Are we planning to implement CharFilter in the future ? #2800
Description
Is your feature request related to a problem? Please describe.
Yes. The problem is that currently we can't specify a CharFilter like ES when creating index.
Describe the solution you'd like
Add support for CharFilter when build an Analyzer.
Currently only tokenizer and CharFilter are supported during initialize an Analyzer. And then we can implement the following pre-processing by CharFilter when adding a doc to index.
Furthermore, maybe we need to build an interface like Tokenizer for users to implment their own CharFilter. And at the same time we can make several buildin CharFilters.