v0.4.1

Want to migrate? We have prepared a guide to help you migrate from v0.3.

Release highlights

• Extended search - this allows you to put special patterns to further narrow down your search
• We now have two fuzzy algorithms for you. "v2" (default) gives you better highlighting for matched characters while "v1" compensates that for speed.
• You can turn off fuzzy matching to do an exact match on strings instead
• If scores of two matches are tied and you prefer one to appear above the other, you can provide a tiebreaker to do this. The library ships with two of them.
• If matched characters appear at more than one place in the string, you can now choose whether to highlight the characters that appear at the end of the string or the ones that appear at the start. This is controlled by forward.

Breaking changes

• options.cache is removed. As a result, no caching is performed.
• options.maxResultItems has now been renamed to options.limit
• normalize is now on by default. This means we'll automatically remove diacritics/accents from Caffè to make it Caffe for example. This can be toggled off in options.
• forward is now on by default. This means earlier querying "abc" on "xabc yabc" used to highlight last occurrence (abc of yabc). Now first occurence (abc of xabc) will be highlighted instead. This can be toggled off in options.
• positions is now a Set. Earlier it could either be an array or a null.
• Result item is now flattened. This means entry.result.score will now be written as entry.score. Same applies for start and end.

Fixes

• The package now correctly resolves as ESM in Node.js environments