Hello Mike, thank you for the answer. Currently I hold this datastructure in a HashTable in Memory but the ressource consumtion is very high. Lucene looks easy in the way of usage and is suposed to be extremely perfomant. I though also to use later the abilities to lucene to attach parameters to search results so I do not neccessarily need to process the record.
I checked berkley db but it is not free for comercial use. What would be a implementation for btree? Thank you, Tobias > On 13-Dec-07, at 3:26 PM, Tobias Rothe wrote: > >> I got a quick question. I am handling hughe CSV files. They start >> with a key in the first column and are followed by data. >> I need to retrieve randomly this data based on the key. So it is >> kind of a search where I give a unique key and ideally access to >> the right line. >> The file contains about 200.000 lines or more. I am not sure if >> Lucene can handle things like that and I did not really find a hint >> on this topic. So I hope to find help here. > > Is that all you're doing? You could accomplish the above with Lucene > but it isn't really needed for that. You need some kind of on-disk > key->value mapper. Something like a berkeley db hashtable or btree > should work (store each line as a key/value pair). > > -Mike > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
