Notion Org Chart Generator

Thank You for trying out my actor and the feedback!

1. I'm afraid it is the limitation of a Notion API I'm using in the actor. But there is a possibility for a lengthy workaround - scrape the URL. Note: If only URL is given and if the database is private, then it cannot be opened, so the solution wouldn't be perfect either.
2. I updated the actor and now user can input either ID or the URL.
3. Updated.
4. Based on image I think the problem is the Relation input. Maybe the relationName in the input does not correspond to the name of the column.

Cheers

Just checking, @tuan, you are planning to update number 4. so that it works as expected? Can number 1, be automated? I mean creating custom integration via the actor?

Thanks for the feedback. I tried it again, and it's much better now. But there is still one problem. Becaus the "Name" is entered and "Team lead" is using a list of Notion users, the names sometimes don't match. See run ID [REMOVED]. Not sure how to fix this though... it's hard to make sure the names will match

@zuzka, 4: based on what @jan just sent, I'm thinking about a little heuristics which can join people, for example Matej Sochor and Matěj Sochor (based on diacritics). Or I could add something like 'possible prefixes and postfixes', like we had '__name' or 'ZX name'. I'm not sure how to approach this, because it is not actor's fault for not recognizing same people, because it was given different names. I think this should be fixed on the users' side. What do you think, guys? AIso I've just thought of an opposite problem. What if two different people have the same name :D 1: yes, it can be automated, I'm going to work on it now.

That's a great idea! Maybe when you remove diacritics and there's no perfect match, you could find the name that is closest using https://www.npmjs.com/package/fast-levenshtein. And report if there's no good enough match, which would mean some user is missing. Some heuristic like that.

Regarding 1), it took just a few seconds to set up the integration, so it's good enough IMHO, unless simplifying it would be something easy.

Thanks again, I think this will be a nice useful actor.

Thanks for the suggestion! I added diacritics nad levenshtein (I'm not sure what threshold to use, so I set it to 2) and also check for nicknames (like, Hoang "spidoosho" Anh Tuan).

To be honest, I would rather skip 1) too. But if customers will want it, it's no problem to implement it.

Tried it again, but it fails now: