Here is the final diff which takes care of the discussed changes and also deals correctly with relative url results (eg. the Hein Online links).
Let me know if anyone has any questions. /Mark On Sun, Aug 17, 2008 at 11:38 AM, Mark Aufflick <[EMAIL PROTECTED]> wrote: > Hi, > > The patch as is doesn't stop getting an item when the target url is > missing. As long as there is a valid bibtex url the item will be > added. The logging was really just for checking my xpath was working > right. > > Fair point about overwriting an existing field - I guess at some point > google could add the URL field to the bibtex entry so I'll add a check > for that (I assume the value for the key will be nil if not present, > but I can figure that out). > > I'll post a new patch later with these changes and the relative url change. > > /Mark > > On Sat, Aug 16, 2008 at 11:33 PM, Christiaan Hofman <[EMAIL PROTECTED]> wrote: >> To start: thanks for your efforts, it makes sense to add the URL if >> it's available. I only had a short look at your patch, and the basic >> thing looked OK. Though I wouldn't stop getting an item when the URL >> is missing, and also don't log or return an error, because the URL >> should be optional. Also, I don't think it should overwrite an >> existing field if present. >> >> I don't think the URL should be added as a linked URL AOT a URL field >> (that would be possible through a files: argument of the BibItem >> initializer). We also don't do that for other imports. Moreover it >> will be automatically converted to a linked URL when it's actually >> imported, taking into account the preferences in the Default Fields >> prefs. >> >> Christiaan >> >> >> On 16 Aug 2008, at 3:17 PM, Mark Aufflick wrote: >> >>> Also, I would appreciate some feedback on whether the google scholar >>> result url should go in the URL field (as it is currently) or should >>> only be added as a remote url resource. >>> >>> I'm not yet sure how to do the latter via the BibItem object, but it >>> might be the right way to do it now that I think about it since you >>> don't necessarily want that url becomming part of your bibliography >>> unless it is an online only journal. >>> >>> /Mark >>> >>> On Sat, Aug 16, 2008 at 9:33 PM, Mark Aufflick <[EMAIL PROTECTED] >>> > wrote: >>>> Hi again, >>>> >>>> I have realised that some google scholar result links are relative >>>> urls - I'll work on additional changes to deal with that once I have >>>> finished this essay draft! >>>> >>>> /Mark >>>> >>>> On Sat, Aug 16, 2008 at 5:48 PM, Mark Aufflick <[EMAIL PROTECTED] >>>> > wrote: >>>>> Hi all, >>>>> >>>>> I just started using BibDesk (as I have just become a humanities >>>>> student and have to figure out all this Harvard referencing mumbo >>>>> jumbo!). >>>>> >>>>> It was frustrating me that the google scholar web importing didn't >>>>> add >>>>> the url since I can often use one of the nifty applescripts to >>>>> automatically fetch the pdf from jstor etc. >>>>> >>>>> I'm not sure if goog have cleaned up their syntax since the parser >>>>> was >>>>> first written, but it was fairly straightforward. Each result is >>>>> in a >>>>> <p class="g"> and each such result has (potentially) one BibTex link >>>>> and also (potentially) a target url. The target url isn't always >>>>> there >>>>> for eg. [Book] or [Citation] search results. >>>>> >>>>> So my change is that first the results are extracted, then each >>>>> result >>>>> is parsed for the two components. This adds some overhead, but on my >>>>> G4 12" powerbook it doesn't seem too much different so I'm sure most >>>>> people will notice it even less. >>>>> >>>>> I have attached a preliminary patch to this email that works for >>>>> me. I >>>>> need to tweak the error handling and would appreciate any feedback. >>>>> >>>>> Also - should I be asking for svn commit rights or does someone want >>>>> to commit it for me (once the error handling is done properly of >>>>> course). >>>>> >>>>> Mark. >>>>> >>>>> PS: thanks for the excellent program - it's going to save me a lot >>>>> of time! >>>>> PPS: Sorry if the list eventually gets this twice - I used the wrong >>>>> email first time. >>>>> >>>>> -- >>>>> Mark Aufflick >>>>> contact info at http://mark.aufflick.com/about/contact >>>>> >>>> >>>> >>>> >>>> -- >>>> Mark Aufflick >>>> contact info at http://mark.aufflick.com/about/contact >>>> >>> >>> >>> >>> -- >>> Mark Aufflick >>> contact info at http://mark.aufflick.com/about/contact >>> >>> ------------------------------------------------------------------------- >>> This SF.Net email is sponsored by the Moblin Your Move Developer's >>> challenge >>> Build the coolest Linux based applications with Moblin SDK & win >>> great prizes >>> Grand prize is a trip for two to an Open Source event anywhere in >>> the world >>> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >>> _______________________________________________ >>> Bibdesk-develop mailing list >>> [email protected] >>> https://lists.sourceforge.net/lists/listinfo/bibdesk-develop >> >> >> ------------------------------------------------------------------------- >> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge >> Build the coolest Linux based applications with Moblin SDK & win great prizes >> Grand prize is a trip for two to an Open Source event anywhere in the world >> http://moblin-contest.org/redirect.php?banner_id=100&url=/ >> _______________________________________________ >> Bibdesk-develop mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/bibdesk-develop >> > > > > -- > Mark Aufflick > contact info at http://mark.aufflick.com/about/contact > -- Mark Aufflick contact info at http://mark.aufflick.com/about/contact
BDSKGoogleScholarParser.diff
Description: Binary data
------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________ Bibdesk-develop mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bibdesk-develop
