feat: Strip HTML from rich text search results #3783
Conversation
| @@ -27,6 +33,7 @@ export const useSearch = <T extends object>({ | |||
| useExtendedSearch: true, | |||
| includeMatches: true, | |||
| minMatchCharLength: 3, | |||
| ignoreLocation: true, | |||
There was a problem hiding this comment.
By default, just the first 60 characters are searched which was giving me some unexpected false negative results.
The testing I've done doesn't make this feel much slower performance-wise - will flag this when this PR is tested.
DafyddLlyr
changed the title
|
Removed vultr server and associated DNS entries |
DafyddLlyr
force-pushed
the
dp/search-strip-html
branch
from
a7aa02a to
d80c5f8
Compare
DafyddLlyr
force-pushed
the
dp/search-strip-html
branch
from
1c631c2 to
9298ae6
Compare
| @@ -307,7 +241,7 @@ const defaultFormatter: SearchResultFormatter = { | |||
| getIconKey: ({ item }) => item.type, | |||
| getTitle: ({ item }) => | |||
| (item.data?.title as string) || (item.data?.text as string) || "", | |||
| getHeadline: ({ item, key }) => get(item, key)?.toString() || "", | |||
| getHeadline: ({ matchValue }) => matchValue, | |||
There was a problem hiding this comment.
This is a nice simplification! Rather than try to reverse engineer the match back via multiple different getHeadline() methods, we can just use the original matchValue which fuse.js used.
|
Looks nice and clean without the markdown in the search result :) I do find it a bit odd that several seemingly unrelated strings are sometimes thrown up in bold (e.g |
What does this PR do?
SearchResultCard