One question:
The similarity class is abstract. Are there default implementations like in
other parts of this API (Analysers for example) available and how can I use
it i.e. to calculate weights? Are there some default implementations hidden in
other classes (i.e. in Hits) to calculate the term weighting. Does somebody
have some code which demonstrates the usage?
Well, nothing is "hidden" when you have a nice IDE :))
In IntelliJ, I did this:
- Ctrl-n, typed "simi" followed by return, which took me to Similarity.java
- Ctrl-h
Two subclasses appeared: DefaultSimilarity, and TestSimilarity.SimpleSimilarity.
So there are two subclasses within Lucene's codebase, one used for testing. To see it's usage, have a look at TestSimilarity. The test cases are a great place to look for "documentation".
As for weighting - the scoring process uses weighting based on the type of query - look at the Query class and its subclasses to see those details.
Erik
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
