it already is really great with the links, so would be awesome to have the same for searches!
And I wish all of the search was combined into one feature.
I love the way that nValt - https://brettterpstra.com/projects/nvalt/ - handles search.
There is only one box for typing - Search or Create. It not only searches the titles (like CMD+O in Obsidian), but also searches in files. The search results give precedence to matching files while still displaying results that only contain matches in files.
If your search DOES come back with a file name, pressing enter will open that file, otherwise it will create it (again, like CMD+O in Obsidian). Behavior is slightly different here, and, unlike Obsidian, you can create notes that are substrings of other notes (By pressing backspace after it auto-completes the name).
All of this combines to a reduced cognitive load when editing notes. If I’m watching a video about, say, MongoDB databases, I don’t have to know whether I already have existing notes about that topic, I just start typing “Mongo,” and if nothing comes up, I can press enter to create a new note. Seconds saved every time you open the app add up!
totally agree with you.
I hope to use several #tags as filters in a single search.
I have some tags as basic ideas and other tags as topics. Now I’m not able to search 2 tags simultaneously if they are not next to each other in a note. This problem would be solved if there were FUZZY SEARCH in ‘search in files’.
I really NEED this.
I need this.
This was implemented
I don’t see it ether. Any chance you can provide more details @WhiteNoise? Thanks!
this unimplemented feature prevent me coming back to obsidian, I use obsidian for my second brain, but what’s point if I can’t search easily to find my note? current search function is not enough
Where is it?
@miguel @jackiexiao I am not sure I understand what you want.
Quick switcer performs a fuzzy search over filenames
The global search is fuzzy by default over filenames and contents and supports operator to refine/change the behaviour.
What is that you want exactly?
A fuzzy search returns a list of results based on likely relevance even though search argument words and spellings may not exactly match. Eg. searching for “misisipi” should return results for “mississippi”. Google does a great job of this is their search.
I think there is a second element though and that is search ranking. If you search for “Mississippi” results should be ranked by a combination of title, content, links, tags, date created and date modified eg. a result that has the title “Mississippi” should be highly ranked and similarly one that has the word many times in the content should be more highly ranked that one that only has it once. In an ideal world, semantic understanding could also add context by knowing a word is a city or a river or a song as well and suggest these contexts to further refine the search.
Perfect, I understand now.
Of the fuzzy search obsidian does:
- Substring matching
- Insertion (cot → coat)
It doesn’t do well deletion (coat → cot) and substitution (coat → cost).
Please, open a feature request for this.
The ranking of the results is whole other subject.
Here is the relevant FR https://forum.obsidian.md/t/sort-search-results-by-relevance-and-what-relevance-is/5933
Bear in mind that there is a heavy computation cost on performing these actions on the whole vault (filenames, content and metadata).