I often have to create very large (hundreds of millions of records) but simple read-only databases, with sorted keys and values. I usually use BerkeleyDB for the database, and then build my own objects and methods on top of BerkeleyDB.pm. BerkeleyDB is very simple and fast, and much more suitable for these problems (IMO) than a full-blown DBMS.
I always want to retrieve a range of indexed keys and their associated values, so using tie is ineffective. Similarly, performing a (one-time) fetch of all the keys is much too inefficient. Using BerkeleyDB.pm works fine, and yet ...
... it seems that there should be a more generic way to perform these indexed-based retrievals, which accepts a range of index values as input, and allows you to iterate over the associated keys(indices) and values. Is there such a module out there?
In reply to fast+generic interface for range-based indexed retrieval by jae_63
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |