Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: rewind keyword search #1330

Open
wants to merge 5 commits into
base: main
Choose a base branch
from

Conversation

shaik-zeeshan
Copy link
Contributor

related: #1271
/claim #1271

Screen.Recording.2025-02-11.at.3.50.14.PM.1.mp4
Screen.Recording.2025-02-11.at.4.13.59.PM.1.mp4

Copy link

vercel bot commented Feb 11, 2025

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Comments Updated (UTC)
screenpipe ✅ Ready (Inspect) Visit Preview 💬 Add feedback Feb 11, 2025 10:55am


// Add conditions for each keyword
if fuzzy_match {
conditions.extend(keywords.iter().map(|_| "o.text LIKE '%' || ? || '%'"));
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

isn't the FTS 100000x faster than LIKE?

O(log(n))

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah you are right

changing the implementation to use FTS

@louis030195
Copy link
Collaborator

trying to understand

is it full refactor of timeline or different page in it?

like ideas here, however the query is way too slow to generate suggestion, most people don't have this attention span

how do you extract these logos? this is quite interesting as some people ask for "universal file search" or like typical use case:

"ok im trying to find my rent document is it in dropbox, google drive, gmail, icloud, somewhere else...?"

and also this is a use case for teams, for example if my friend saw a document and i want to find it, it would be so useful to be able to search and find it in seconds

#1293 related

this is kinda out of scope of the initial search idea i wrote but kind of extension of it

we've also explored idea of using models like https://github.com/microsoft/OmniParser to index images/logos more precisely than OCR for the future FYI

PS: cannot test right now - focused on #1318

@shaik-zeeshan
Copy link
Contributor Author

shaik-zeeshan commented Feb 12, 2025

is it full refactor of timeline or different page in it?

it is different page

however the query is way too slow to generate suggestion, most people don't have this attention span

using local model for suggestions, maybe I could generate suggestions based on app_name without using models.
so it doesn't too long (app name based filtering). also since this is keyword search, the generated suggestion might not have any data

how do you extract these logos? this is quite interesting as some people ask for "universal file search" or like typical use case:

getting links from text and fetching favicon from google

"ok im trying to find my rent document is it in dropbox, google drive, gmail, icloud, somewhere else...?"

and also this is a use case for teams, for example if my friend saw a document and i want to find it, it would be so useful to be able to search and find it in seconds

this is cool, maybe we could use models to get related info or use embeddings for this

Copy link

vercel bot commented Feb 12, 2025

Someone is attempting to deploy a commit to the louis030195's projects Team on Vercel.

A member of the Team first needs to authorize it.

@louis030195
Copy link
Collaborator

hows the AI settings config?

Screenshot 2025-02-13 at 8 38 39 AM Screenshot 2025-02-13 at 8 38 23 AM

@louis030195
Copy link
Collaborator

Screen.Recording.2025-02-13.at.8.41.04.AM.mov

also would appreciate if u can fix the top bar navigation

@shaik-zeeshan
Copy link
Contributor Author

Screenshot 2025-02-13 at 8 38 23 AM

looking into this

also would appreciate if u can fix the top bar navigation

sure

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