Dear a,

If your pdf's are in a single branch of the file system and if the naming of 
the pdf's has followed some rules, you should be able using an AppleScript to 
recreate the links, even if none are exported by Papers. I have written such 
AppleScripts, which can reconstruct all links. However, I am following strict 
pdf naming rules, i.e. the cite key is the name of the first pdf file, and then 
_S is added, e.g. for supplementary material. Ex.:  <cite key>.pdf, <cite 
key>_S.pdf, <cite key>_S2.pdf, <cite key>_S3.pdf, ....

I can easily ls (old fashioned Unix command) all files belonging to a given 
cite key and following above naming rules using within the AppleScript 
following shell script command:

set theShellScript to "cd " & quoted form of workDir & " ; ls -1 " & theCiteKey 
& "*." & ext & " | grep -E \"" & theCiteKey & "[_]?[S]?[0-9]*" & "." & ext & 
"\""
set theResult to (do shell script theShellScript)

And then I link all files returned by above shell script for each publication 
(thePub) similar to this:

-- add or insert file links as given by allFilesToFix
set fixedFileLinks to 0
repeat with theFile in allFilesToFix
-- do this outside the tell block
set theMacFileToLink to POSIX file theFile
tell application "BibDesk"
set thePub to contents of thePub
tell thePub
add theMacFileToLink to end of linked files
end tell
end tell
set fixedFileLinks to fixedFileLinks + 1
end repeat

Rather straightforward. A similar algorithm can be made to work given some 
information in the record is consistently used to name the pdf's belonging to a 
record. You would just need to alter the grep expression accordingly.  If no 
pdf naming rules were followed, you would need Paper to export the file name in 
some form or you are lost.

To sum up: It is most likely doable and likely to be relatively easy many 
cases. Since I do not know Papers I do not know what its limitations are, but 
you probably know that best.

If you wish my AppleScript, write to me directly and I am happy to send it.

Just one caveat: With thousands of records abaove approach becomes very slow, 
can take hours and hours. Seems the thread goes to some energy saving mode or 
whatever. But should not be of much concern to your problem now. This getting 
slow problem is something I might anyway have to chase down with probably 
Christiaan ;-).

Despite all the caveats, most likely doable.

Regards,
Andreas



On 23/Mar/2010, at 19:14 , Christiaan Hofman wrote:


On Mar 23, 2010, at 18:50, Adam Megacz wrote:


"Adam R. Maxwell" <[email protected]<mailto:[email protected]>> writes:
Does it write out links to PDF files in any form?

Hrm, looks like it does produce a Local-Url field.  Is there any way I
can get bibdesk to use this field to *copy* (not move) each PDF to the
location I have designated for auto-file purposes?  Once that process is
complete I would be able to discard my Papers archive.


AppleScript.

I guess the only loose end would be the collections; sadly Papers leaves
no trace of this data in its BibTeX or RIS exports.  But if I do the
export one-collection-at-a-time I might be able to reduce the amount of
work needed.


Neither bibTeX nor RIS have support for any collection information.

Thanks!

- a


------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Bibdesk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bibdesk-users


------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Bibdesk-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bibdesk-users

Reply via email to