At 11:24 Uhr -0400 02.07.1998, Chris Nandor wrote: >At 11.16 -0400 1998.07.02, Dave Emme wrote: >>I'm just learning Perl and becoming familiar with MacPerl. I'm confused >>about the DBM facility supplied with the MacPerl distribution. I see >>references to Berkeley DB and don't even understand if that's what I >>currently have... >> >>More specifically: If I attempt to build a database with 10,000 records >>with 256-byte ASCII keys, will this database handle it well, or do I need >>something more robust that the default database? > >Berkeley DB is what is under the hood of the DB_File module. It can surely >handle a DB that big, or bigger. There have been some reports that >creating such a large database can be slow in MacPerl, but I believe key >access is very fast in such a large database. > That's true, but only under the condition that you have enough memory. I recently tried to create an index for an 8MB flat text file, but I had to give it up since I cannot give MacPerl 20MB or more for creating/merging index databases. 10.000 records should not be a problem, though. If it works, it is extremely fast. See the pod contained in DB_File, where you'll find everything you need to know. Note the remark at the end, concerning reopening of a DB file. BTW: Is there anybody out there who could point me to an alternative method/algorithm for generating an index file for full text search, other than using a DB_File. It shouldn't be too memory hungry and index files should be around the size of the original file. The original file's content itself cannot easily be converted into a database. I'd like to be able to access every single word. Every offer is greatly appreciated. Peter Hartmann Visiting Scholar Ritsumeikan University, Kyoto, Japan e-mail: hartmann@mbox.kyoto-inet.or.jp (preferred) phv00542@askic.kic.ritsumei.ac.jp ***** Want to unsubscribe from this list? ***** Send mail with body "unsubscribe" to mac-perl-request@iis.ee.ethz.ch