After exporting data in ISO file in manageable number, and reimporting them
in Winisis, we were able to isolate the records with errors. Actually,
Winisis automatically cleaned our data for us, removing those with errors
when we imported it back to our database.
I am happy to say that we were able to import 4405 records out of 4700 into
The strange characters on some of the records were causing the errors.
We are looking at the records with errors in our backed-up database that
contained 4700 records. We will be editing them so that we can recover the
300 records we lost during the import/export. Surely, this a lot better than
just having 29 records :)
On Mon, May 18, 2009 at 10:36 AM, Katherine Don <firstname.lastname@example.org>wrote:
> Are the documents you get valid records? Or are they just chunks of
> multiple records?
> Do you need to set the split_exp option to ISISPlugin?
> -split_exp <regexp> A perl regular expression to split
> files into segments.
> Default: r?
> Maybe its not detecting the record boundaries properly?
> Have you tried setting GLI to expert mode to see if there are any errors
> Ramon Sampang wrote:
> We have been trying to import our Winisis data to Greenstone 2.81 running
> on Linux but wea are having problems importing all the data. We have around
> 4,000+ records accessible in Winisis. But after creating a collection in
> Greenstone we only end up with 29 documents.
> We are already using the patch GLI.jar for 2.81, and we have already tried
> exploding the metadata, but still the same number of documents. There is no
> explicit error during the process.
> Could anyone help us figure out how to solve this problem?
> greenstone-users mailing email@example.com://list.scms.waikato.ac.nz/mailman/listinfo/greenstone-users
-------------- next part --------------
An HTML attachment was scrubbed...