Re: Bug in xgettext?

[Bruno Haible]

Hi,

Jakob Damgaard Møller wrote in
<https://lists.gnu.org/archive/html/bug-gettext/2019-10/msg00009.html>:
> i have this issue:
> 
> I have a bundle.js, output from a bigger react / typescript webpack project.
> 
> If i run: xgettext bundle.js --from-code=utf-8 --language=JavaScript it
> finds some of the translated strings, but not all of them.
> 
> I have tried to change --language to different others: PHP, Python, Perl, C
> etc. and I am still missing translations.
> 
> Is there a way to debug xgettext to see what it is doing step by step?
> 
> Or anyone have an idea to find all translated strings in my bundle.js ?
> Bundle.js - https://paste.ee/p/XmML7

This URL is expired.

> And one of the strings that are skipped, is the string: "Control
> Information"
> 
> I can find gettext("...") and gettext('...') 352 times in the bundle. But I
> only get 216 strings in my message.po file after running xgettext command
> above.
> This is the part of the full bundle.js, that contains the string: "Control
> Information"
> 
> https://pastebin.com/CuXKrDKC

This URL works, but it is minified JavaScript. xgettext is supposed to be used
on source code, that includes comments. Rationale: This is needed in order to
extract comments meant for the translators.

Also, please try to reduce the reproducible example. I'm not going to search
for syntactic oddities in a 100 KB large one-line program.

Bruno