Thread: [Refdb-users] Producing modified bibtex output.
Status: Beta
Brought to you by:
mhoenicka
Menu
▾
▴
refdb-users
|
From: Josef K. <jo...@ta...> - 2007-07-13 00:31:43
|
Having trouble understanding how to create a custom bibtex file using
refdb.
At the moment I'm running:
runbib -d joe -S bibtex-abbrev -t bibtex out
which creates out.bib:
@ARTICLE{ Burd1988,
YEAR = {1988},
AUTHOR = {Burd, A. B. and Barrow, J. D.},
TITLE = {Inflationary Models with Exponential Potentials},
JOURNAL = {Nucl.Phys.B.},
VOLUME = {308},
NUMBER = {4},
PAGES = {929-945}
}
and so on.
I want to add an extra field to each bibtex record, called 'EPRINT'
which would carry the value of one of the refdb tags (although I'm not
sure which one yet), perhap the UR.
How do I go about doing this? Is it a modification to the style xml
file, or do I need to hack code?
Joe
|
|
From: Markus H. <mar...@mh...> - 2007-07-13 07:08:45
|
Hi Joe,
Quoting Josef Karthauser <jo...@ta...>:
> I want to add an extra field to each bibtex record, called 'EPRINT'
> which would carry the value of one of the refdb tags (although I'm not
> sure which one yet), perhap the UR.
>
> How do I go about doing this? Is it a modification to the style xml
> file, or do I need to hack code?
>
The mapping of RIS fields to bibtex fields during data export is =20
currently hard-wired. UR and L1-L4 are being exported as "URL =3D {}" =20
entries. Some fields, most notably M1-M3 and U1-U5 are currently =20
ignored (which is not a good thing).
I believe the mapping of RIS fields to bibtex tags should remain =20
hard-wired on the server side, as RefDB is designed as a multi-user =20
system. The tag names could then be fixed on the client side using sed =20
as every user sees fit.
Support for U1-U5 could be added in no time if it is ok to export them =20
to the pseudo-bibtex fields U1-U5 (which you could then change on the =20
client side to something meaningful). I've been reluctant to export =20
M1-M3 because their contents depend on the reference type. I'm =20
currently in the process of disentangling these along the lines of the =20
proposed rbib data format (see =20
http://refdb.sourceforge.net/rbib.html). I'm not going to switch to an =20
internal rbib representation in 0.9.9 yet, but some features are =20
likely to get implemented anyway. This holds true for the M1-M3 fields =20
which will be split into several fields according to their true =20
meaning. Once that is implemented, it will be a snap to export them =20
properly to bibtex.
regards,
Markus
--=20
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de
|
|
From: Josef K. <jo...@ta...> - 2007-07-13 09:11:07
|
On Fri, Jul 13, 2007 at 09:08:36AM +0200, Markus Hoenicka wrote:
> Hi Joe,
>=20
> Quoting Josef Karthauser <jo...@ta...>:
>=20
> > I want to add an extra field to each bibtex record, called 'EPRINT'
> > which would carry the value of one of the refdb tags (although I'm not
> > sure which one yet), perhap the UR.
> >
> > How do I go about doing this? Is it a modification to the style xml
> > file, or do I need to hack code?
> >
>=20
> The mapping of RIS fields to bibtex fields during data export is =20
> currently hard-wired. UR and L1-L4 are being exported as "URL =3D {}" =20
> entries. Some fields, most notably M1-M3 and U1-U5 are currently =20
> ignored (which is not a good thing).
>=20
What's the first version which exports the URL tags? I'm using 0.9.6
currently, as when I upgraded to 0.9.8 I couldn't get the refdbc to work
- it complained about not being able to connect to the database, despite
the refdbd running without startup errors.
Joe
|
|
From: Markus H. <mar...@mh...> - 2007-07-13 09:30:03
|
Quoting Josef Karthauser <jo...@ta...>: > > What's the first version which exports the URL tags? I'm using 0.9.6 > currently, as when I upgraded to 0.9.8 I couldn't get the refdbc to work > - it complained about not being able to connect to the database, despite > the refdbd running without startup errors. > I believe 0.9.7 did export the URL tags. I can see the code in the =20 first subversion revision after we migrated from CVS to subversion. =20 This was a month or two before 0.9.7. However, instead of looking for older versions I'd recommend to fix =20 the 0.9.8 installation problems. I'm pretty sure this is a setup =20 problem caused by the migration of 0.9.6 to 0.9.8, and you might bump =20 into the very same problem when attempting to upgrade to 0.9.7. =20 Usually these problems can be fixed as long as the debug logs of =20 refdbd are available. Would you mind to reinstall 0.9.8 and send in =20 the refdbd log with the log level set to 7? regards, Markus --=20 Markus Hoenicka mar...@ca... (Spam-protected email: replace the quadrupeds with "mhoenicka") http://www.mhoenicka.de |
|
From: Markus H. <mar...@mh...> - 2007-07-13 09:32:41
|
Quoting Markus Hoenicka <mar...@mh...>: > Would you mind to reinstall 0.9.8 and send in > the refdbd log with the log level set to 7? Just to make sure: use 0.9.8-1 instead of 0.9.8 as the latter contained a few bugs. regards, Markus -- Markus Hoenicka mar...@ca... (Spam-protected email: replace the quadrupeds with "mhoenicka") http://www.mhoenicka.de |
|
From: Markus H. <mar...@mh...> - 2007-07-16 21:06:38
|
Markus Hoenicka writes:
> The mapping of RIS fields to bibtex fields during data export is
> currently hard-wired. UR and L1-L4 are being exported as "URL = {}"
> entries. Some fields, most notably M1-M3 and U1-U5 are currently
> ignored (which is not a good thing).
>
The bibtex backend now exports the U1-U5 fields as USER1 = {} and so
on entries. This is available as of revision 417.
regards,
Markus
--
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de
|
|
From: Markus H. <mar...@mh...> - 2007-08-05 15:21:27
|
Picking up an old thread...
The new prerelease (0.9.9-pre3, see the separate announcement on this
list) now supports the export of the contents of the U1-U5 and M1-M3
fields into aptly named bibtex fields.
regards,
Markus
Markus Hoenicka writes:
> Markus Hoenicka writes:
> > The mapping of RIS fields to bibtex fields during data export is
> > currently hard-wired. UR and L1-L4 are being exported as "URL = {}"
> > entries. Some fields, most notably M1-M3 and U1-U5 are currently
> > ignored (which is not a good thing).
> >
>
--
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de
|
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X