Underlying mechanism of extractors

[nick-lehmann]

Hello 👋🏻

I'm really intrigued by the extractor mechanism. It is fascinating to me that I can simply create functions with arguments of some type that implements FromRequest and all of the sudden, the data is available inside the function body. However, I have no clue how this works under the hood.

For a personal project (some kind of data loader and test runner), I would like to implement something similar. I wanted to take the extractor mechanism as a base to implement a my own "magic" conversion mechanism. After almost an hour of studying the actix-web source code, I was unable to locate the place where this magic happens. By that I mean the place where, depending on the type of a function argument, some code is generated that calls the SomeExtractor::from_request method and passes this information to the function at the correct argument position. May anyone give my a hint on where to find this?

[robjtede]

A couple of links you may find enlightening.

• https://youtu.be/7DOYtnCXucw?t=1057
• https://docs.rs/actix-web/latest/actix_web/trait.Handler.html

[nick-lehmann]

Thank you very much, that was exactly the guidance I needed 👍🏻