From ffesti at redhat.com Wed Jul 12 11:50:30 2006 From: ffesti at redhat.com (Florian Festi) Date: Wed, 12 Jul 2006 13:50:30 +0200 Subject: [Fedora-i18n-list] Numeric values Message-ID: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> Hi! I am currently working on the i18n of the foomatic printer database. Messages that need to be translated are printer/printing option names and values. These often contain numerical values like "600 DPI", "256 MB Memory extention", "Tray 4", "1.125x3.5in, 28x89mm", "-30?" and so on. To reduce to number of messages I'd like to merge string that only differ by the numeric values to some thing like "%s DPI", "Tray %s", "%sx%sin. %sx%smm". Are there any special issues I have to take care of. Do numbers need to be translated into something else than [0-9]* in some languages and if so is this done automatically by sprintf if LC_NUMERIC is set? Thanks for your help Florian Festi From petersen at redhat.com Wed Jul 12 12:57:21 2006 From: petersen at redhat.com (Jens Petersen) Date: Wed, 12 Jul 2006 21:57:21 +0900 Subject: [Fedora-i18n-list] Re: Tamil input In-Reply-To: References: Message-ID: <44B4F1B1.2080007@redhat.com> [cc to fedora-i18n-list] Hi, U. Sudhakar wrote: > i am sudhakar,i am working in OS-Localication(Tamil,Hindi) using dcc > based... Okay nice to hear. Sorry, what is dcc? > i want On-screen keyboard and input methods for select language tamil > typewriter keyboard.....what can i do.. > i am installing scim-table-addtional,how to work scim..in Tamil pls help > me sir.. Are you using Fedora Core 5? This is discussed in the release notes, but I suggest installing m17n-db-tamil and scim-m17n instead of scim-tables. If you can use Yum, you can also do it by running "yum groupinstall tamil-support". Hope that helps, Jens From abelcheung at gmail.com Tue Jul 25 19:24:52 2006 From: abelcheung at gmail.com (Abel Cheung) Date: Wed, 26 Jul 2006 03:24:52 +0800 Subject: [Fedora-i18n-list] Numeric values In-Reply-To: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> References: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> Message-ID: On 7/12/06, Florian Festi wrote: > To reduce to number of messages I'd like to merge string that only > differ by the numeric values to some thing like "%s DPI", "Tray %s", > "%sx%sin. %sx%smm". > > Are there any special issues I have to take care of. Do numbers need to > be translated into something else than [0-9]* in some languages and if > so is this done automatically by sprintf if LC_NUMERIC is set? AFAIK a few languages, especially Persian, may want to take advantage of it. However it's better to ask people speaking Persian directly though. Abel > > Thanks for your help > > > Florian Festi > > -- > Fedora-i18n-list mailing list > Fedora-i18n-list at redhat.com > https://www.redhat.com/mailman/listinfo/fedora-i18n-list > -- Abel Cheung (GPG Key: 0xC67186FF) Key fingerprint: 671C C7AE EFB5 110C D6D1 41EE 4152 E1F1 C671 86FF -------------------------------------------------------------------- * GNOME Hong Kong - http://www.gnome.hk/ * Opensource Application Knowledge Assoc. - http://oaka.org/ From ankit at redhat.com Fri Jul 28 08:04:16 2006 From: ankit at redhat.com (Ankitkumar Rameshchandra Patel) Date: Fri, 28 Jul 2006 13:34:16 +0530 Subject: [Fedora-i18n-list] Numeric values In-Reply-To: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> References: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> Message-ID: <44C9C500.50400@redhat.com> Florian Festi wrote: > Hi! > > I am currently working on the i18n of the foomatic printer database. > Messages that need to be translated are printer/printing option names > and values. These often contain numerical values like "600 DPI", "256 MB > Memory extention", "Tray 4", "1.125x3.5in, 28x89mm", "-30?" and so on. > > To reduce to number of messages I'd like to merge string that only > differ by the numeric values to some thing like "%s DPI", "Tray %s", > "%sx%sin. %sx%smm". > > Are there any special issues I have to take care of. Do numbers need to > be translated into something else than [0-9]* in some languages and if > so is this done automatically by sprintf if LC_NUMERIC is set? > > Thanks for your help > > > Florian Festi > > -- > Fedora-i18n-list mailing list > Fedora-i18n-list at redhat.com > https://www.redhat.com/mailman/listinfo/fedora-i18n-list > Hi, Regarding the messages about which you are not sure what to do with them, i suggest you to extract all the messages as it is. And leave those messages up to translators as they better understand their language, and they can decide better what to write for particular message. If you don't would like to give those messages for translation and would like keep them in English then don't extract those messages, otherwise there is no way AFAIK to reduce the number of messages. Leaving the decision on translator is a better choice i would suggest you rather than doing automation using LC_NUMERIC or anything else... Let me know, if you need any more clarification... Thanks, -- Ankit Patel From roozbeh at farsiweb.info Sat Jul 29 12:53:55 2006 From: roozbeh at farsiweb.info (Roozbeh Pournader) Date: Sat, 29 Jul 2006 16:23:55 +0330 Subject: [Fedora-i18n-list] Numeric values In-Reply-To: References: <1152705030.7130.121.camel@triefenried.stuttgart.redhat.com> Message-ID: <1154177635.5190.12.camel@tameshk.farsiweb.info> ??? ????????? 2006-07-26 ???? 03:24 +0800? Abel Cheung ????: > On 7/12/06, Florian Festi wrote: > > To reduce to number of messages I'd like to merge string that only > > differ by the numeric values to some thing like "%s DPI", "Tray %s", > > "%sx%sin. %sx%smm". > > > > Are there any special issues I have to take care of. Do numbers need to > > be translated into something else than [0-9]* in some languages and if > > so is this done automatically by sprintf if LC_NUMERIC is set? > > AFAIK a few languages, especially Persian, may want to take > advantage of it. However it's better to ask people speaking > Persian directly though. Yes, numbers need to be translated. Only Persian uses that currently, but I expect the number of such languages to grow. Urdu, Pashto, and some of the Arabic locales are candidates. In Persian, we localize some, but not all of the "%d" and such markup to "%Id" and such. Sometimes it gets more complicated, like the case of messages like "%2d" and "%02d", as glibc counts bytes and not characters. Not merging the messages is better for Persian localizers, but if you insist, tell me the exact problem that you are trying to fix and I may be able to help more. Roozbeh