In which cases will be words completely ignored?

[Zsolt]

Hi,

I have a strange problem: I have the impression that in some cases (but not
always) sometime words will be completely ignored.

I read plain text files and store most of them with:
  ... Field.UnStored("text", string);

Does lucene really index everything? I understand that some English words
are ignored (but they are always ignored).

I use lucene-1.4.3 under tomcat-5.5.15.

Zsolt