Skip to content

Fuzzy search#13

Open
umitdincel wants to merge 2 commits intoHipo:masterfrom
umitdincel:master
Open

Fuzzy search#13
umitdincel wants to merge 2 commits intoHipo:masterfrom
umitdincel:master

Conversation

@umitdincel
Copy link
Copy Markdown
Contributor

Previous algorithm was only searching for full word. With fuzzy search algorithm we are also able to find words with absent characters in it.

Complex is better than complicated.

when you search cmplx the algorithm is able to identify complex word in this text.

You can see my python code in the link below
https://gist.github.com/umitdincel/89960b4f9c302c604443

@ybrs
Copy link
Copy Markdown
Contributor

ybrs commented Jan 4, 2016

@umitdincel thanks for the pull request, but there are a few things about this.

i. when we make this kind of match, it takes a while to get suggestions. i can't easily type in "ssh myserver" because my history is between ~30k-50k lines. i am not sure if this is the most used case, but it completely ruins my user experience :) that was the main reason i didnt add fuzzy search in the first place.

maybe we can add fuzzy search, if the data is not that much, but that adds inconsistency. or we can add a flag to enable fuzzy search ? what do you think ?

ii. the default is weighting by occurence, but i think this bypasses that - eg: if i have a ton of "ssh someserver" in my history, i expect to see "ssh someserver" but this brings in "ssh someserver someserver" because its a better match :)

iii. also i see many false positive search results with this, see https://www.evernote.com/shard/s29/sh/d753e08c-2b22-439c-b6a3-ae75286a1830/d3d4a2c14b512db0/res/bed44491-9b3c-4cd5-9685-2b9df8b21b44/skitch.png
app.py is the exact match but for some reason its below some app-holidays.py

please don't hesitate to comment or discuss.

thank you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants