Created attachment 81702 [details] hr_HR.xml local file Tried to improve decimals and time specs, most done on dates - added (hope so) genitive case for log date formats, added spaces present date formats where needed.
Please put this on hold until we recheck it.
Please note that we already have [hr-HR] locale data in i18npool/source/localedata/data/hr_HR.xml, see also http://cgit.freedesktop.org/libreoffice/core/tree/i18npool/source/localedata/data/hr_HR.xml For already existing files please submit a diff instead of entire new files, preferably against the latest revision available. I just sorted the FormatElement entries in that committed file by formatindex for comparability, but even with the indentation we use applied to the data you provided there are too many changes, I'll attach the diff so you can check. Specifically I have doubts about the change of the group separator from '.' dot to ' ' space, http://www.unicode.org/cldr/charts/by_type/patterns.numbers.html#a1ef41eaeb6982d lists the '.' dot as group separator. Additionally, the negative currency formats introduce the () parentheses formats, which I think are wrong, see also https://wiki.documentfoundation.org/LibreOffice_Localization_Guide/How_To_Submit_New_Locale_Data Pitfalls. Not sure about the change of LongDateDaySeparator. - <LC_FORMAT replaceFrom="[CURRENCY]" replaceTo="[$kn-41A]"> + <LC_FORMAT replaceFrom="[CURRENCY]" replaceTo="[$Kn-041a]"> changes currency symbol from 'kn' to 'Kn' (uppercase K). According to https://en.wikipedia.org/wiki/Croatian_kuna the curency symbol is 'kn' all lowercase. Note also that the hexadecimal number must be uppercase, so 41A. + <IndexKey phonetic="false" default="true" unoid="alphanumeric">A-B-C-Č-Ć-D-Đ-DŽ-E-F-G-H-I-J-K-L-LJ-M-N-NJ-O-P-R-S-Š-T-U-V-Z-Ž</IndexKey> is wrong in using the '-' between all letters, which is reserved to express a range between two letters, please see documentation in i18npool/source/localedata/data/locale.dtd http://cgit.freedesktop.org/libreoffice/core/tree/i18npool/source/localedata/data/locale.dtd In the calendar the month names and abbreviations are now capitalized, not sure if that is right or wrong, the ICU locale explorer says they aren't, https://ssl.icu-project.org/icu-bin/locexp?d_=en&_=hr_HR Changes in <LC_TRANSLITERATION> element probably just due to overwriting existing data. Elements of <LC_NumberingLevel> are different. Elements of <LC_OutLineNumberingLevel> are removed and replaced by <LC_OutLineNumberingLevel ref="en_US"/>, most certainly not correct. Summary: this change can't be applied.
Created attachment 81826 [details] diff of re-indented file against current revision
Thanks for your comment. Does current hr_HR.xml file which is in use in LO contain all needed / possible fields? Want to know if we need to recheck if all xml fields are there and we can just edit current file, or we maybe need to add new xml fields?
It contains all elements. Only that it has <LC_COLLATION ref="sh_RS"/> which means that for the LC_COLLATION element it inherits the content defined in the sh_RS.xml file.
Created attachment 82620 [details] few modification to a croatian localization file Added few date formats, modified LongDateDaySeparator and more...
Looks good, except some nitpicks I have: * FormatElement formatindex 50 and 51 are inserted instead of appended to the list (really minor nitpick ;-) * FormatElement formatindex="24" was changed from <FormatCode>D, MMM YYYY</FormatCode> to <FormatCode>D. M. YYYY.</FormatCode> * I suggest to keep the original element and add another element instead. * Reason: when formats <50 are used with system standard locale (not hard set to a specific locale), and the standard locale is switched to another locale without reloading the document, the numerically equivalent format of that other locale is applied. Formats' meanings should match across locales, this is not enforced but it should.. * FormatElement formatindex="52" I thought this should use non-breaking spaces, as you asked about it in mail? The diff is with ordinary space. * DateAcceptancePattern elements for the new formats with spaces are missing. The rest looks good. If you have problems with any of these just tell me and I'll do the necessary changes. Last but not least: As this seems to have been your first contribution to the code base could you please send us a blanket statement that you contribute all your past and future patches under LGPLv3+ and MPL 1.1 licenses? Best on the dev (or l10n) mailing list so we can link to it from https://wiki.documentfoundation.org/Development/Developers Something like this does nicely: All of my past & future contributions to LibreOffice may be licensed under the MPLv2/LGPLv3+ dual license. We need it to commit your contribution to the code repository. Thanks.
For non-breaking character, should   do? so need add <DateAcceptancePattern>D. M.</DateAcceptancePattern> <DateAcceptancePattern>D. M. YYYY</DateAcceptancePattern> <DateAcceptancePattern>DD. MM. YYYY.</DateAcceptancePattern> and </FormatElement> <FormatElement msgid="DateFormatskey26" default="false" type="medium" usage="DATE" formatindex="54"> <FormatCode>D. M. YYYY</FormatCode> <FormatElement msgid="DateFormatskey24" default="false" type="medium" usage="DATE" formatindex="52"> <FormatCode>DD. MM. YYYY.</FormatCode> </FormatElement>? format index kes 24 back to D, MMM YYYY and formatindex key 23 and 25 to the end of that section?
Yes, I do have hard time doing it right. Please, help me and change what needs to be changed, I just can't get it right. When everything is in place, let me no so I'll send a contribution statement to mailing list. Thanks
Eike, formatindex="24" was changed because it was 100% indentical as formatindex="23" (except afcourse formatindex number and title).
(In reply to comment #8) > For non-breaking character, should   do? Numeric character entities indeed seem to work. For readability I prefer the literal character though. (In reply to comment #9) > Yes, I do have hard time doing it right. Please, help me and change what > needs to be changed, No problem, I'll do. I'll commit your change and then adapt to the needs with a followup-commit. (In reply to comment #10) > Eike, formatindex="24" was changed because it was 100% indentical as > formatindex="23" (except afcourse formatindex number and title). Yes, there may be a few duplicate entries with the current conventions, but nothing to worry about, they don't harm. In case you're interested in the historical reason see http://cgit.freedesktop.org/libreoffice/core/tree/offapi/com/sun/star/i18n/NumberFormatIndex.idl There are future plans to revive the original meanings of the SYSTEM/SYS/DEF formats mentioned and introduce a system settings layer, so we're not weeding everything out.
Please leave DateAcceptancePattern as they are. We discussed this in another bug when you were introducing this and kicked out few of them. This should be fine, especially if it only affect Calc. Concerning formatindex="24", I don't see why we should keep 2 indentical formattings in UI. If it's all the same to you, we will change it. If there are some other reasons and they must stay same, then we will add new one. Finaly, does LC_COLLATION (this is new, copied from serbian xml file and edited) and LC_INDEX (changed to represent croatian alphabet) look ok to you? We kicked out A-Z mark in LC_INDEX which included english only letters like Q, X, W and Y. Will that affect anything?
(In reply to comment #12) > Please leave DateAcceptancePattern as they are. We discussed this in another > bug when you were introducing this and kicked out few of them. This should > be fine, especially if it only affect Calc. Ok if dates are not to be entered with spaces in between. > Concerning formatindex="24", I don't see why we should keep 2 indentical > formattings in UI. If it's all the same to you, we will change it. If there > are some other reasons and they must stay same, then we will add new one. They must not stay the same. I know it may look odd to have duplicated entries. The originally intended purpose also may have resulted in duplicated entries for some locales depending on the current system's Regional Settings. Currently there is no compelling reason though to keep the duplicated entries identical. > Finaly, does LC_COLLATION (this is new, copied from serbian xml file and > edited) and LC_INDEX (changed to represent croatian alphabet) look ok to you? Yes, looks good. > We kicked out A-Z mark in LC_INDEX which included english only letters like > Q, X, W and Y. Will that affect anything? That's ok, index terms starting with a letter not defined in IndexKey will be appended after the last entry with a defined IndexKey.
Krunoslav Å ebetiÄ committed a patch related to this issue. It has been pushed to "master": http://cgit.freedesktop.org/libreoffice/core/commit/?id=38fc4d86d96c8d4c9b79299f83fd005ac3bef85c changed Croatian [hr-HR] locale data, fdo#66376 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.
Eike Rathke committed a patch related to this issue. It has been pushed to "master": http://cgit.freedesktop.org/libreoffice/core/commit/?id=2cce78137bbf9810421b9fdedd45127b019b1188 small cleanup in [hr-HR] locale data, fdo#66376 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.
Looks good to me, thank you. Will this be included in 4.1.1 maybe or this kind of change only follows major releases?
I can backport and submit for 4-1 review.
Pending review for 4-1 as https://gerrit.libreoffice.org/4977
Thank you.
Eike Rathke committed a patch related to this issue. It has been pushed to "libreoffice-4-1": http://cgit.freedesktop.org/libreoffice/core/commit/?id=faf6e4c29463c28fd8c14f72542126139e0c9fbf&h=libreoffice-4-1 changed Croatian [hr-HR] locale data, fdo#66376 It will be available in LibreOffice 4.1.1. 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.
Thanks Eike, Kruno