Hi Alexander,
On 06/29/2011 10:03 AM, Alexander Wagner wrote:
[...] it would be handy if I could just drop a file to
etc/bibformat/output_formats and be done. I can ticketize this in
case you agree.
Not sure if it is really worth: it should already be the case that you
can simply drop an output format in etc/bibformat/output_formats dir.
It will just miss the following bit of information (stored in the DB):
1. The content-type used to serve the output format.
2. The visibility bit switching on/off the output format on the user
search interface (hidden does not mean that the output format is
not available: it is just not in the list).
3. Names and description (mostly useful when format is visible).
(1) is the most annoying but can be worked around in some cases by
using the fact that some default content-type is used depending on the
first letter of the output format code:
- 'h': text/html
- 't': text/plain
- 'x': text/xml
(2) and (3) can be "worked around" thanks to the fact they are more or
less only here for information (this is mostly true for (3)), and also
because these values are not often changed: if you keep the output
format files in some repo and periodically deploy the changes by
copying the files to your production environment, there are chances
that most of the time you want to keep (2) and (3) as they were on the
production server (don't change the values in the DB).
It is also not frequent that somebody wants to add new output formats,
compared to the need to update them (which is in this case well covered
with the current implementation).
Of course one can imagine moving (1), (2) and (3) inside the output
format files. But do you see any use case for which the above is not
exactly suitable?
Best regards
--
Jerome Caffaro
