Re: iso-codes .pot msgid strings contain non-ASCII characters

[Paul Eggert]

Alastair McKinstry <address@hidden> writes
re <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=330990>:

> Two partial solutions I can think of, are:
>
>     (1) transcode in Glibc  (e.g. in the iso-codes-utils code, gnome, etc. ),
>     so the algorithm is:
>         - read "?land Islands" from XML country list
>         - transcode in C to "Aaland islands"
>         - look up in .po file
>     With comments in the .pot file, so that the translator sees:
>
>     # AX, ?land Islands
>     msgid "Aaland islands"
>     msgstr ""

I suppose that might work.  However, you should ensure that the
comments use the correct encoding for the .po file in question.  This
might not be UTF-8.  It might be easier just to put in an ASCII
comment saying that the msgid has been ASCII-ized.

Can the format of the XML country list be extended to contain two
spellings, one in UTF-8, one ASCII-ized?  Then the algorithm wouldn't
need to transcode.  The ASCII-ized spelling could be optional to save
space.

> or,
>
>     (2) Ensure in the .po files shipped, for these entries in particular, 
> there
>     are no empty
>     entries, e.g. a script to do:
>     # ,fuzzy
>     msgid "?land Islands"
>     msgstr "Aaland Islands"
>
>     wherever there was an empty msgstr entry.

I don't see how that would work, since gettext msgids should be ASCII.

I'll CC: this to address@hidden, to give them a heads-up on the problem.