Re: [Pnet-developers] [bug] resgen produces junk for valid PO files

[Rhys Weatherley]

On Sunday 28 December 2003 12:50 am, Bruno Haible wrote:
> > resgen will interpret all input files, including .po files, as UTF-8,
> > unless you explicitly provide the "--latin1" option.
>
> Since the format of PO files is documented for 8 years in the GNU gettext
> manual, it means that 'resgen' supports only some particular PO files, and
> the developer has to make some extra steps (calling "msgconv -t UTF-8")
> before feeding them to 'resgen'. Can you please update the documentation
> and the "resgen --help" text, so that
>
>        --po-input    or -p
>            Input files contain GNU gettext .po resources.
>
> becomes
>
>        --po-input    or -p
>            Input files contain GNU gettext .po resources in UTF-8 encoding.

The documentation was already updated in CVS when I added the "--latin1" 
option to clearly indicate that the default is UTF-8 unless "--latin1" is 
supplied, and "resgen --help" will now produce the following:

    --latin1      or -l
        Interpret text and .po files as Latin-1 rather than UTF-8.

It would be incorrect to add the UTF-8 qualification on the "--po-input" 
option as its character set is dependent upon the "--latin1" option.  The 
"--latin1" option also affects "--po-output".

Please use the CVS version when testing these kinds of issues.  I've already 
addressed your concerns, including fixing the other .po bugs you found.

I realize that the UTF-8 default may be a little surprising if you are used to 
everything being "whatever my system uses by default", but the behaviour is 
documented and consistent with all of the other .NET tools that use UTF-8 by 
default unless otherwise qualified.

Cheers,

Rhys.