Sphinx For Full-text Indexing
For a few projects at work, I’ve been using Sphinx as a separate full-text engine. The configuration to get going is about as easy as I could ask for, and indexing is lightning-fast:
collected 28851 docs, 6.1 MB
sorted 6.0 Mhits, 100.0% done
total 28851 docs, 6134126 bytestotal 7.432 sec, 825404.12 bytes/sec, 3882.17 docs/sec
rotating indices: succesfully sent SIGHUP to searchd (pid=27222).
For an example of usage, check out a project I made for aggregating news from a few select sources based on keyword sets: http://feedwords.net/
I realize the similarity to Google’s blog notification system, but this was built to index with less noise.
