Last active
March 3, 2025 11:45
-
-
Save evenfrost/1ba123656ded32fb7a0cd4651efd4db0 to your computer and use it in GitHub Desktop.
Fuse.js with highlight
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
const highlight = (fuseSearchResult: any, highlightClassName: string = 'highlight') => { | |
const set = (obj: object, path: string, value: any) => { | |
const pathValue = path.split('.'); | |
let i; | |
for (i = 0; i < pathValue.length - 1; i++) { | |
obj = obj[pathValue[i]]; | |
} | |
obj[pathValue[i]] = value; | |
}; | |
const generateHighlightedText = (inputText: string, regions: number[] = []) => { | |
let content = ''; | |
let nextUnhighlightedRegionStartingIndex = 0; | |
regions.forEach(region => { | |
const lastRegionNextIndex = region[1] + 1; | |
content += [ | |
inputText.substring(nextUnhighlightedRegionStartingIndex, region[0]), | |
`<span class="${highlightClassName}">`, | |
inputText.substring(region[0], lastRegionNextIndex), | |
'</span>', | |
].join(''); | |
nextUnhighlightedRegionStartingIndex = lastRegionNextIndex; | |
}); | |
content += inputText.substring(nextUnhighlightedRegionStartingIndex); | |
return content; | |
}; | |
return fuseSearchResult | |
.filter(({ matches }: any) => matches && matches.length) | |
.map(({ item, matches }: any) => { | |
const highlightedItem = { ...item }; | |
matches.forEach((match: any) => { | |
set(highlightedItem, match.key, generateHighlightedText(match.value, match.indices)); | |
}); | |
return highlightedItem; | |
}); | |
}; | |
// usage: | |
const res = highlight(fuse.search(text)); // array of items with highlighted fields |
This is great, many thanks!
My two cents: While implementing on jekyll I saw a couple of issues:
-
Sometimes
regions
are unordered: (e.g.[[3,4],[1,2]]
), that breaks the layout. I think this happens whenuseExtendedSearch: true
. -
My database has strings with html, and I saw that if the
class
is captured on the search (e.g, searching formyclass
would hit on"<div class='myclass'>Some text</div>"
the layout breaks as well (would result in something like<div class='<span class="${highlightClassName}">myclass</span>'>Some text</div>
I solved both as follows:
const generateHighlightedText = (inputText: string, regions: number[] = []) => {
let content = '';
let nextUnhighlightedRegionStartingIndex = 0;
// Sort regions to avoid breaking layout
regions = regions.sort(function(a, b) {
return a[0] - b[0];
});
regions.forEach(region => {
if(nextUnhighlightedRegionStartingIndex > region[0]){
return content;
}
const lastRegionNextIndex = region[1] + 1;
// Try escape html in database
const theInput = inputText.substring(region[0]);
const indexIniTag = theInput.indexOf("<");
const indexCloseTag = theInput.indexOf(">");
// Case of first '>' before first '<', we are inside of '<...>'
if (indexCloseTag < indexIniTag) {
return content;
}
// Corner case when we're in the last html tag of the string, since indexIniTagCase not found and then -1
if (indexIniTag == -1 && indexCloseTag > 0) {
return content;
}
content += [
inputText.substring(nextUnhighlightedRegionStartingIndex, region[0]),
`<span class="${highlightClassName}">`,
inputText.substring(region[0], lastRegionNextIndex),
'</span>',
].join('');
nextUnhighlightedRegionStartingIndex = lastRegionNextIndex;
});
Quite newby in Javascript but this did the work
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The type on the regions is actually a bit off. And in React, any leading or trailing spaces are ignored, so you lose spaces between matches.. Here's adapted version that solved the problems for me for use in react.