My software looks like this:
[Content]----- {0..* ordered} ->[Classification]
tokenStreamFactory():TokenStream
I need to create a new Field for each Classification of MyDocument.
Every Classification can have an own tokenization scheme, and they will
in some cases use the same field name.
I could not figure out a good way to send the tokens to Lucene, so I
stored the indexOf order as field value and look it up from my analyzer:
for (final Content content : contents) {
Document doc = documentFactory();
PerFieldsAnalyzer perFieldsAnalyzer = new PerFieldsAnalyzer();
Analyzer classifierAnalyzer = new Analyzer() {
public TokenStream tokenStream(String fn, final Reader r) {
return new TokenStream() {
TokenStream ts = null;
public Token next() throws IOException {
if (ts == null) {
ts =
content.getClassifications().get(r.read()).tokenStreamFactory();
}
return ts.next();
}
};
};
int cnt = 0;
for (Classification c : myDocument) {
doc.add(new Field(c.getFieldName(), new String(new char[]{cnt}, ...
perFieldsAnalyzer.add(c.getFieldName(), clazzifierAnalyzer);
}
indexWriter.add(doc, perFieldsAnalyzer);
}
Do I have any alternatives?
What I really want is:
{
for (Classification c : myDocument) {
doc.add(new Field(c.getFieldName(), c.tokenStreamFactory()...
}
indexWriter.add(doc, perFieldsAnalyzer);
}
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
