这是indexloc提供的服务,不要输入任何密码
Skip to content

Conversation

@EugenMayer
Copy link
Contributor

@EugenMayer EugenMayer commented Dec 8, 2024

Pull Request Type

  • ✨ feat

What is in this change?

As dicussed recently, we would like the contribute the new Drupal Wiki content collector.

Additional Information

Drupal Wiki content collector for Drupal Wiki on-prem and cloud instances.

Developer Validations

  • I ran yarn lint from the root of the repo & committed changes
  • Relevant documentation has been updated
  • I have tested my code functionality
  • Docker build succeeds locally

@EugenMayer
Copy link
Contributor Author

Anybody that can hint me where possible updates in regard of

Relevant documentation has been updated

the documentation could be needed?

@andul
Copy link

andul commented Dec 9, 2024

Hi @timothycarambat , my colleague implemented the data connector mentioned in Issue 115. Can you please review and advise?

@EugenMayer EugenMayer force-pushed the feature/drupalwiki-collector branch from 182c70d to 2924651 Compare December 9, 2024 17:02
@EugenMayer
Copy link
Contributor Author

Regarding the docs all i could find mentioning the collectors is https://github.com/Mintplex-Labs/anythingllm-docs/blob/d3920da5259e83eb8896eea6266eb95332661f10/pages/beta-preview/active-features/live-document-sync.mdx#scope-of-documents - but those are examples. I can an Drupal Wiki is suggested anyway, let me know

@timothycarambat timothycarambat self-assigned this Dec 11, 2024
@timothycarambat timothycarambat added Integration Request Request for support of a new LLM, Embedder, or Vector database core-team-only labels Dec 11, 2024
@timothycarambat
Copy link
Member

Going to keep this PR open and change it into a data connector plugin when that is live - so that these same things can happen without direct PR review. It will work the same way as agent-skill plugins that are on hub.anythingllm.com

@EugenMayer
Copy link
Contributor Author

@timothycarambat thank you for the heads up. Are data-collectors already suppported on the hub and can be selected? If yes, are there any hints how i need to change the structure to match the marketplace-plugin arch? This way i can take up the work

@timothycarambat
Copy link
Member

They are not supported in the hub yet, but the object and framework exists for it already, we just need to figure out what a data-connector package will look like.

Conceptually it will basically be a bundle just like agent-skills and we will hot load them in the collector so they become available. This includes the UI elements in the data connector UI as well.

So there will be basically

  • config.js
  • handler.js (again)
  • any other supporting files

and it should all work the same way as agents since its super portable when all we are doing is unzipping content into a plugins/dataconnector folder in the storage.

I can share more once its set in stone what it looks like!

@EugenMayer
Copy link
Contributor Author

Thanks. So i assume right now, there is nothing i can do except waiting for either the specs for the new hub-based data-connectors to come out or for you to port it. In addition, the support to install those on the hub must be implemented.

Do you have any timeline on when the specs for such a bundle including the ability to install it via the hub could be ready?

Thank you for the effort

@EugenMayer
Copy link
Contributor Author

I have seen that you now also added a gitlab collector using the old API too. Any reason we could not go the same router with DrupalWiki until the hub is ready to take in official plugins? AFAICS https://hub.anythingllm.com/ does not support collectors yet, and the roadmap at https://docs.anythingllm.com/roadmap does not really give a timeline for this to happen.

Thanks

@timothycarambat timothycarambat removed their assignment Jan 25, 2025
@andul
Copy link

andul commented Feb 25, 2025

Hi @timothycarambat,
my colleague had finished the data collector back in december. We tested the functionality intensively and it works as expected.

uess the community hub for data connectors will take a while. As we really need this connector available in the standard release of AnythingLLM, it would be nice to merge the pull request. As soon the data collector is available we can transfer it to an addon.

@rexjohannes
Copy link
Contributor

They are not supported in the hub yet, but the object and framework exists for it already, we just need to figure out what a data-connector package will look like.

Conceptually it will basically be a bundle just like agent-skills and we will hot load them in the collector so they become available. This includes the UI elements in the data connector UI as well.

So there will be basically

  • config.js
  • handler.js (again)
  • any other supporting files

and it should all work the same way as agents since its super portable when all we are doing is unzipping content into a plugins/dataconnector folder in the storage.

I can share more once its set in stone what it looks like!

Any updates? It has been some time now and we would really like to use this feature. Could you merge this PR for now and we can convert the connector to a plugin later?

@EugenMayer EugenMayer force-pushed the feature/drupalwiki-collector branch from 4c07dc1 to 8d76a7f Compare April 15, 2025 14:07
@timothycarambat timothycarambat self-assigned this Apr 15, 2025
@timothycarambat timothycarambat added PR:needs review Needs review by core team blocked labels Apr 15, 2025
@timothycarambat
Copy link
Member

@EugenMayer - can you allow maintainer edits in this PR? It will be blocked until that is resolved since I cannot make any modifications to this PR without having to make an entirely new PR

@EugenMayer
Copy link
Contributor Author

@timothycarambat would love to, but there is no such checkbox for me on this PR:

image

Not sure why this option is missing.

If it is not easier for you to just do a review and i do the changes, the only workaround i can think of is that you create a PR on source branch : https://github.com/KontextWork/anything-llm/tree/feature/drupalwiki-collector

@timothycarambat
Copy link
Member

Weird, okay well just going to move to #3693

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

Labels

blocked core-team-only Integration Request Request for support of a new LLM, Embedder, or Vector database PR:needs review Needs review by core team

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants