In article <[EMAIL PROTECTED]>,
 Christiaan Hofman <[EMAIL PROTECTED]> wrote:

> On 7 May 2008, at 10:33 PM, Adam R. Maxwell wrote:
> 
> > In article <[EMAIL PROTECTED]>,
> > James Howison <[EMAIL PROTECTED]> wrote:
> >
> >> On May 6, 2008, at 9:55 PM, Adam R. Maxwell wrote:
> >>
> >>>
> >>> On May 6, 2008, at 6:17 PM, James Howison wrote:
> >>>
> >>>> I usually use Copy as RIS to exchange references with the poor sods
> >>>> using Endnote.  However recently (I think) I've noticed that  
> >>>> BibDesk
> >>>> doesn't include the TY (type) of reference, which confuses  
> >>>> Endnote no
> >>>> end.  I'm pretty sure that RIS should always have that filled out:
> >>>>
> >>>> http://www.refman.com/support/risformat_tags_01.asp
> >>>
> >>> You are absolutely right, and I just checked in a fix for it; it's
> >>> been broken for several months.  I only tested it with your example,
> >>> so more testing is welcome.  Incidentally, I fixed another random  
> >>> bug
> >>> in RIS export that's been around since March 2006, due to a change  
> >>> to
> >>> support PubMed's bastard RIS.  Apparently RIS doesn't get a lot of
> >>> use...
> >>
> >> Hi,  using today's daily I tried this with my correspondent.  I see
> >> that TY is now included but he still had issues.  We both realize  
> >> that
> >> RIS is poorly defined, and it might be an Endnote importing thing.
> >
> > RIS at least has a published spec; unfortunately, BibDesk is doing a
> > crappy job exporting it.  I never noticed this because BibDesk can  
> > read
> > its own output back in (and I never tried Endnote).
> >
> > Basically the problem is that BibDesk has a dictionary mapping RIS  
> > tags
> > to BibTeX, so TI = title; T1 = title; T2 = title; and does a reverse
> > dictionary lookup to find that title = (TI, T1, T2), and picks the  
> > first
> > one.  However, those are actually in random order, so the result  
> > varies
> > wildly.
> >
> > What really needs to be done:
> >
> > 1) create a dictionary mapping BibTeX field->RIS tag
> 
> Just added this one, so that should be better now. Though I'm not sure  
> I've done it right.
> 
>      PubMedTagsForBibTeXFieldNames = {
>          Abstract = AB;
>          Address = AD;
>          Annote = N1;
>          Author = AU;
>          Booktitle = BT;
>          Editor = ED;
>          Isbn = SN;
>          Journal = JO;
>          Keywords = KW;
>          Language = LA;
>          "Local-Url" = L1;
>          Medlineref = UI;
>          Pages = PG;
>          Title = TI;
>          Publisher = PB;
>          Status = PS;
>          Url = UR;
>          Volume = VL;
>          Year = PY;
>      };

I'm not sure either, but that will at least give consistent results :).  
Hopefully someone has the Copious Free Time to check against the spec.  
We have a mix of PubMed and RIS in the type dictionaries, unfortunately, 
and that adds some extra confusion.


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Bibdesk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bibdesk-users

Reply via email to