>>>>> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> Jean-Marc, what do you think of this approach/code? It enables Angus> us to generate a valid UTF-8 encoded i18n.php3 with minimal Angus> human intervention. Angus> If you're happy, then I'd like to commit this. (To the devel Angus> tree too of course.) Angus> The attached patch adds a comment of the form: Angus> # HTML-Last-Translator: Zdenko Podobný Can't you just use a tool like recode to put the name in the right shape? I do not know how common recode is, but grep Last-Translator src/lyx/1.3.x/po/eu.po| sed 's/^.*Last-Translator: //' |recode -d l1..h4 gives something that is not too bad. And you can get the charset from the po file too. Angus> However, I see that someone has manually removed the entries Angus> for bg.po, ca.po, cs.po, he.po, hu.po, pt.po, sv.po, tr.po and Angus> wa.po from the web page (all the "red" entries). Nothing has been manually removed. The i18n.php3 runs the script on POFILES, which contains the list of .po files that we actually distribute. It happens that my criterion for choosing these translations is indeed not being in red. Angus> Rather than do that, I modified the generated .php3 file to Angus> only print out "non-red" table rows. Compare Angus> http://www.devel.lyx.org/i18n.php3 and Angus> http://www.devel.lyx.org/~leeming/www-devel/i18n.php3.) They Angus> look the same, but my local version requires no manipulation at Angus> all. Everything is done by postats.sh. No, I want the table to reflect what is actually supported by LyX. The rest is kept in cvs for reference. JMarc
