feature: Allow computing fine-grained/weighted-based scores #11
Labels
lvl-2-medium
Medium-ranking issue
p1-low
Priority 1: Generally no one plans to work on the task, but it would be nice if someone decides to.
t-feature-request
Type: Idea/request of an enhancement towards a library/framework
If you were to compare
battle
andcattle
, they would be 83.33...% similar (5/6). However, if you were to compareBattle
andbattle
, they would still technically be 83.33...% similar as well. Of course though, this type of substitution isn't a change of letters, it's a change of letter case (upper-to-lower & vice versa).What if instead, it'd be possible to have a weighted score? Full letter substitutions have a 1.0 penalty, and letter-case substitutions would have a 0.5 penalty.
This would give a relatively more accurate similarity than before, which would be a 5.5/6 instead, or 91.66...%.
The text was updated successfully, but these errors were encountered: