[greenstone-users] Questions about searching function with using the CGI URL

From feng wang
DateThu Jun 26 03:19:15 2008
Subject [greenstone-users] Questions about searching function with using the CGI URL
Hi,

We want to write a application which can access and manage the collection in the GSDL. We know,that man can use the CGI+URL to implement the search function in any way. And with the Java+Perl makes the functionality of managing the collection possible. We have some questions about the function like searching and getting the document in the GSDL Collection.

1.Man can find a description about the generic CGI arguments in the manual.pdf(page 53). But it appears still some CGI arguments in the URL Query, which i didn't know what its means. I think i must at first make a full map about all CGI in the GSDL Query. Then i can easy build the any query what i will. Do you know how can i get it in Details.
2.GSDL offers two ways to access the Docs in the GSDL(search or Classifier browser). We encountered a problem by building the demo collection. After building the new PDF Documents in the GSDL Collection we wanna know whether it works. So we used the search function in the localsite to access the PDF Doc.

The Problem is that the PDF File(maybe all the files) is automatically converted in the HTML Format. It's great that we can get the HTML File directly. Then we don't need to consider an other software to open the original source in the GSDL Collection. But for our Application we want the both(converted HTML and the original source)!!! In the Result-Page of search/query there is just one converted HTML. And we can only in the Browser page find the both source.
We thought it is just the question about the design style for displaying the docs. So hope someone can us, how we should do,thanks.

3.GSDL supports various Format of the Metadaten. So we selected the DC as standard by building the doc in the GSDL Collection. After building the Docs in the collection the GSDL GLI will give the ex-Metadaten by itself. Problem is at some time the ex.titel and dc.titel are not equal after the building phase(Not only the dc.titel and ex.titel). And the result-page for the search function displays only the context of ex.titel not the dc.titel. We understand why is the ex.titel displayed. What should we do to prevent the problem.

4.We konw that after sending the query in URL Format we get a metalist about all the docs which match the query and we can access any doc in the list. But we want more!! Our teacher hopes that the GSDL can offer the metadata like creator, page,year, description about the single doc additionally by Searching. Does anyone know what we should do to realize this function. Adding some CGI arguments in the URL Query extra??

5.How can we build an URL+CGI Query to access and search the Docs in/across the several Collections in the GSDL Server. With the CGI Argument c=collection1+collection2 or some else?

6.For Test i have installed the GSDL 3 in my Linux. But i found a problem that until now i didn't get a right solution. The GSDL 3 is successfully installed only when i packed the Java in the folder of /opt!! I have tried to give the right Java path by launching the install command. But the Install Package can't find the Java JDK or JRE! I know that i asked it in the mailing list yet. Please help me , i must find the answer because i should install the GSDL 3 in our Server and could install the Java in /opt too! I am not the Admin!

Regards
Feng
--
GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen!
Jetzt dabei sein: http://www.shortview.de/?mc=sv_ext_mf@gmx