Custom Tokenizer - preserving previous tokens

Elastic Stack Elasticsearch
1

Does anyone know if it's possible to tokenize the following string as such?

"myField": "123-4-5678-9012-3"

I want my tokens to be:

123
123-4
123-4-5678
123-4-5678-9012
123-4-5678-9012-3

I've been playing around w the EdgeNGram and pattern tokenizers but they don't do quite what I want.

I guess my other option is to store the variations in the source prior to ingestion.

2

Have you tried the path hierarchy tokenizer?

3

THATS IT!!

Thanks Christian