Also related to #398067, I think we should fix intltool-update to allow registering any arbitrary file types to scan rather than just a predefined fixed set.

I have tried to use the same solution as for bug #398067, but no success: after adding this line to POTFILES.in:

[type: gettext/xml]plugins/project-wizard/templates/cpp.wiz

intltool-update -m doesn't complains, but intltool-update --pot bails out with the following:

gabor@gabor-laptop:~/anjuta/po$ intltool-update --pot

junk after document element at line 8, column 0, byte 165 at /usr/lib/perl5/XML/Parser.pm line 187
/usr/bin/xgettext: error while opening "../plugins/project-wizard/templates/cpp.wiz.h" for reading: Nincs ilyen fájl vagy könyvtár
ERROR: xgettext failed to generate PO template file. Please consult
       error message above if there is any.

Perhaps these .wiz files are not well-formed xml?

I think a proper xml file mandates only one root node in a file which these .wiz files do not have (they have several root nodes). I don't recall what the reason was but Sebastien can tell more.

I don't remember any particular reason for having several root nodes. I have probably done this because it's simpler and works fine with GMarkup parser. I think I can change the format to put everything in one root node, but I will need a several days to do this change.

This problem has been fixed in the development version. The fix will be available in the next major software release. Thank you for your bug report.

Closing all old bugs. Apologies for the spam.