On 15 Mar 2008, at 7:20 PM, A wrote: > Hi all, > > I've been using Bibdesk with great pleasure. I work in computer > science, where DBLP is kind of an authoritative source for > bibliography. > > Hence my request: is it possible to add DBLP > (http://en.wikipedia.org/wiki/Digital_Bibliography_%26_Library_Project > , http://www.informatik.uni-trier.de/~ley/db/) support in Bibdesk? > Ideally, it would be a search group, but could also be as a webgroup. > I have no idea on how to implement such a feature, but the way this > site works is simplistic, so it should be no pain at all to implement > it and it would bring so much added value to all of us who are working > with BibDesk in Computer Science. > > For those who are not familiar with DBLP, here is how it works. The > result page of a query is very basically structured with HTML tables, > one row per result. Now this result page is inconsistent depending on > the type of search you did. With an author search (from > http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/index.html > or by only putting an author in > http://www.informatik.uni-trier.de/~ley/db/indices/query.html) > , you get a clickable number as the first element of the row. For > example, searching Alan Turing with either method gives the page > http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/t/Turing:Alan_M=.html > Clicking on the number, for example 2, leads to the corresponding > bibtex entry, for example > http://dblp.uni-trier.de/rec/bibtex/journals/jsyml/Turing37 > On the other hand, with a title search, things are a bit more > tricky. For example, searching the previous publication (from > http://www.informatik.uni-trier.de/~ley/db/indices/t-form.html > or by putting a title in > http://www.informatik.uni-trier.de/~ley/db/indices/query.html) > , we obtain the following page > http://www.informatik.uni-trier.de/ley/dbbin/dblpquery.cgi?title=computability+and+lambda+definability > Here, the number is not clickable, but we have the DBLP entry. The > part between "[DBLP:" and "]" is what you have to add to > http://dblp.uni-trier.de/rec/bibtex/ > to find back the page with the bibtex entry (that is > http://dblp.uni-trier.de/rec/bibtex/journals/jsyml/Turing37) > . > > It might be that there are two bibtex entries in the bibtex page (the > one starting by http://dblp.uni-trier.de/rec/bibtex/), when there are > cross-references. > > So I would see two ways of integrating this in BibDesk. As a search > group, it would be ideal, BibDesk would just serve as an interface to > the query page, and present the results (parsing the bibtex entries as > described). > > Or as a webgroup, it could just parse the result page. > > So my questions are: can this be accepted as a feature request? Is it > something that can easily be implemented in BibDesk? Also, I can't > find any documentation on webgroups. Is it something which is hard > coded in the source code of BibDesk? Is there any easy way to add a > webgroup? > > Thank you very much for having read this long email. > > A
This does not look like something that could be put into a search group. Those work for services that have a query syntax that returns machine-parseable bibliography info (AOT human readable pages). This looks more like something for a web group scraper, similar to what we have for google-scholar. Of course you can add this as a feature request. But whether it will be implemented depends wholly on anyone wanting to take it on, and chances for that are pretty low ATM. Unless of course you (or anyone else) want to take a shot at it yourself, we're happy to assist ;-) You could take our Google Scholar scraper as a starting point. Currently we don't have an SDK to add define new web group plugins. Adam was talking about it, and I do think it would be cool to have. But I don't think it is very difficult to write code directly for a scraper, compare <https://bibdesk.svn.sourceforge.net/svnroot/bibdesk/trunk/bibdesk/BDSKGoogleScholarParser.m > Christiaan ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Bibdesk-develop mailing list Bibdesk-develop@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bibdesk-develop
