Information générale
Catégorie:
Base de données, Tables, Vues, Index et syntaxe SQL
>The data comes in as all text. Even “numeric” data is in text format. Getting rid of any padding doesn’t seem to help much because I had created an ASCII Delimited Text file from my 2 Gig dbf and the text file was almost 2 Gig itself.
Perhaps I did not make myself clear.
(Text) "Numerics" can be "packed". For example, if the Zip code is a separate Field, of say 9 or 10 digits, it can be converted to a 4 byte binary field; eg.
X = BINTOC( VAL( "123456789" ) )
You've just reduced the storage requirements for zip codes by over 50%; the same applies for other numerics.
I'm not familiar with the contents of your text files, but if street names, building names, etc. are being duplicated, they should be spun off into a separate table and reduced to a simple binary key in the main table.
You can't assume that other vendors have simply taken some "off the shelf" solution; you may have to develop a custom pack-and-search engine (as they might have), or license one.
>Do you know of any other database who’s data is this compressed? Paradox (DB) or whatever?
Paradox extensions are .DB, and their indexes are "huge", often bigger than the data.
Paradox supports binary fields, as does VFP. You cannot simply work with only text if you hope to reduce storage requirements.
Précédent
Suivant
Répondre
Voir le fil de ce thread
Voir le fil de ce thread à partir de ce message seulement
Voir tous les messages de ce thread
Voir tous les messages de ce thread à partir de ce message seulement