lucene/sandbox/contributions/indyo/docs/introduction.txt

30 lines
1.3 KiB
Plaintext

Indyo is a datasource-independent Lucene indexing framework.
What this means, is that Indyo allows a myriad of sources from which
data is fed to the search engine to be indexed. Datasources can take
the form of traditional storage mediums (filesystem, database, web
site, etc), objects, complex datasources which consist of a mixture of
objects and storage medium, and pretty much anything which implements
com.relevanz.indyo.IndexDataSource. If it's a file that's being
indexed (via com.relevanz.indyo.FSDataSource), the contents of the
file can be indexed by a class which implements
com.relevanz.indyo.contenthandler.FileContentHandler (e.g.
TextHandler, ZIPHandler, etc). Via the datasource, applications can
also associate a search result object with the object that was indexed
(or optionally use Peter's SearchBean contribution), for display
purposes.
To summarize, if you:
a) Want a way of indexing various sources of data, and even nested
datasources (like indexing a HTML file, which spawns a custom
datasource, say RemoteHTMLDataSource, for every link it encounters)
b) Simply want a pluggable system of indexing different types of file
content (currently plain text, Zip, Tar, GZip file formats are
supported, but writing new file content handlers are easy)
then Indyo may be worth checking out.