Hi Florent,

sorry for letting you wait.

Incremental full-text updates are on our agenda, but they currently have low priority. As you already indicated, one solution with the current architecture would be to split up your data into more than one database and. Recently updated documents would then be stored in and queried from a temporary db without full-text index, and data that is large and/or static would be indexed, and only optimized if merged with temporary instances.

Christian

On Thu, Jan 2, 2014 at 7:14 PM, Florent Heyworth florent@napalu.ch wrote:

Hi all

I've been experimenting with BaseX - and I'm quite pleased with its capabilities - however I'm hitting a wall where the fulltext index is concerned. Here's the use-case:

• Web application analyzing media (news, etc.) on a large scale
• Frequent inserts (metadata about news) by many clients (ca. 40+ million

documents per year)

• Online Full-text search is a must

Every time data is inserted into a database the full-text index is invalidated. Because of the online full-text search capability requirement a 2-stage update process (store to temporary database, batch update to main database and optimize full-text index afterwards) is unfortunately not an option.

Are there plans to provide for incremental full-text update capacity or is this not an option?

Thanks for any feedback and kind regards

Florent _______________________________________________ BaseX-Talk mailing list BaseX-Talk@mailman.uni-konstanz.de https://mailman.uni-konstanz.de/mailman/listinfo/basex-talk