On Thu, 29 Mar 2007, Jonas Sandman wrote:
I guess I could split it into more tables, but would that really make anything faster?
It would make upgrades, changes, and searching faster.
The whole key is the filename, I query for files in specific filepaths, so it doesn't matter if something is mis-spelled. Rather I encourage the possibility that it can be.
So you never query by artist, title, genre, or any other field in the table?
So what should I use as a primary key then? The path?
That's what you wrote that you use. I would make the title the primary key, unless you have the same title by different artists. In that case, I'd use a numeric key.
Since one file is one row in the database I rather liked the idea that I didn't have to scatter around updating a lot of tables. I am using the database instead of a large memory table like I used before.
Then you might as well use a spreadsheet. With a relational database, multiple tables are related by foreign keys (which sqlite3 implements as triggers). If you have an interface other than the sqlite3 command line editor then you enter information in the appropriate data entry widgets and use your middleware code to store/retrieve from appropriate tables. -- Richard B. Shepard, Ph.D. | The Environmental Permitting Applied Ecosystem Services, Inc. | Accelerator(TM) <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863 ----------------------------------------------------------------------------- To unsubscribe, send email to [EMAIL PROTECTED] -----------------------------------------------------------------------------
