- OpenStack-I18n - lists.openstack.org

Re: [Openstack-i18n] Would like Zanata branch to correct 1494819
by Douglas Fish 23 Sep '15

23 Sep '15

1 0

Re: [Openstack-i18n] Would like Zanata branch to correct 1494819
by Ying Chun Guo 23 Sep '15

23 Sep '15

Thank you for looking into it, Illya. I think we need to fix it, although the manually process might be a little complex. I'm going to: 1. Create a new version in Zanata 2. Mark master as read only 2. Pull the source files and the translation files. 3. Change the pluralities settings in Russian translation files. 4. Upload the source files and translation files to the new version. 5. Check if the all things are right, change master version name to "master-old", and then change the new version name to "master" How do you want to update these pluralities then? By program ? Or manually ? Is it possible to make the changes through Zanata UI, other than uploading po files ? Best regards Ying Chun Guo (Daisy) Douglas Fish/Rochester/IBM wrote on 09/23/2015 04:38:19 AM: > From: Douglas Fish/Rochester/IBM@IBMUS > To: ilyaalekseyev(a)acm.org > Cc: guoyingc(a)cn.ibm.com, openstack-i18n(a)lists.openstack.org > Date: 09/23/2015 04:38 AM > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > Hi Ilya, > > Thanks for looking at this! Unfortunately having 4 pluralities > confuses Horizon. I've tested out stripping one of the pluralities > and restoring the expected Plural-forms definition (just like the > one you linked) and it works great. > > I had hoped to just repair the PO files in this way and upload the > corrected translations to Zanata. > > Unfortunately, I understand from Daisy's research that once a > language/project has been created in Zanata in this way we cannot > just upload a PO file with corrected pluralities . More > investigation will be needed to understand the process to fix this. > > Doug > > > ----- Original message ----- > From: Ilya Alekseyev <ilyaalekseyev(a)acm.org> > Sent by: ilya(a)phpcoder.ru > To: Douglas Fish/Rochester/IBM@IBMUS > Cc: Ying Chun Guo <guoyingc(a)cn.ibm.com>, "openstack- > i18n(a)lists.openstack.org" <openstack-i18n(a)lists.openstack.org> > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > Date: Tue, Sep 22, 2015 2:39 PM > > Hi Doug and Daisy! > > Looks like we have wrong pluralities in current release: > in https://github.com/openstack/horizon/blob/master/ > openstack_dashboard/locale/ru/LC_MESSAGES/django.po > > we have: > "Plural-Forms: nplurals=4; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n" > "%10<=4 && (n%100<12 || n%100>14) ? 1 : n%10==0 || (n%10>=5 && n%10<=9) || (n" > "%100>=11 && n%100<=14)? 2 : 3);\n" > > but in previuos releses we have: > "Plural-Forms: nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n% > 10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2);\n" > > Last one is right for Russian. > > We can make translation in Zanata for 4 pluralities, but it can give > in some cases bugs with wrong plural forms. > > Kind regards, > Ilya Alekseyev. > > 2015-09-22 14:22 GMT+03:00 Douglas Fish <drfish(a)us.ibm.com>: > Hi Daisy, > > I am certain this issue is caused by plural forms. The issue with > missing menus is seen with the current 4 plural form translations; > when I correct the PO file to have 3 forms Horizon properly displaysthe menus. > > Doug > > > ----- Original message ----- > From: Ying Chun Guo/China/IBM > To: Douglas Fish/Rochester/IBM@IBMUS@IBMAU > Cc: openstack-i18n(a)lists.openstack.org > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > Date: Tue, Sep 22, 2015 12:49 AM > > Hi, Doug > > We need to consider it carefully. > > After working with Zanata dev team, I got to understand: > > 1. Zanata has its own plural forms. If po file has its own plural > form, Zanata will use that. > Otherwise, Zanata will use its own plural form. Pluralforms in Zanata: > nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && > (n%100<10 || n%100>=20) ? 1 : 2). > So it looks like that OpenStack po files overrode the pluralforms in Zanata. > > 2. Zanata currently don't allow update of plural forms once the > document has been uploaded to Zanata. > Even with deletion of that source document and re-upload (in the > same project-version, same doc name and same path), > Zanata will still reuse the old pluralform. > > 3. If we want to update the pluralforms, we will need to do it in > new project-version > (don't use copy-from previous version function). > > I think we need to carefully design the steps to change pluralforms > of Russian. > At first, I want to confirm with you, are you sure this issue is > caused by plural forms? > > Best regards > Ying Chun Guo (Daisy) > > > Douglas Fish/Rochester/IBM wrote on 2015/09/18 23:38:22: > > > From: Douglas Fish/Rochester/IBM@IBMUS > > To: guoyingc(a)cn.ibm.com > > Cc: openstack-i18n(a)lists.openstack.org > > Date: 2015/09/18 23:38 > > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > > > Hi Daisy, > > > > Yes, I understand - I don't think I can even propose my branch > > unless I am added as a translator by the Russian team. > > > > My proposal would include removing one plural form for > > 4 messages in horizon django.po > > 1 message in horizon djangojs.po > > 144 messages in openstack_dashboard django.po > > 0 messages in openstack_dashboard djangojs.po > > plus the pluralization specification would be updated in all 4 files. > > > > Doug > > > > ----- Original message ----- > > From: Ying Chun Guo <guoyingc(a)cn.ibm.com> > > To: Douglas Fish/Rochester/IBM@IBMUS > > Cc: openstack-i18n(a)lists.openstack.org > > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > Date: Fri, Sep 18, 2015 3:25 AM > > > > Are there a lot of changes in your proposed po files ? > > I'm still waiting for Russian team's response. > > > > Best regards > > Ying Chun Guo (Daisy) > > > > > > "Douglas Fish" <drfish(a)us.ibm.com> wrote on 09/18/2015 05:37:07 AM: > > > > > From: "Douglas Fish" <drfish(a)us.ibm.com> > > > To: openstack-i18n(a)lists.openstack.org > > > Date: 09/18/2015 05:38 AM > > > Subject: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > > > > > Hi translation friends, > > > > > > I'd like to contribute 4 PO files for the Horizon Russian > > > translation to correct the issue > > > https://bugs.launchpad.net/openstack-i18n/+bug/1494819 > > > > > > These uploads won't contain any new translations, but instead will > > > be an adjustment to the PO files so they have 3 pluralities, which > > is standard > > > http://localization-guide.readthedocs.org/en/latest/l10n/pluralforms.html > > > > > > Specifically, I plan to remove the msgstr[2] translation from the > > > files (because msgstr[2] is blank for many cases), rename the msgstr > > > [3] translation to msgstr[2] and adjust the plurality specification to > > > nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && > > > (n%100<10 || n%100>=20) ? 1 : 2); > > > > > > As discussed in the team meeting http://eavesdrop.openstack.org/ > > > meetings/openstack_i18n_meeting/2015/openstack_i18n_meeting. > > > 2015-09-17-13.00.log.html > > > please create a Zanata branch so I can propose these corrections > for review. > > > > > > Thanks! > > > > > > Doug > > > _______________________________________________ > > > Openstack-i18n mailing list > > > Openstack-i18n(a)lists.openstack.org > > > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n > > > _______________________________________________ > Openstack-i18n mailing list > Openstack-i18n(a)lists.openstack.org > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n >

1 0

Re: [Openstack-i18n] IBM translations to contribute: update
by Ying Chun Guo 23 Sep '15

23 Sep '15

Hi, team I have made enough tests on the test server with the help from Doug and his colleagues. In the test server, I created versions for nova, glance, cinder, swift, ceilometer, heat, keystone, and ironic. Doug and his colleagues uploaded their po files for languages de, es, fr, it, ja ko-KR, pr-BR, ru, zh-CN, and zh-TW. I tested merging two versions. Things are running well. The test server is still in good condition. So there is no technical issues block me to do the same in the production server. I even tried "Group" in the test server. The "Group" feature looks wonderful. So if there is no objections, I will go to production server, doing the same as what I did in the test server. Best regards Ying Chun Guo (Daisy) Douglas Fish/Rochester/IBM wrote on 09/18/2015 05:14:22 AM: > From: Douglas Fish/Rochester/IBM@IBMUS > To: guoyingc(a)cn.ibm.com > Cc: openstack-i18n(a)lists.openstack.org > Date: 09/18/2015 05:14 AM > Subject: Re: [Openstack-i18n] IBM translations to contribute: update > > Hi Daisy, > > We have translations to contribute for ceilometer, cinder, glance, > heat, ironic, keystone, nova and swift for languages de, es, fr, it, > ja ko-KR, pr-BR, ru, zh-CN, and zh-TW > > We are working to upload those to the Zanata test server to make > sure there are no technical issues. > > Doug > > ----- Original message ----- > From: Ying Chun Guo <guoyingc(a)cn.ibm.com> > To: Douglas Fish/Rochester/IBM@IBMUS > Cc: "Openstack-i18n Openstack-i18n" <openstack-i18n(a)lists.openstack.org> > Subject: Re: [Openstack-i18n] IBM translations to contribute: update > Date: Thu, Sep 17, 2015 11:04 AM > > Hi Doug, Hi team, > > In today's IRC meeting, we talked about the process to get IBM translations. > I'm going to normalize the process to contribute translations by > uploading po files. > > Here are some key points during the meeting: > > 1. It is strongly discouraged to use the offline editor other than web tool. > Because the offline editor could not take the advantages of > translation memory and glossary, > which might affect the translation quality. > 2. Simply uploading po files are not allowed because it has a risk > that the po files might contain bad translations. > 3. In some special cases, if people who want to contribute > translations by uploading po files, following below steps > 3.1 Send email to I18n mailing list, describing the source > documents and languages he/she is going to upload. > 3.2 Zanata admin will create a branch. Then contributors could > upload po files to the proposed branch. > 3.3 When the uploading is done, language coordinators could > review the translations. > The coordinators could have two kinds of review results: > agree to merge or deny to merge. > If the coordinators deny to merge, coordinators should > give some comments, either to delete this branch, or to suggestions > to improve. > 3.4 If the coordinators agree to merge to the master branch, > Zanata admin will make the merge. > A option could set during the branches merging to make sure not > overwrite translated/approved translations in the source branch. > After merging, the proposed branch will be deleted. > > I have been testing branches in the Zanata test server for several days. > When I make sure it works fine, I will go to the production server. > > @Doug, please let us know which resources and which languages you > are going to propose ? > > Best regards > Ying Chun Guo (Daisy) > > > "Douglas Fish" <drfish(a)us.ibm.com> wrote on 09/16/2015 08:05:52 PM: > > > From: "Douglas Fish" <drfish(a)us.ibm.com> > > To: "Openstack-i18n Openstack-i18n" <openstack-i18n(a)lists.openstack.org> > > Date: 09/16/2015 08:06 PM > > Subject: [Openstack-i18n] IBM translations to contribute: update > > > > Here's the revised approach we are taking to contributing our IBM > > translations to the community, based on feedback on suggestions: > > > > We are planning to make the proposed translations available for a > > pre-review as a separate branch in Zanata, but they are not there > > yet. While testing out this upload on the development server we > > discovered that uploading PO files caused the server to become > > unresponsive for a while. Thanks to pleia2 and the infra team for > > sorting that out and addressing the issue. We'll try the upload on > > dev again tomorrow and if it goes well we will make the translations > > available for pre-review in a separate Zanata branch. > > > > I had originally proposed ending the pre-review on Monday. Obviously > > that's not going to happen! I understand that different language > > teams have differing levels of readiness to accept these > > contributions. I understand some groups are ready at this time, > > while others would like to have more time, possibly even into Mitaka > > to review. We aren't going to upload PO files for any language were > > the team does not want them. > > > > Language coordinators, please let me know how you'd like to proceed > > with these contributions for your language. Do you want > > 1) us to upload them to Zanata's current master translations for > > normal review and editing as soon as we can, > > 2) us to upload them to Zanata as a separate branch for further > pre-review, or > > 3) hold off any submission or preview at this time. > > > > Please note that it important to us that we upload these > > translations. The translations are a valued IBM asset, it's > > important to us that they are recognized as contributed by IBM. > > > > I'll be at the Thursday i18n team meeting, and am available by IRC > > (doug-fish) or email if you have questions or further concerns. > > > > Thanks! > > > > Doug_______________________________________________ > > Openstack-i18n mailing list > > Openstack-i18n(a)lists.openstack.org > > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n

6 17

Arabic Script mixed with Gurmukhi Script in Translation Memory
by ਅਨਪੜ੍ਹ ਜਿਮੀਂਦਾਰ 23 Sep '15

23 Sep '15

Hi there, As I don't know how this mix up happened so someone please respond to this issue. After the migration has done and i am doing translation in Zanata I am quite surprised to see Arabic script in my Translation Memory of many projects. I have clarified before, and explaining again, that Punjabi has two entirely different scripts i.e. Gurmukhi (ਪੈਂਤੀ) and Shahmukhi (اردو), while former is widely used in Indian side of Punjab aka Eastern Punjab the latter is used in Pakistani side of Punjab aka Western Punjab. I guess while migration someone has mistaken both the scripts as one and joined them. As this is my assumption so it is you guys, will have to help me out. Thanks Amandeep Singh Saini Co-ordinator Punjabi Team (pa-IN)

3 5

Re: [Openstack-i18n] Russian translations with 4 pluralities?
by Douglas Fish 22 Sep '15

22 Sep '15

3 2

Re: [Openstack-i18n] Would like Zanata branch to correct 1494819
by Douglas Fish 22 Sep '15

22 Sep '15

2 2

Automatic copying of translations in Zanata?
by Andreas Jaeger 22 Sep '15

22 Sep '15

Today the majority of the IBM translations got merged into the the normal versions since zanata by default searches at every upload with the zanata client for translation of untranslated strings and merges them [1]. So, when ceilometer got uploaded as master version, the IBM ceilometer strings where searched for already translated strings and untranslated ceilometer strings where automatically translated with the IBM version. Daisy, Clark, Lyz, and myself just had some discussion on #openstack-i18n and decided to disable for now the automatic copying of translations while we discuss what to do next. The question for me is, do we want this kind of automatic copying of translations - and if, how do we want it. Previously enabled was copying only in a project between versions. So, for example if there is a stable/liberty and a master version (or an IBM and a master version) of the same document, and one contains a translated string and the other gets updated, the translated string gets copied over if the string is not translated yet. This feature can also be enabled for copying between projects and files. So, we could enable it and then when a string gets translated in let's say nova and exactly the same string is in neutron and untranslated, the translation would be copied from nova to neutron. So, question to discuss: 1) Do we translators want such automatic translations? 2) And if yes, only between: a) versions of same document b) inside a project (so, only inside nova) c) across all projects and documents Note that this setting is the same for all languages. Note that there's an option as well to do this automatic translation manually, for example at project creation time. So, if a new project gets setup, Zanata could try to copy automaticly translations across projects for a quick start. What do you think? Do you want to translate each string manually - or let Zanata do the work and to which extend? I hope I summarized everything correctly, if not please correct my understanding, Andreas References: [1] http://zanata.org/help/copytrans-explained/ -- Andreas Jaeger aj(a){suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126

1 0

Please review IBM translations in the master branch
by Ying Chun Guo 22 Sep '15

22 Sep '15

Hi, friends Something unexpected happened tonight. IBM translations were copied into master branches. The copy action is triggered by uploading of pot files, which are executed by automatic Jenkins jobs when the source pot files in code repo are updated. Nobody expects it. Infrastructure team have figure out a way to stop auto copy while uploading pot files so this kind of things won't happen the second time. IBM translations only fill in blanks without overwriting the others. I hope translation teams could take a look at IBM translations in the master branches. IBM contributions could be located by searching "last-modified-after: 2015-09-21". You could * improve translations if they are easily to improve. Different terminologies could easily be fixed by "project-wide Search & Replace". ( in the Editor view, click the magnifying glass icon on the right top corner to open the search&replace view.) * reject some translations if you don't like. * If you don't like the whole translations at all, I may work out a way to revert it but it will cost time. The relevant translation teams include: de es fr it ja ko_KR pt_BR ru zh_CN zh_TW. The relevant documents are: nova.pot, glance.pot, swift.pot, ceilometer.pot, cinder.pot, heat.pot, ironic.pot, keystone.pot and neutron.pot. Sorry for the troubles and thank you for your work. Best regards Ying Chun Guo (Daisy)

1 0

IP capitalization change in nova - ok to merge?
by Andreas Jaeger 22 Sep '15

22 Sep '15

Translation team, I've just send a patch for review that fixes lots of user visible capitalization strings in nova as I noticed during translation that it should be "IP" and not "ip" for example in "fixed IP". Is it ok to get this patch in now - or should I wait after liberty and mark it as WIP for now? https://review.openstack.org/225923 Andreas -- Andreas Jaeger aj(a){suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126

2 1

Re: [Openstack-i18n] Would like Zanata branch to correct 1494819
by Ying Chun Guo 22 Sep '15

22 Sep '15

Hi, Doug We need to consider it carefully. After working with Zanata dev team, I got to understand: 1. Zanata has its own plural forms. If po file has its own plural form, Zanata will use that. Otherwise, Zanata will use its own plural form. Pluralforms in Zanata: nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && (n%100<10 || n%100>=20) ? 1 : 2). So it looks like that OpenStack po files overrode the pluralforms in Zanata. 2. Zanata currently don't allow update of plural forms once the document has been uploaded to Zanata. Even with deletion of that source document and re-upload (in the same project-version, same doc name and same path), Zanata will still reuse the old pluralform. 3. If we want to update the pluralforms, we will need to do it in new project-version (don't use copy-from previous version function). I think we need to carefully design the steps to change pluralforms of Russian. At first, I want to confirm with you, are you sure this issue is caused by plural forms? Best regards Ying Chun Guo (Daisy) Douglas Fish/Rochester/IBM wrote on 2015/09/18 23:38:22: > From: Douglas Fish/Rochester/IBM@IBMUS > To: guoyingc(a)cn.ibm.com > Cc: openstack-i18n(a)lists.openstack.org > Date: 2015/09/18 23:38 > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > Hi Daisy, > > Yes, I understand - I don't think I can even propose my branch > unless I am added as a translator by the Russian team. > > My proposal would include removing one plural form for > 4 messages in horizon django.po > 1 message in horizon djangojs.po > 144 messages in openstack_dashboard django.po > 0 messages in openstack_dashboard djangojs.po > plus the pluralization specification would be updated in all 4 files. > > Doug > > ----- Original message ----- > From: Ying Chun Guo <guoyingc(a)cn.ibm.com> > To: Douglas Fish/Rochester/IBM@IBMUS > Cc: openstack-i18n(a)lists.openstack.org > Subject: Re: [Openstack-i18n] Would like Zanata branch to correct 1494819 > Date: Fri, Sep 18, 2015 3:25 AM > > Are there a lot of changes in your proposed po files ? > I'm still waiting for Russian team's response. > > Best regards > Ying Chun Guo (Daisy) > > > "Douglas Fish" <drfish(a)us.ibm.com> wrote on 09/18/2015 05:37:07 AM: > > > From: "Douglas Fish" <drfish(a)us.ibm.com> > > To: openstack-i18n(a)lists.openstack.org > > Date: 09/18/2015 05:38 AM > > Subject: [Openstack-i18n] Would like Zanata branch to correct 1494819 > > > > Hi translation friends, > > > > I'd like to contribute 4 PO files for the Horizon Russian > > translation to correct the issue > > https://bugs.launchpad.net/openstack-i18n/+bug/1494819 > > > > These uploads won't contain any new translations, but instead will > > be an adjustment to the PO files so they have 3 pluralities, which > is standard > > http://localization-guide.readthedocs.org/en/latest/l10n/pluralforms.html > > > > Specifically, I plan to remove the msgstr[2] translation from the > > files (because msgstr[2] is blank for many cases), rename the msgstr > > [3] translation to msgstr[2] and adjust the plurality specification to > > nplurals=3; plural=(n%10==1 && n%100!=11 ? 0 : n%10>=2 && n%10<=4 && > > (n%100<10 || n%100>=20) ? 1 : 2); > > > > As discussed in the team meeting http://eavesdrop.openstack.org/ > > meetings/openstack_i18n_meeting/2015/openstack_i18n_meeting. > > 2015-09-17-13.00.log.html > > please create a Zanata branch so I can propose these corrections for review. > > > > Thanks! > > > > Doug > > _______________________________________________ > > Openstack-i18n mailing list > > Openstack-i18n(a)lists.openstack.org > > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n

1 0