Removing commas when using whitespace tokenizer

Question

When using the whitespace tokenizer a text like "there, he is." would be split to "there," "he" and "is.". Naturally I would want to remove those punctuation that the standard tokenizer would had removed automatically.

My questions are:

1. How to trim those punctuation marks? (in elasticsearch setting, like adding another token filter or charfilter)
2. I need to use the whitespace tokenizer mainly because I don't want hyphenated words to be splited. Is there a way I can achieve this while still using the standard tokenizer?

Answer 1

You can use the char filter to remove the the ",". Char Filter