substring fragmentation algorithm

[xdxxxdx]

Hello,

I have some difficulties in understanding the substring fragmentation algorithm.
e.g. I have a company's name is: "ACKERMANS & VAN HAAREN" ;
The substring fragmentation property is set to the company name.
Root fragment is full.
The bucketizer 'A' is not full yet, so the member should be ONLY added under http://localhost:8080/kbo/by-name?substring=a right?

Cause what I got how is the member is added under:

Actually, from what I understand from the https://github.com/Informatievlaanderen/VSDS-LDESServer4J/blob/main/ldes-fragmentisers/ldes-fragmentisers-substring/README.MD
Each member should be only added one time to the fragmented stream right?

Thanks

[Yalz]

Hi @xdxxxdx ,
I do believe this is intended behavior. It considers every word as a separate part to do the fragmentation on.
Since your name is "ACKERMANS & VAN HAAREN", it will put it in a bucket of "A", "&", "V" and "H".

[Yalz]

@WLefever-Cegeka @ddvlanck feel free to correct me

[Tomvbe]

This is correct, each member is added one time per token. Looking back at the readme I agree that the example can cause some confusion. This is because the use of tokens was added later to the algorithm. I will make a PR to improve the example description.

[xdxxxdx]

Thanks a lot for the reply.

[Tomvbe]

cfr. #781