-update Enumerating objects: 15, done. Counting objects: 100% (15/15), done. Delta compression using up to 4 threads Compressing objects: 100% (8/8), done. Writing objects: 100% (8/8), 420.88 KiB | 2.65 MiB/s, done. Total 8 (delta 6), reused 0 (delta 0) remote: Resolving deltas: 100% (6/6) remote: error: branch refs/heads/sk-update: remote: You need 'Create' rights to create new references. remote: User: milossramek2 remote: Contact an administrator to fix the permissions remote: Processing changes: refs: 1, done To ssh://gerrit.libreoffice.org/dictionaries ! [remote rejected] sk-update -> sk-update (prohibited by Gerrit: not permitted: create) error: failed to push some refs to 'ssh://milossram...@gerrit.libreoffice.org/dictionaries' Milos On 6/7/20 8:41 PM, Andras Timar wrote: > Hi Milos, > > There may be multiple solutions. Here is mine: > > You did right when you checked out dictionaries repo. But instead of > anongit.freedesktop.org <http://anongit.freedesktop.org> I would have > used git://gerrit.libreoffice.org/dictionaries > <http://gerrit.libreoffice.org/dictionaries> > There you make your commit. Then: > > git push origin master:refs/for/master > > This command pushes the patch to gerrit. logerrit script does the same > IMHO. I hope it will work for you, but I never know, because I have > rights to do a lot of things. > > Best regards, > Andras > > On Sun, Jun 7, 2020 at 5:54 PM Milos Sramek <mailto:sramek.mi...@gmail.com>> wrote: > > Hi, > I've tried to find instructions on how to commit something to > dictionaries. > > I've checked out 'core', but the 'dictionaries' directory is empty > there. I've read > https://wiki.documentfoundation.org/Development/Submodules, but I > did not understand anything. > > Then I checked out > git://anongit.freedesktop.org/libreoffice/dictionaries > <http://anongit.freedesktop.org/libreoffice/dictionaries> and > followed > https://wiki.documentfoundation.org/Development/gerrit/SubmitPatch. > There is, however no the 'logerrit' script. Should I use the > version from core? Or should clone to dictionaries to > core.dictionaries? Should I call ../logerrit in the disctionaries > directory of ./logerrit in core? > > thanks for help > Milos > > > > On 6/6/20 10:10 PM, Milos Sramek wrote: >> thank you, Zdenko! >> I'll try submit it >> Milos >> >> On 6/6/20 9:52 AM, ZdPo Ster wrote: >>> Hello, >>> >>> users complain to me that Libreoffice Slovakian dictionaries are >>> very old. >>> Can somebody download and apply patch from >>> http://sk-spell.sk.cx/files/sk_SK-20200510.patch.gz ? >>> >>> I am not interested in setting up gerrit or any other system for >>> time to time update of dictionaries. >>> Thanks for understanding. >>> >>> Zdenko >>> >>> >>> ___ >>> LibreOffice mailing list >>> LibreOffice@lists.freedesktop.org >>> <mailto:LibreOffice@lists.freedesktop.org> >>> https://lists.freedesktop.org/mailman/listinfo/libreoffice >> >> >> -- >> Milos Sramek, msrame...@gmail.com <mailto:msrame...@gmail.com> >> >> ___ >> LibreOffice mailing list >> LibreOffice@lists.freedesktop.org >> <mailto:LibreOffice@lists.freedesktop.org> >> https://lists.freedesktop.org/mailman/listinfo/libreoffice > > > -- > email & jabber: sramek.mi...@gmail.com <mailto:sramek.mi...@gmail.com> > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > <mailto:LibreOffice@lists.freedesktop.org> > https://lists.freedesktop.org/mailman/listinfo/libreoffice > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Update of sk_SK dictionary
Hi, I've tried to find instructions on how to commit something to dictionaries. I've checked out 'core', but the 'dictionaries' directory is empty there. I've read https://wiki.documentfoundation.org/Development/Submodules, but I did not understand anything. Then I checked out git://anongit.freedesktop.org/libreoffice/dictionaries and followed https://wiki.documentfoundation.org/Development/gerrit/SubmitPatch. There is, however no the 'logerrit' script. Should I use the version from core? Or should clone to dictionaries to core.dictionaries? Should I call ../logerrit in the disctionaries directory of ./logerrit in core? thanks for help Milos On 6/6/20 10:10 PM, Milos Sramek wrote: > thank you, Zdenko! > I'll try submit it > Milos > > On 6/6/20 9:52 AM, ZdPo Ster wrote: >> Hello, >> >> users complain to me that Libreoffice Slovakian dictionaries are very >> old. >> Can somebody download and apply patch from >> http://sk-spell.sk.cx/files/sk_SK-20200510.patch.gz ? >> >> I am not interested in setting up gerrit or any other system for time >> to time update of dictionaries. >> Thanks for understanding. >> >> Zdenko >> >> >> ___ >> LibreOffice mailing list >> LibreOffice@lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/libreoffice > > > -- > Milos Sramek, msrame...@gmail.com > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Update of sk_SK dictionary
thank you, Zdenko! I'll try submit it Milos On 6/6/20 9:52 AM, ZdPo Ster wrote: > Hello, > > users complain to me that Libreoffice Slovakian dictionaries are very > old. > Can somebody download and apply patch from > http://sk-spell.sk.cx/files/sk_SK-20200510.patch.gz ? > > I am not interested in setting up gerrit or any other system for time > to time update of dictionaries. > Thanks for understanding. > > Zdenko > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- Milos Sramek, msrame...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
LibreOffice booth at conference in Bratislava
Hi, we are looking for help at a LibreOffice booth at Open Source Weekend (June 17-18, http://ossvikend.sk/) in Bratislava, Slovakia. Maybe there is somebody on this list from or not far from Bratislava. The problem is that I am the organizer of the event and have no time to be in the booth. Without help we will not be able to organize it :( best Milos -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: testing commonsallayout branch
Hi, you can find results of my testing in http://bender.dam.fmph.uniba.sk/~milos/roundtrip/layouttext/rslt-web-mod.ods I've tested 149 files obtained from the officeshots repository at http://dav.officeshots.org/testsuites/. I've tested the last 5.2 released version and the master version on Windows. Linux and Mac.In the table they are separated by a black vertical bar. Since there is no gold standard, I took the 5.2 version on Linux as a reference to which I've compared all other versions. Maybe that a different one should have been taken as the reference. The results are sorted, with the worst graded (largest differences) on top. To see the difference it is best to view the document overlay in the second column (labeled as THE). The first column (LND) shows the documents side-by-side, where just large differences are visible. 3-rd and 4-th columns show details. Observations: Rendering on MAC differs from that on Linux more than rendering on Windows. This seems to be partially caused by a slightly large line spacing on Mac. On Win and Lin letter 'g' touches table lines below which changes number of detected lines in the document. Examples: lines 8 and 9. It is unclear if the Mac version is OK or the Win Another reason: incorrect drop caps in master on Mac, lines 3, 7 and 9 Improvements: lines 12, 13, 16, 17, 26-29, only on Windows, on Mac and LINUX were OK before Line 18: something has changed in master, which applies to all platforms. It is not related to the layout engine I think that each difference should be analyzed separately - sometimes it may be caused by the layout engine, but sometimes perhaps by something different. I will be glad to test a different set of files, which may be eventually more appropriate. If interpretation of the results in unclear, please ask. best Milos On 2016-10-25 01:15, Khaled Hosny wrote: > Hi, > > I think the daily builds would be the better as they should have any new > fixes: > http://dev-builds.libreoffice.org/daily/master/ > > To enable the new layout you need to set SAL_USE_COMMON_LAYOUT > environment variable, to disable it you need to unset the variable > (setting it to 0 will NOT disable it as the code just check the presence > of the variable not its value). > > Regards, > Khaled > > On Mon, Oct 24, 2016 at 01:40:19PM +0200, Milos Sramek wrote: >> Hi, >> >> to be sure that I do everything right: >> >> The master build should be downloaded from >> http://dev-builds.libreoffice.org/pre-releases/win/x86_64/ >> http://dev-builds.libreoffice.org/pre-releases/mac/x86_64/ >> http://dev-builds.libreoffice.org/pre-releases/deb/x86_64/ >> >> The builds are from October 21-st - are they already the right ones? >> >> How to select the layout code? Which environment settings to use? >> >> best >> Milos >> >> On 2016-10-24 00:30, Khaled Hosny wrote: >>> Hi, >>> >>> Things I’d like to test at this point: >>> * master with and without the new layout code activated on the three >>> platforms. To see if we have any regressions or, hopefully, progressions. >>> * master with the new layout code across platforms (to see if we really >>> getting identical results as we hope). >>> >>> Regards, >>> Khaled >>> >>> On Fri, Oct 21, 2016 at 09:18:01PM +0200, Milos Sramek wrote: >>>> Hi, >>>> >>>> >>>> I've changed subject, since the discussion went in a different direction. >>>> >>>> Meanwhile I've discovered, that one can run OSX in a virtual machine :). >>>> That fits to my environment well. >>>> >>>> I've tested a few random files with a few random version on Linux, Mac >>>> and Windows. The results: >>>> http://bender.dam.fmph.uniba.sk/~milos/roundtrip/odttest-mac-win/rslt-web.ods >>>> >>>> Brief explanantion: >>>> grade 0: pixel level identical >>>> grade 1: some differences, probably not visible >>>> ... >>>> Grade meaning is explained in a comment >>>> >>>>> is a link to pdf: >>>> column LND: side by side >>>> column THE: overlay >>>> column HLPE: vertically aligned lines >>>> column FDE: vertically and horizontally aligned >>>> >>>> first sheet: print test: printed by tested version (name in header) and >>>> in reference version (LO43 on Linux) and compared >>>> second sheet: roundtrip test: saved to odt and printed by the reference >>>> version. No differences here >>>> >>>> Which versions should I test? The engines can be specified by >>>
Re: testing commonsallayout branch
Hi, to be sure that I do everything right: The master build should be downloaded from http://dev-builds.libreoffice.org/pre-releases/win/x86_64/ http://dev-builds.libreoffice.org/pre-releases/mac/x86_64/ http://dev-builds.libreoffice.org/pre-releases/deb/x86_64/ The builds are from October 21-st - are they already the right ones? How to select the layout code? Which environment settings to use? best Milos On 2016-10-24 00:30, Khaled Hosny wrote: > Hi, > > Things I’d like to test at this point: > * master with and without the new layout code activated on the three > platforms. To see if we have any regressions or, hopefully, progressions. > * master with the new layout code across platforms (to see if we really > getting identical results as we hope). > > Regards, > Khaled > > On Fri, Oct 21, 2016 at 09:18:01PM +0200, Milos Sramek wrote: >> Hi, >> >> >> I've changed subject, since the discussion went in a different direction. >> >> Meanwhile I've discovered, that one can run OSX in a virtual machine :). >> That fits to my environment well. >> >> I've tested a few random files with a few random version on Linux, Mac >> and Windows. The results: >> http://bender.dam.fmph.uniba.sk/~milos/roundtrip/odttest-mac-win/rslt-web.ods >> >> Brief explanantion: >> grade 0: pixel level identical >> grade 1: some differences, probably not visible >> ... >> Grade meaning is explained in a comment >> >>> is a link to pdf: >> column LND: side by side >> column THE: overlay >> column HLPE: vertically aligned lines >> column FDE: vertically and horizontally aligned >> >> first sheet: print test: printed by tested version (name in header) and >> in reference version (LO43 on Linux) and compared >> second sheet: roundtrip test: saved to odt and printed by the reference >> version. No differences here >> >> Which versions should I test? The engines can be specified by >> environment variable - I can test that too. >> >> Is there a way how to install multiple versions in parallel on MAC? I >> can install only production and development one. >> >> best >> Milos >> >> >> On 2016-10-18 18:36, Khaled Hosny wrote: >>> On Tue, Oct 18, 2016 at 09:06:42AM +0200, Milos Sramek wrote: >>>> Dear Khaled, >>>> >>>> do you think that extensive comparison of rendering of numerous >>>> documents with and without the new layout engine makes sense? >>>> If yes, I can run my tests in such a way to see if there are any >>>> differences. >>> This would be extremely helpful, and more so if it can run on Windows or >>> Mac. >>> >>>> My test documents use latin script except for a few ones in Chinese. So, >>>> if you happen to have sume documents in Arabic I can add them to my test >>>> set. >>> I’ve a random collection of documents somewhere, I’ll clean it up and >>> send it to you. >>> >>> Regards, >>> Khaled >>> >>> >>> ___ >>> LibreOffice mailing list >>> LibreOffice@lists.freedesktop.org >>> https://lists.freedesktop.org/mailman/listinfo/libreoffice >> >> -- >> email & jabber: sramek.mi...@gmail.com >> > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
testing commonsallayout branch
Hi, I've changed subject, since the discussion went in a different direction. Meanwhile I've discovered, that one can run OSX in a virtual machine :). That fits to my environment well. I've tested a few random files with a few random version on Linux, Mac and Windows. The results: http://bender.dam.fmph.uniba.sk/~milos/roundtrip/odttest-mac-win/rslt-web.ods Brief explanantion: grade 0: pixel level identical grade 1: some differences, probably not visible ... Grade meaning is explained in a comment > is a link to pdf: column LND: side by side column THE: overlay column HLPE: vertically aligned lines column FDE: vertically and horizontally aligned first sheet: print test: printed by tested version (name in header) and in reference version (LO43 on Linux) and compared second sheet: roundtrip test: saved to odt and printed by the reference version. No differences here Which versions should I test? The engines can be specified by environment variable - I can test that too. Is there a way how to install multiple versions in parallel on MAC? I can install only production and development one. best Milos On 2016-10-18 18:36, Khaled Hosny wrote: > On Tue, Oct 18, 2016 at 09:06:42AM +0200, Milos Sramek wrote: >> Dear Khaled, >> >> do you think that extensive comparison of rendering of numerous >> documents with and without the new layout engine makes sense? >> If yes, I can run my tests in such a way to see if there are any >> differences. > This would be extremely helpful, and more so if it can run on Windows or > Mac. > >> My test documents use latin script except for a few ones in Chinese. So, >> if you happen to have sume documents in Arabic I can add them to my test >> set. > I’ve a random collection of documents somewhere, I’ll clean it up and > send it to you. > > Regards, > Khaled > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Merging feature/commonsallayout branch
On 2016-10-18 18:36, Khaled Hosny wrote: > On Tue, Oct 18, 2016 at 09:06:42AM +0200, Milos Sramek wrote: >> Dear Khaled, >> >> do you think that extensive comparison of rendering of numerous >> documents with and without the new layout engine makes sense? >> If yes, I can run my tests in such a way to see if there are any >> differences. > This would be extremely helpful, and more so if it can run on Windows or > Mac. I can run the tests on Linux and Windows. I do not have Mac, but I think that there should not be a problem with that. On windows or mac one just converts/prints files by means of a script - they are then compared on a system with running Python - in my case a Linux machine. I run Windows in a virtual machine, which has access to the directory, where all data and scripts reside. So I even do not have to copy files. I've checked the web and see that it is possible to install MAC on virtual box: https://techsviewer.com/how-to-install-mac-os-x-el-capitan-on-pc-on-virtualbox/ I'll try it best milos > >> My test documents use latin script except for a few ones in Chinese. So, >> if you happen to have sume documents in Arabic I can add them to my test >> set. > I’ve a random collection of documents somewhere, I’ll clean it up and > send it to you. > > Regards, > Khaled > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Merging feature/commonsallayout branch
Dear Khaled, do you think that extensive comparison of rendering of numerous documents with and without the new layout engine makes sense? If yes, I can run my tests in such a way to see if there are any differences. My test documents use latin script except for a few ones in Chinese. So, if you happen to have sume documents in Arabic I can add them to my test set. best Milos On 2016-10-17 21:30, Khaled Hosny wrote: > I believe that feature/commonsallayout (AKA unified text layout) is now > feature complete with no known major bugs, and should be ready to be > merged on master. I’ll try merge it tomorrow night and hope for the > best, unless someone objects loudly. > > Currently the new layout is off by default and can be enabled at runtime > by setting SAL_USE_COMMON_LAYOUT env variable. After merging with > master, I’m going to wait a week or so for any potential build issues > then swap the default. > > There are too main issues with the new code: > - Type 1 fonts are not supported. They can be supported with some > effort, but Type 1 fonts have been obsolete for more than 15 years and > I’d like to use this opportunity to drop support for them and cleanup > some of the ugly code we have. > - We use a bit of DirectWrite to load fonts on Windows, so Windows XP is > not supported as well. Again it can be fixed with some effort, but I > don’t think anyone will miss XP. Ideally we should do a full switch to > DirectWrite and modernise our Windows font rendering, but that is > another story. > > Regards, > Khaled > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- Milos Sramek, msrame...@gmail.com signature.asc Description: OpenPGP digital signature ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
[Libreoffice-commits] core.git: writerfilter/inc
writerfilter/inc/dmapper/resourcemodel.hxx |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) New commits: commit 5c71ae07229b1a76b6ad54c485c8cde55ea309e5 Author: Milos Sramek <milos.sra...@soit.sk> Date: Mon Sep 19 17:13:52 2016 +0200 Explanation of the SPRM acronym added to definition of the Sprm class Change-Id: I536cd6ba2d531292d5148819dfd914b76f4b Reviewed-on: https://gerrit.libreoffice.org/29032 Reviewed-by: Katarina Behrens <katarina.behr...@cib.de> Tested-by: Katarina Behrens <katarina.behr...@cib.de> diff --git a/writerfilter/inc/dmapper/resourcemodel.hxx b/writerfilter/inc/dmapper/resourcemodel.hxx index 043c160..252d833 100644 --- a/writerfilter/inc/dmapper/resourcemodel.hxx +++ b/writerfilter/inc/dmapper/resourcemodel.hxx @@ -347,7 +347,7 @@ public: }; /** - An SPRM. + An SPRM: Section, Paragraph and Run Modifier */ class Sprm ___ Libreoffice-commits mailing list libreoffice-comm...@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice-commits
Re: How debug OOXML import / export?
Thank you! m On 2016-08-23 03:21, Mark Hung wrote: > Hi, > > For import, it's in writerfilter/source/dmapper . > > For export, it's in sw/source/filter/ww8/docxattributeout.cxx > > 2016年8月23日 01:21於 "Milos Sramek" <sramek.mi...@gmail.com>寫道： > >> Hi, >> >> I think I've found a bug in OOXML export/import (related to kerning) and >> I would like to look at it. However, I have no idea how to start. I've >> spent several hours by looking for the related code with no success. >> >> I assume there is a place where the OOXML code is converted in internal >> representation on load and vice versa on save. Or, maybe there is a >> related howto somewhere, which I failed to find. >> >> Please help. >> >> Milos >> >> >> >> >> -- >> Milos Sramek, msrame...@gmail.com >> >> >> >> ___ >> LibreOffice mailing list >> LibreOffice@lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/libreoffice >> >> > > > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- Milos Sramek, msrame...@gmail.com signature.asc Description: OpenPGP digital signature ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
How debug OOXML import / export?
Hi, I think I've found a bug in OOXML export/import (related to kerning) and I would like to look at it. However, I have no idea how to start. I've spent several hours by looking for the related code with no success. I assume there is a place where the OOXML code is converted in internal representation on load and vice versa on save. Or, maybe there is a related howto somewhere, which I failed to find. Please help. Milos -- Milos Sramek, msrame...@gmail.com signature.asc Description: OpenPGP digital signature ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Help with an old regression in text rendering needed
Hi, thank you for your response and for pointing me towards kerning. It looks like the problem resides in kerning and not in harfbuzz. LO ignores setting of kerning in docx documents: disregarding its setting in the input document the setting (Character->Position->Pair Kerning) is always ON and when saved it is in the output file always OFF (verified in MSOffice). This has been so in LO since version 3.3 (verified right now) and is still so in the most recent one. Therefore, roundtrip tests (open docx in LO, save, open saved on MSO and compare with the original docx) with documents with kerning set to OFF always pass and documents with kerning set to ON always fail. The situation is different in print test (open original and print by LO and compare with printed by MSO). Bibisection shows that up to revision 59a69f7 in the bibisect-43all repository documents with Kerning ON fail and those with kerning OFF pass. >From the bibisection log: * 59a69f7 tag: source-hash-48ad2f61fe71edc1a8967b322d3e0f368f4be06f, refs/bisect/bad) * 96f14f1 -tag: source-hash-f0393d7ff69011a16b100541ef18e5090544e4a1, refs/bisect/good-96f14f1bf7505b59a2731dcba1615568f03fd68f I wanted to verify this by compiling the sources (I've tried f0393d7ff69011a16b100541ef18e5090544e4a1), but compilation failed in unit tests. I have used gcc-4.8 and switches from the file autogen.log: CC=/usr/bin/gcc-4.8 CXX=/usr/bin/g++-4.8 ./autogen.sh --disable-dependency-tracking --disable-linkoo --disable-option-checking --without-junit --without-help --without-myspell-dicts --without-doxygen --disable-gnome-vfs --disable-odk --without-system-jpeg Can I switch the unit test off somehow? Maybe I do something wrong, since the first bad commit: https://github.com/LibreOffice/core/commit/48ad2f61fe71edc1a8967b322d3e0f368f4be06f?diff=split does not seem to be related to docx at all :( My results and also files with kerning switched ON and OFF are attached. Thank in advance for your help best Milos On 2016-07-29 14:09, Caolán McNamara wrote: > On Fri, 2016-07-29 at 10:34 +0200, Milos Sramek wrote: >> Thank you, Caolán, for your suggestions. >> >> >> The first thing I've tried was to compile the current version with >> and without harfbuzz using >> ./autogen.sh --with-harfbuzz=no >> The result was the same (bad). >> >> I am, however, not sure if harfbuzz was really switched off > Yeah, you can't turn off harfbuzz under Linux, the option is only to > enable turning it *on* for other platforms since it became the only > implementation under Linux. > >> I've tried to investigate it bit deeper: >> >> revision mentioned: >> bff8fa9 - (2013-05-06 11:08:29 +0200) Enable HarfBuzz by default - >> Khaled Hosny >> >> The last good revision according to my bibisection >> f0393d7 - (2013-05-06 16:54:53 +0200) [harfbuzz] Fix text width >> calculation, 3rd try - Khaled Hosny > Note that the very next commit after f0393d7 is actually bff8fa9, so > you can probably short circuit bisecting and jump straight > between bff8fa9 and f0393d7 > >> The first bad revision according to my bibisection >> 48ad2f6 - (2013-05-06 12:48:23 +0200) fdo#46808, Convert some code to >> getProcessComponentContext >> >> So, they really are from the same day. > >> I unfortunately cannot compile these revisions, maybe because my >> compiter is too new (gcc version 5.4.0 20160609 (Ubuntu >> 5.4.0-6ubuntu1~16.04.1)). I happens because of std::auto_ptr: >> /workdir/unxlngx6.pro/UnpackedTarball/boost/boost/smart_ptr/scoped_pt >> r.hpp:66:31: >> warning: ‘template class std::auto_ptr’ is deprecated >> [-Wdeprecated-declarations] > That's only a warning. So unless you have used --enable-werror to makes > warnings fatal then that warning shouldn't be a problem. But seeing as > f0393d7 built, presumably bff8fa9 (because its the next commit) builds > as well for comparison. > > C. > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com File999-results.ods Description: application/vnd.oasis.opendocument.spreadsheet File_999l-kernon.docx Description: MS-Word 2007 document File_999l-kernoff.docx Description: MS-Word 2007 document ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
Re: Help with an old regression in text rendering needed
Thank you, Caolán, for your suggestions. The first thing I've tried was to compile the current version with and without harfbuzz using ./autogen.sh --with-harfbuzz=no The result was the same (bad). I am, however, not sure if harfbuzz was really switched off, since in config.log I can see $ ./configure --with-harfbuzz=no --srcdir=/mnt/data/milos/sbox/libreoffice.git --enable-option-checking=fatal but later configure:30367: checking which harfbuzz to use configure:30468: result: internal and later also -I/usr/include/harfbuzz -lharfbuzz -lharfbuzz-icu I've tried to investigate it bit deeper: revision mentioned: bff8fa9 - (2013-05-06 11:08:29 +0200) Enable HarfBuzz by default - Khaled Hosny The last good revision according to my bibisection f0393d7 - (2013-05-06 16:54:53 +0200) [harfbuzz] Fix text width calculation, 3rd try - Khaled Hosny The first bad revision according to my bibisection 48ad2f6 - (2013-05-06 12:48:23 +0200) fdo#46808, Convert some code to getProcessComponentContext So, they really are from the same day. I unfortunately cannot compile these revisions, maybe because my compiter is too new (gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.1)). I happens because of std::auto_ptr: /workdir/unxlngx6.pro/UnpackedTarball/boost/boost/smart_ptr/scoped_ptr.hpp:66:31: warning: ‘template class std::auto_ptr’ is deprecated [-Wdeprecated-declarations] How should I use the -Wdeprecated-declarations switch? thanks Milos On 2016-07-20 18:27, Caolán McNamara wrote: > On Wed, 2016-07-20 at 08:55 +0200, Milos Sramek wrote: >> Hi, >> >> In Brno I would like to present my tools for automated detection and >> bibisection of interoperability problems between LO an MS Office. My >> method is currently of limited utility, because there is a problem >> with >> rendering of plain text, which affects most documents and overshadows >> other bugs. >> >> I managed to localized the problem by bibisection. I also wanted to >> investigate it more deeply, but was unable to compile that version. >> Thus, I would need help from you :). >> >> The problem resides in that the text layout changed (as bibisected >> using bibisect-43all): > You don't state the platform you were working on. Was it Linux, if so > then probably "bff8fa97e16f0f06fddc5545ea36c8bd2b18a580: Enable > HarfBuzz by default" is what changed during that period. > > With text layout there's the vcl level stuff, which is now backed by > harfbuzz on Linux, and different things on the other platforms. > Typically we have had "little" differences between out platforms when > it comes to text layout already. There's a GSOC underway to experiment > with using harfbuzz on all platforms. In this stack of code different > features enabled/disabled in the underlying fontconfig can cause other > differences too. > > I presume you've ensured the same fonts are available to LibreOffice as > MSOffice otherwise you'd not have gotten this far anyway I guess. > > There's also some extra "wiggling" of glyphs that writer does on top of > all this. I never could quite figure out what writer does there or why > it does it, but that adds some extra variability to things. > > I could never get an exact pixel-to-pixel match in the past (ages ago, > before the harfbuzz rework) so I guess this was also a "slightly-fuzzy" > approach in the close-match case. > > Maybe there's some way to try disabling kerning in the harfbuzz layer > to see if that makes a difference in any way. > > C. > > as an aside, if I was going down the road again of investigating our > differences in layout vs msoffice I'd probably try getting harfbuzz up > and running under windows LibreOffice and then see what the differences > there were between LibreOffice/MSOffice after throwing out all the > mysterious writer adjustment code. Possibly creating some ultra-dumb > font that is plain black blocks or something like that. > ___ > LibreOffice mailing list > LibreOffice@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/libreoffice -- email & jabber: sramek.mi...@gmail.com ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
LO - MSO interoperability test
Hi, this is my first post to this ML :) At the Milano conference I've presented my tools for interoperability testing (http://conference.libreoffice.org/talks/2013/content/sessions/009/files/2013-Milano-Sramek.pdf) . Being influenced by Adam Fyne's talks I've modified them to roundtrip tests and have used the set of documents he has collected. You can find the results with description here: https://copy.com/CZ3WBWtpe7jmQeQp (use LO 4.4.0.0 to 4.4.1.0 to open the spreadsheet - see https://bugs.documentfoundation.org/show_bug.cgi?id=92773). The tools comprise of a few python and bash scripts and should be run on Linux and Windows machines. The current tests are focused on LO interoperability with MSO on the basis of docx documents (roundtrip and print tests). I can, however, implement also other test (e.g. MSO or Google docs vs. LO on the basis of ODT, doc, docx, rtf and other files). In fact any office software can be used, if it has a commad line interface for file conversion. The software is of course open, I just I not distribute it, since it requires quite complex setup to work with (several computers with LO and MSO installations). You can get it, if you want. If you find this interesting, I can in future run the test for new releases or simplify it to be used by a developer to compare files during the develoment process. Or, maybe, we can set up a web service doing this automatically (processing of all files takes about one day). I'll gladly answer your questions. best Milos -- Milos Sramek, msrame...@gmail.com signature.asc Description: OpenPGP digital signature ___ LibreOffice mailing list LibreOffice@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice
