database driven search

[wnoordhoek]

Are there plans for databaseserver driven (for example mysql) search methods in the future?

I would see the advantage of custom queries for statistics generation and ofcourse current and dynamic statistics

the statistics page would get loaded and the most recent statistics are pulled up without having to download the searchlog file analyse it, rename the files and then upload the generated files.....

with the database u can have users submit their own interpretation of queries for statistics, or graphical or just textual without u having to think of the queries :D



i know that right now the own database is portable over different systems (i would expect this as an argument y u dont build it in ;))
and ofcourse i could figure the build up of the files and write a php script myself for statistics generation on the fly, but it would be a great deal easier in a DB like mysql (since im into php, i guess it would be mssql or access db)

in the db there could also be a list of pages and for example pdf files that got indexed (see my thread about pdf's not getting indexed ;) )
for checking if every page or file got indexed (wich are wanted to be indexed)

just venting some ideas i got on the fly :D

thxn for reading

[Ray]

There are no plans for using a database server such as MySQL for the server-side searching at this point.

The main reason is that our internal data format (the .zdat files) is already optimized using techniques that are similar to the way a relational database works. This is why Zoom produces such fast results. And combined with the fact that we avoid the overhead of having to connect to a database, we can produce even faster results than a database driven search engine.

Other reasons are cross platform capabilities (as you mentioned), portability (moving between servers or web hosts), and the simplicity of not requiring a database setup.

However, what you're actually thinking of is more related to the ability to store the search LOG in a database. This does not depend on how the index data is stored or whether Zoom uses a database for searching.

The search log is stored entirely in a text CSV (Comma Seperated Value) file. This means that you can actually import the file into say, Microsoft Excel, or Access, etc. and perform personalized statistic queries. You can really do anything at all with this data, from sorting search terms alphabetically, to retreiving all search terms that contain certain words, draw charts, ... the list goes on.