Adjustment to bloom filter size calc, slight refactoring IndexDictionary and Index Shard, Improvements to guessLocaleFromString to handle hindi, more work on nwordgram filters, simplify and speed up each summarizer, also refactor and get sentences more accurately, improve english tokenizer's parser for question answering, tweak some test cases, a=chris

Adjustment to bloom filter size calc, slight refactoring IndexDictionary and Index Shard, Improvements to guessLocaleFromString to handle hindi, more work on nwordgram filters, simplify and speed up each summarizer, also refactor and get sentences more accurately, improve english tokenizer's parser for question answering, tweak some test cases, a=chris