That was too fast. There is a problem with the new format. It works
just fine for standard output for one layer, but it has broken some
modules and wxGUI which rely on one line per layer in the form
<layer/name>;<table>;<key>;<database>;<driver>
<layer/name>;<table>;<key>;<database>;<driver>
etc.
That was too fast. There is a problem with the new format. It works
just fine for standard output for one layer, but it has broken some
modules and wxGUI which rely on one line per layer in the form
<layer/name>;<table>;<key>;<database>;<driver>
<layer/name>;<table>;<key>;<database>;<driver>
etc.
That was too fast. There is a problem with the new format. It
works just fine for standard output for one layer, but it has
broken some modules and wxGUI which rely on one line per layer
in the form
<layer/name>;<table>;<key>;<database>;<driver>
<layer/name>;<table>;<key>;<database>;<driver>
etc.
ah, yes. it's easy to forget about layer 2+ ...
(see also r.univar + multiple input maps, similar problems)
btw, please do not add _() i18n to module output sent to stdout,
scripts parsing the result get confused. only for G_message()s.
btw, please do not add _() i18n to module output sent to stdout,
scripts parsing the result get confused. only for G_message()s.
Hmm, original and restored behaviour is to print non-i18n for -g and
i18n for -p output. The -p output does go to stdout but is much more
difficult to parse (translated or not) than -g output. IMHO, it's ok
to say that -g output must be parsable whereas -p output must be
readable, also translated if possible?