Unfortunately I keep getting 'out of memory' errors when I run this converter script on my local PC (P133,32MB,Win98,ActivePerl).

There is approx. 1GB available as virtual memory but the script bombs out when the swapfile grows to approx. 200 MB.

Is there anything I can do or should I just forget about using (some categories) of the Open Directory as a basis for my Links directory
Running the script via telnet is out of the question as my provider would surely kill me for that ...
Also, is there an easy way to split the rdf-files into smaller chunks ?

Any suggestions are highly appreciated.

Thanks,
Rudolfo
---
Sorry this post was meant to be a reply to http://www.gossamer-threads.com/scripts/forum/resources/Forum11/HTML/000984.html

[This message has been edited by rudolfo (edited September 05, 1999).]

Same thing happened to me. I started to use GREP and it searched within 25 seconds. But, when perl started my whole computer locked up. I hope better when i get Red Hat 6.0 installed.

On a Unix System you can "split" the file ...

split -10000 <infile>

That would make files named "xaa .. xab ... etc" each with 10000 lines from the master file. You can use the convert program one by one on them (or write a shell script to do that for you).

Doing it in smaller chunks would take a lot longer, but would ensure non-use of available memory.

John

Awesome! Will try later