what is the way to translate Markdown-formatted text with po4a? Explicit type 'markdown' is not supported, and the only mention of Markdown I found is in /usr/share/perl5/Locale/Po4a/Text.pm, my Perl-fu is not enough to dig harder.

Am 27.01.2015 um 13:39 schrieb Jonas Smedegaard: > Here's how to enable the mardown support: > > po4a-gettextize -f text -o markdown ... > Admittedly, first I thought this command is completely bogus... I've never read about Markdown support in po4a before. Some things are hidden under the hood, and the somewhat strange man page hierarchy of po4a doesn't even make things easier. But indeed, it exists and works fine! Many thanks for the hint!

Jonas Smedegaard ☫ → To Sergey Korobitsin @ Tue, Jan 27, 2015 13:39 +0100 >> what is the way to translate Markdown-formatted text with po4a? >> Explicit type 'markdown' is not supported, and the only mention of >> Markdown I found is in /usr/share/perl5/Locale/Po4a/Text.pm, my >> Perl-fu is not enough to dig harder. > > Thanks for your interest in po4a. > > Here's how to enable the mardown support: > > po4a-gettextize -f text -o markdown ... > > Hope that is enough of clue for you. Thanks! I'm using integrated po4a binary, but I've already found a way to pass the option ("opt:" in config file).

> Also, I hope you don't mind that I shared my response with the public > po4a mailinglist so that others might stumble across our conversation > too. Of course. P.S. As I've found, not all features of Markdown are supported, the first one is old style headers (==== and ----). But they are preserved in PO file, so that's not a big problem.

Jonas Smedegaard ☫ → To Sergey Korobitsin @ Wed, Jan 28, 2015 09:04 +0100 > Hmm, not sure if you mean to say that it isn't working or that you > solved the problem on your own. Obviously I hope it is the latter > :-) You pointed me to an option of po4a-updatepo, and I don't use it directly, but via main po4a executable. I've solved the problem with your help, thanks :-)

>> P.S. As I've found, not all features of Markdown are supported, the >> first one is old style headers (==== and ----). But they are >> preserved in PO file, so that's not a big problem. > > My goal was to parse not all Markdown markup but the least possible > to not break when tumbling it through po4a. OK, that's clear to me. > As I recall, I found old-style headers to cause problems treated as > plain text - what (non-big) problem do you see with it? A I see, non-sophisticated translator can omit the '=====' and '----' lines and leave only translated header itself, and at the output we'll get broken structure. Then, it can be workarounded by some kind of Translator's HOWTO, but people are people, and it would better if this case processes automatically.

Jonas Smedegaard ☫ → To Sergey Korobitsin @ Wed, Jan 28, 2015 09:42 +0100 >> A I see, non-sophisticated translator can omit the '=====' and >> '----' lines and leave only translated header itself, and at the >> output we'll get broken structure. Then, it can be workarounded by >> some kind of Translator's HOWTO, but people are people, and it would >> better if this case processes automatically. > > I don't follow how current loose parsing leads to broken structure - > i.e. how it is an actual (not just an academic) problem. Can you > elaborate on that - perhaps give an example of how it can break (even > if, as you already wrote, the breakage is minor)? See, I have a Markdown document started as follows: Область применения документа ============================ Данный документ является руководством для администраторов платформы «ARTA SYNERGY» (далее Система), предназначенной для совместной работы на предприятиях и управления по целям. Здесь описываются процессы, связанные с выполнением задач администрирования Системы. Then, when I apply po4a, I get following content at PO file: #. type: Plain text #: admin-manual.md:54 #, fuzzy, no-wrap #| msgid "Область применения документа" msgid "" "Область применения документа\n" "============================\n" msgstr "Scope of document" And then, what will do some forgetful translator - just remove the "fuzzy" mark. If so, I'll get the following in output document: Scope of document The document is guide for Administrators of the ARTA SYNERGY platform (hereinafter referred to as the System) designed for groupwork on enterprises and target-oriented management. It describes processes connected to execution of the System administration tasks. And, bingo! I've got broken stucture of document. This problem not about bugs, but about usability, and its minor. Maybe I've got some professional deformation, because my scope of work for now is usability :-)