feat: data sources #2262

Open
opened 2025-11-11 15:03:39 -06:00 by GiteaMirror · 15 comments
Owner

Originally created by @tjbck on GitHub (Oct 2, 2024).

https://developers.google.com/drive/picker/guides/overview

https://learn.microsoft.com/en-us/onedrive/developer/controls/file-pickers/js-v72/?view=odsp-graph-online

  • Google Drive
  • Notion
  • Dropbox
  • Confluence
  • Onedrive/Sharepoint
  • NextCloud
Originally created by @tjbck on GitHub (Oct 2, 2024). https://developers.google.com/drive/picker/guides/overview https://learn.microsoft.com/en-us/onedrive/developer/controls/file-pickers/js-v72/?view=odsp-graph-online - [ ] Google Drive - [ ] Notion - [ ] Dropbox - [ ] Confluence - [ ] Onedrive/Sharepoint - [ ] NextCloud
GiteaMirror added the enhancementcore labels 2025-11-11 15:03:40 -06:00
Author
Owner

@flefevre commented on GitHub (Oct 2, 2024):

+1 for onedrive and SharePoint
If it could be accessible for the workspace feature at team level
And also directly from a chat, Additionnelaly to the upload file feature.

@flefevre commented on GitHub (Oct 2, 2024): +1 for onedrive and SharePoint If it could be accessible for the workspace feature at team level And also directly from a chat, Additionnelaly to the upload file feature.
Author
Owner

@rgaricano commented on GitHub (Oct 2, 2024):

I don't now if it's usefull, since files are read, analized & vectorized to "db", ...., I think that, right now, there isn't a verification for uploaded docs, & it do the process again .... expending resources

But just in case this is half done with pipelines (RAG sample)
& pyGooAPI: https://github.com/googleworkspace/python-samples/tree/main/drive/driveapp
pyOnedrive: https://github.com/OneDrive/onedrive-sdk-python
pydropbox: https://pypi.org/project/dropbox/ https://github.com/dropbox/dropbox-sdk-python

by the way, notion have AI, (and probably plug for connect to ollama API or compatible, i don`t know, obsidian have it & have connection to notion too ... another way to do it)

@rgaricano commented on GitHub (Oct 2, 2024): I don't now if it's usefull, since files are read, analized & vectorized to "db", ...., I think that, right now, there isn't a verification for uploaded docs, & it do the process again .... expending resources But just in case this is half done with [**pipelines**](https://github.com/open-webui/pipelines/tree/main) [(RAG sample)](https://github.com/open-webui/pipelines/tree/main/examples/pipelines/rag) & pyGooAPI: https://github.com/googleworkspace/python-samples/tree/main/drive/driveapp pyOnedrive: https://github.com/OneDrive/onedrive-sdk-python pydropbox: https://pypi.org/project/dropbox/ https://github.com/dropbox/dropbox-sdk-python by the way, notion have AI, (and probably plug for connect to ollama API or compatible, i don`t know, obsidian have it & have connection to notion too ... another way to do it)
Author
Owner

@rgaricano commented on GitHub (Oct 6, 2024):

tool for access google drive: https://openwebui.com/t/suciocerdo/google_drive/

@rgaricano commented on GitHub (Oct 6, 2024): tool for access google drive: https://openwebui.com/t/suciocerdo/google_drive/
Author
Owner

@jeandelest commented on GitHub (Oct 16, 2024):

Maybe also think about GitHub / GitLab connector like it's done in anythingllm ?

@jeandelest commented on GitHub (Oct 16, 2024): Maybe also think about GitHub / GitLab connector like it's done in anythingllm ?
Author
Owner

@Lanhild commented on GitHub (Oct 28, 2024):

  • DBMSs (PostgreSQL, etc.)
@Lanhild commented on GitHub (Oct 28, 2024): - DBMSs (PostgreSQL, etc.)
Author
Owner

@harikrish-s commented on GitHub (Nov 1, 2024):

Relevant: #1293

@harikrish-s commented on GitHub (Nov 1, 2024): Relevant: #1293
Author
Owner

@longfei-zhang commented on GitHub (Nov 28, 2024):

+1 for confluence

@longfei-zhang commented on GitHub (Nov 28, 2024): +1 for confluence
Author
Owner

@taylorwilsdon commented on GitHub (Dec 16, 2024):

hey @tjbck - I have created a fully functional Google drive implementation that I think looks very clean, will submit a PR after I clean up the debug logging. Any preference on capturing google drive client id + secret (env var vs a config option in app ui?)

https://github.com/user-attachments/assets/5d0f201f-1768-43d9-bfbe-dfe5bc17d93e

@taylorwilsdon commented on GitHub (Dec 16, 2024): hey @tjbck - I have created a fully functional Google drive implementation that I think looks very clean, will submit a PR after I clean up the debug logging. Any preference on capturing google drive client id + secret (env var vs a config option in app ui?) https://github.com/user-attachments/assets/5d0f201f-1768-43d9-bfbe-dfe5bc17d93e
Author
Owner

@gabriel-ecegi commented on GitHub (Dec 16, 2024):

Question might be what is data source:
A) Source for syncing embeddings
B) Instant Retrieval/Embedding per chat

For me it is A.
Starting discussion here https://github.com/open-webui/open-webui/discussions/7888

@gabriel-ecegi commented on GitHub (Dec 16, 2024): Question might be what is data source: A) Source for syncing embeddings B) Instant Retrieval/Embedding per chat For me it is A. Starting discussion here https://github.com/open-webui/open-webui/discussions/7888
Author
Owner

@flefevre commented on GitHub (Dec 18, 2024):

Beautiful vidéo for the Google drive integration!
Well-done.

About the UX, you put a button to add manually a file into the chat but perhaps we could also imagine to add 'a knwoledge' or a Google file/directory button ? So you can create a knowledge directly to a Google drive directory? Or to add files into a knowledge from a Google drive?

@flefevre commented on GitHub (Dec 18, 2024): Beautiful vidéo for the Google drive integration! Well-done. About the UX, you put a button to add manually a file into the chat but perhaps we could also imagine to add 'a knwoledge' or a Google file/directory button ? So you can create a knowledge directly to a Google drive directory? Or to add files into a knowledge from a Google drive?
Author
Owner

@flefevre commented on GitHub (Dec 18, 2024):

Any idea about the difficulty to do exactly the same for SharePoint ?

@flefevre commented on GitHub (Dec 18, 2024): Any idea about the difficulty to do exactly the same for SharePoint ?
Author
Owner

@papiche commented on GitHub (Feb 20, 2025):

In my opinion & user case, it would be great to retriever documents from self hosted solution NextCloud

I am not an expert, but maybe could be done using
https://pypi.org/project/webdavclient3/
or https://pypi.org/project/nc-py-api/

Let me know what do you think about this idea ?

@papiche commented on GitHub (Feb 20, 2025): In my opinion & user case, it would be great to retriever documents from self hosted solution **NextCloud** I am not an expert, but maybe could be done using https://pypi.org/project/webdavclient3/ or https://pypi.org/project/nc-py-api/ Let me know what do you think about this idea ?
Author
Owner

@flefevre commented on GitHub (Feb 20, 2025):

Whatever décision you take, please add an admin dashboard to autorise or not such drive to enhance security and avoid data leak. Thanks in advance

@flefevre commented on GitHub (Feb 20, 2025): Whatever décision you take, please add an admin dashboard to autorise or not such drive to enhance security and avoid data leak. Thanks in advance
Author
Owner

@frenzybiscuit commented on GitHub (Aug 9, 2025):

Can we get integration in for nextcloud, if at all possible?

@frenzybiscuit commented on GitHub (Aug 9, 2025): Can we get integration in for nextcloud, if at all possible?
Author
Owner

@Pounii25 commented on GitHub (Sep 18, 2025):

NextCloud integration will be awesome for our private data <3

@Pounii25 commented on GitHub (Sep 18, 2025): NextCloud integration will be awesome for our private data <3
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#2262