How has SO implemented the tagged search? Is it using Lucene or any other open-source search engine library for tagged searching?
What is the best way to search document (PDF, XML, HTML, MS Word) or database?
A:
So, yes, it is using Lucene.NET, though I'm not sure exactly how. The "best" way is a whole 'nother story.
Matthew Flaschen
2009-04-26 05:21:13
A:
Searching tags is very different than searching text. A tagged search is searching for an association where questions are all associated with a particular tag. This can be implemented with a full-text engine where the tags are all appended in a single large entry, but a relational database will probably be best in this situation (assuming the tagged data is in a relational database to start with).
For searching other documents like PDF, XLS, HTML, then you need full text like Lucene. You'll need a parser that can extract just the relevant text from each source (i.e., separate text from markup).
Sam
2010-01-14 02:47:24
A:
The last time this was discussed (on the podcast) it was mentioned that Stackoverflow uses SQL Server's full-text search feature, not Lucene.
Don
2010-01-14 02:54:08
A:
SO doesn't use Lucene.
If you want to index documents and are running Windows, then IFilters would be my first choice.
Michael Stum
2010-01-14 02:59:43