Date: Tue, 18 Feb 2014 10:12:59 +0100 From: Benedict Reuschling <bcr@FreeBSD.org> To: d@delphij.net Cc: "freebsd-doc@freebsd.org" <freebsd-doc@FreeBSD.org>, Warren Block <wblock@FreeBSD.org> Subject: Re: Translations (was Re: svn commit: r43974 - head/en_US.ISO8859-1/books/handbook/advanced-networking) Message-ID: <5303241B.8080906@FreeBSD.org> In-Reply-To: <alpine.BSF.2.00.1402172145010.42338@wonkity.com> References: <201402180226.s1I2QS0x076422@svn.freebsd.org> <5302C7B9.7090208@delphij.net> <alpine.BSF.2.00.1402171940210.42338@wonkity.com> <5302E05F.1050200@delphij.net> <alpine.BSF.2.00.1402172145010.42338@wonkity.com>
next in thread | previous in thread | raw e-mail | index | archive | help
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi guys, thanks for your interest in translation. I think that with the improvements that we can make, we can also make it easier for doc committers making changes to the doc not worry too much about translations with regards to whitespace. I could be wrong, but I think we could combine whitespace and content changes in the future. For now, my primary focus is on getting the translation process more automated and easier to catch up. >>> If there is any way I or other committers can make this easier >>> for translators, please post on the -doc mailing list or to me >>> invididually. >> >> One thing I would recommend is to separate the contents that do >> not need translation (e.g. tags, entities, etc.) and contents >> that needs translation. This way, these contents would serve as >> a positioning blocks when merging from upstream (English). > > That sounds interesting, and leads well into the next part: > >>> We are also trying to modernize the translation process, and >>> automate some of the work that translators are currently forced >>> to do. Anyone who would like to help with that is welcome. >> >> That would be great! How can we help, or is there some kind of >> TODO list? > Well, I have a couple of loose threads that we need to tie together to have a complete and mostly automated system. > There are several things that would be helpful. > > We could use help from people who are experienced with using the > .po/.pot/.mo tools (gettext) on other platforms. > > The basic process is to separate all the content from the markup > automatically. Then an editor can be used to add translations, and > the tool puts a translated file back together from it. This allows > the translation program to remember existing translations, so > translating one document helps translate others. > Yes, that is a good description. I've put a (hopefully) more thorough description on the wiki page below. > textproc/itstool is one of the automatic separator programs. I've > been somewhat stymied trying to figure out how to get the Python > libxml2 implementation used by it to find FreeBSD documentation XML > catalogs. Documentation on this is... let's just say sparse. > Another tool is textproc/po4a that Thomas Abthorpe showed me on one of the hacker lounges two years ago in Canada. He set up a test project for a translation into en_GB to see how these tools fit our toolchain. He send me the following mini-howto, which I extended a bit for a first commit to the german doc repo recently: <mini-howto> install textproc/po4a install editors/poedit # optional install devel/gtranslator #optional install lokalize # somewhere in kde4, too lazy to look To prep a new translation, inline, same directory the following works. Substitute en_GB for the language of your choice. po4a-gettextize -f xml -m article.sgml -p article.pot # renders to pot file #<do translation>, save as article_en_GB.po po4a-translate -f xml -m article.sgml -p article_en_GB.po -l article_en_GB.sgml Trying to convert to po files with existing translations, the following should work providing the sgmlised files match tag for tag po4a-gettextize -M ISO8859-1 -f xml -m ../../../en_US.ISO8859-1/articles/<article>/article.sgml -l article.sgml -p article_de_DE.po </mini-howto> The biggest problem right that I came across is that after the translated strings are put back into the xml document, the formatting is screwed up as the po tools don't know our rules. So, it would we really cool if we had a tool that does this for us that we could run after the po4a-translate step. Another issue is that when I tried po4a-gettextize to extract the strings from an already existing translation the tool bailed out early because it can only match string by string. Since the translation and the english original doc have drifted too far apart and sometimes you would make two sentences for one english sentence, the tool cannot cope with it. I don't have a solution for this yet, but it would be a shame to throw away the already existing translations and not use them to feed a translation memory. > The PC-BSD folks are using tools like Pootle, but not for DocBook > (as far as I know). They have a web site for translation, useful > as an example of what can be done: http://pootle.pcbsd.org/ > > Benedict (CCed) has made some progress with some of these tools. > I think there are plans to add a page to the wiki, but don't know > if it is present yet. I put everything in the DevSummit wiki page and will start a separate one with the results from the summit to continue from there: https://wiki.freebsd.org/201405DevSummit/Translation Regards Benedict Reuschling Documentation Committer The FreeBSD Project The FreeBSD Documentation Project -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.22 (Darwin) Comment: GPGTools - https://gpgtools.org Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBCgAGBQJTAyQbAAoJEAQa31nbPD2LLVsH/1DzRh3y2Wm5YTfo+OzmQTz9 9XKZTTaP3kF20HqGamMPOgEJdY8Szo6sBNJyPoPtkzBWHcsUv7s2W49MGwIRkdwc yACdIPrladNXc77BpgwNsqoWGLsSElllqe2vYQNlICUrZbqgHC2a52eTtOi1QJZ5 P98M7wDtHFaxLdxKb17qBaL9/mvx5HhLXc6ZwP1GbKHvPWUtTN08H/BiGRc5cy1s yTPqP/xmQNm7l3VTURfMqyuJykG+nbxpr6FvupDSo25rDfbxi0z+wcUeR1c7hLol Vsh2p3Spx48Cc0Ccfj4vP+4Bdt6eiJ/O8jsofroZZKk1l+zBMhxbuVt3TugVgX4= =u0pT -----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5303241B.8080906>