Hi spellcheck developers,
as users might not always have the administrative power to change
their ulimit settings (on *nix systems), it might be useful to include
two new parameters
indexMergeFactor and indexMaxBufferedDocs
and their respective setter methods in
org.apache.lucene.search.spell.SpellChecker
Especially, as the spellcheck just finishes with an invalid index without
any further message or warning if the ulimit settings don't suffice.
Below is the code i would suggest.
Thanks,
karin
Code changes:
first the additional attributes and setter methods:
private int indexMergeFactor = 300;
private int indexMaxBufferedDocs = 150;
/**
* Set the merge factor for indexing; default 300
*/
public void setIndexMergeFactor(int mf) {
indexMergeFactor = mf;
}
/**
* Set max no. of docs buffered for indexing; default 150
*/
public void setIndexMaxBufferedDocs(int mbd) {
indexMaxBufferedDocs = mbd;
}
here the changes to public void indexDictionary (Dictionary dict) throws
IOException :
replace the hard coded values
writer.setMergeFactor(300);
writer.setMaxBufferedDocs(150);
with:
writer.setMergeFactor(indexMergeFactor);
writer.setMaxBufferedDocs(indexMaxBufferedDocs);
---------------------------------
LLama Gratis a cualquier PC del Mundo.
Llamadas a fijos y móviles desde 1 céntimo por minuto.
http://es.voice.yahoo.com
