search

OpenSlides / openslides-client

Webclient for OpenSlides 4+
MIT License
3 stars 33 forks source link

Enhancement: Fuzzy match for Participants search #2601

Open colinmahns opened 1 year ago

colinmahns commented 1 year ago

Very small but helpful enhancement for my organization would be a fuzzy match for accented characters like é to show up when searching for e. US English keyboards do not have easily accessible accent marks like European keyboards and require a few confusing workarounds to pull up these characters.

Names like Renée don't come up with the current exact match which is typed out when pulling up Renee, making it appear a participant is not present, or is a cause of confusion between tech support and participants when in-person where we can't copy and paste. Seating a backup delegate mid-debate with a similar accented name ended up taking a minute longer than was comfortable as it was as a vote was about to launch.

jsangmeister commented 1 year ago

In the near future, the global search will be re-introduced in OS4. This will support fuzzy search. See https://github.com/OpenSlides/openslides-search-service for more details.

However, we see that filtering in the list view has additional advantages. We will talk about if and how this could be tackled.

bastianjoel commented 3 months ago

Instead of fuzzy matching on the list views we just want to filter character compositions. Fuzzy search should always also include a match ranking which would require further consideration for our already existing sort logic. Also as we now support fuzzy search in the global search.

This could for example be achieved by converting a string to its canonical decomposition and removing specific characters like such: str.normalize('NFD').replace(/[\u0300-\u036f]/g, "") before comparison.