preserve_original.md

preserve_original

preserve_original and preserve_original_recorder are token filters that allow to keep and index original terms. This is very similar to the keyword_repeat and unique filters but will work also on filters that do not support the keyword attribute.

Examples

index :
    analysis :
        analyzer :
            preserve_case :
                type : custom
                tokenizer : whitespace
                filter : [preserve_original_recorder, lowercase, preserve_original]

Will produce the following terms hello, Hello, the, world, World for the input text Hello the World.