chriggi wrote:
There is now only one workflow issue left: I collect references with CiteULike and import them from time to time to Bookends. Now, if I import them all and end up with tons of duplicates.
My workaround is to tag (give a keyword) references with a temporary identifier (like numbers _4) for the next import. Not very nice, though.
Funny. I had exactly the same problem and thought of this solution too. It's nice that it is so easy to remove duplicates in Bookends, but I think a better solution would be at the CiteULike end, if it were possible to show -- and therefore export -- only the references gathered in the last x days in CiteULike. I'll email the guy who runs it and ask if he has thought of doing that.
Jon wrote:
Bookends expects the first item in the Volume field to be the volume, and a subsequent entry to be the issue.
You can fix this by
1. Ignore the issue # (of course, you may not want to).
2. Put the Issue # in a user-defined field instead of the Volume field.
3. If you can, write a Perl script (or similar) to reorder the issue and volume output in the BibTeX file (beyond my knowledge).
Thanks for the very quick response, Jon. Neither (1) nor (2) much appeals -- (1) could get my PhD thesis rejected, and (2) would require reformatting all my existing article references.
I don't know any Perl, so unless someone reading this does and is kind enough to post a script, that's out.
For the moment, then, I'm on option (4): tidy up imported references by hand.
I've discovered since my last post that some tidying is also required for the date field if CiteULike has stored a month and a year, because it formats them as eg 'May 1999' soetimes, at least, not in proper Bookends style as yyyy-mm-dd. I don't see any way round this.
Isn't the problem with issue before volume a general one for BibTeX imports, though, since the fields in a BibTeX reference don't seem to have to come in any particular order? If so, it would be very nice to have a fix in Bookends. (I guess this is a feature request.)