Implement a parser simular youtubei.js

[walking-octopus]

Currently, this project uses error-prone and difficult to maintain manual parsing for each response type. Implementing a YouTube.js-inspired parser would help to get rid of unnecessary unwraps, add a layer of abstraction for creating new API endpoints, and make the project more maintainable. Maybe a macro could be made for parsing the InnerTube responses.

[89Q12]

Hi, I know the parsing could be improved but I simply don't have the time currently to implement that. I'm looking at youtubei.js either way hence they figured out deciphering of some parans but if you however want to open a PR or could maybe provide me a UML diagram of a parser you have in mind it woud truly be very helpful. I will look into it either way I just can't promise it will be in the next release I'm planning.

[walking-octopus]

Thanks for the response. I'll look into it. I just wanted to create an issue for tracking the progress.

[89Q12]

So I thought about it and the main problem we would face is that we have to use type Option<T> anyways or we could generate new() for every type since we need to type out the api anyway since I wanna stay idiomatic and not work with the BTree serde_json provides^^ Providing defaults for every type would create the same 'problem' we currently have with all the unwrap calls

[walking-octopus]

I think fetching any data should return a Result so we could propagate any parsing errors.
As for parsing the InnerTube responses, maybe iterators could parse everything, returning an Option or a Result for any missing data.

YouTube.js's parser is mostly typesafe, so maybe some inspiration could be taken.

[walking-octopus]

Is there any kind of InnerTube schema that can be used to validate the JSON?

[89Q12]

Nope, hence I chose the option to parse via serde_json

[89Q12]

But using serde_json has the disadvantage of not given good error messages but is also easy to use