Re: [greenstone-users] Newbie doubts

From Katherine Don
DateTue, 15 Jul 2003 13:40:39 +1200
Subject Re: [greenstone-users] Newbie doubts
In-Reply-To (BAY8-F117ABePddxu2c0004e8c0-hotmail-com)
hi emiliano

> My first problem is about AZList. It doesn't present any HList containing
> A,B,C,...Z links. It simply presents CL2VList but not any hyperlink guide.
> I've read the docs and looked for the examples and it seems to be
> single...I've seen that using AZCompactList the documents are presented
> grouped by the -metadata field but either could see A,B,C or any letter for
> guiding the navigation. Also, I've used but no understand the meaning of (
> -mingroup 1 -mincompact 5 -minnesting 7 -maxcompact 10 ) AZCompactList's
> options.
> My second question is about DateList, perhaps my xml has no the entire date,
> but is presented like a VList classified by year, and the DateList format
> but I want to present it in a HList with year's hyperlinks like some
> examples that I've seen.

both AZList and DateList only present the horizontal links when there are enough documents - these links are only used
when there are too many documents to easily view in a list. so they will appear once you have more than about 40
with AZCompactList, if there are any documents with the same value for the metadata that the list is built on, they
will be combined into a folder in the VList.

you can run 'perl -S <classifiername>' in a command prompt to see what all the options mean (remember to
run setup in the gsdl directopry first).
heres some of the output for AZCompactList:

-mingroup <int> The smallest value that will cause a group in the
hierarchy to form.
Default: 2

-minnesting <int> The smallest value that will cause a list to converted into nested
Default: 20

-mincompact <int> Minimum number of documents to be displayed per page. Default: 10

-maxcompact <int> Maximum number of documents to be displayed per page. Default: 30

> Third, Hierarchy classifiers, I wish to present the number of items that I
> have in the collection for each category, for example for Subject classifier
> I obtain a first node that perhaps has 4 documents or perhaps has 3 internal
> nodes that each contain 2 documents. I wish to tell to the user the number
> of documents that each category has. I've tried using [numleafdocs] but it
> seems to be a metadata field, I need a calculated field, is there a macro or
> a way to do so?

numleafdocs _is_ a metadata field, but it is calculated when the classifier is constructed, and specifies how many
documents are underneath that node in the classification. so this should be what you are after.

> Four, I will not have to treat with death collections, two or tree times for
> week I will have to add documents to the collection. Simply I must to add
> keepold option for retain the old documents when I import and build
> collections? Perhaps will change the way of access to the collection many
> times, allways I've to re-build the collection or only re-build indexes?
> I've seen that If I follow the recommended method of moving the content of
> building and carrying into index I could not re-build again.

you will need to add the new documents into teh import directory, use -keepold when you import, and then rebuild the
entire collection. (you dont need -keepold for build) you cannot use the old indexes so it doesn't matter that
building is copied over to index. what matters is that the archives directory is kept between builds - this avoids the
need to re-import all the source documents again.

> Five, I've seen that Word plug-in sometimes work well, sometimes not, it
> depends of the document. I will recommend to convert to a "light" HTML and
> after some touches (and Section metadata adding), adding to Greenstone, I
> mean the [Source] 'll be HTML. But the finall user perhaps will request the
> doc file (or pdf, or ps). How could I add an "attached" file or how could I
> instruct Greenstone to add it automatically?

I dont think greenstone handles this type of thing. I would recommend reposting this question to the list by itself
just in case someone has done something like this but hasn't read all of this email.

> Six (and last), I wish to export the metadata of all the documents inside
> the collection and posibly to add a separate OAI server for enabling
> searching (I'm trying to put NTLD thesis using Greenstone for internal
> institutional use and another free OAI server for external searching). Is
> there an automated way to export all the docs.xml? I've seen the OAI plug-in
> but is for accessing from Greenstone to an OAI repository, I will need the
> inverse...

greenstone doesn't have any facility for exporting to other formats as far as I know. Try reposting this question to
teh list as well.

hope this helps,
Katherine Don