Flexible Queryparser has some Locale violation with lowercasing [LUCENE-3317]

When reviewing the flexible queryparser with my GSoC student, I found some String.toLowerCase() calls without locale. Also some code iterates per-char and lowercases. We should fix this to behave exactly like LowerCaseFilter.

Maybe we can make the queryparser pass all tokens to lowercase through a KeywordTokenizer->LowerCaseFilter chain, but this would create dependency to analysis module.

---
Migrated from [LUCENE-3317](https://issues.apache.org/jira/browse/LUCENE-3317) by Uwe Schindler (@uschindler)

Start in the flexible queryparser entry points mentioned by the issue and locate String.toLowerCase() calls and per-character lowercasing. Compare their behavior with LowerCaseFilter, then determine how to achieve identical token lowercasing while resolving whether a dependency on the analysis module is acceptable; done means locale-correct queryparser behavior.

Written by the indexing model from the issue text.