←back to thread

Parse, Don't Validate (2019)

(lexi-lambda.github.io)
389 points melse | 6 comments | | HN request time: 0.913s | source | bottom
Show context
Attummm ◴[] No.27640130[source]
That's how the validation tool for Python Maat works. By creating a completely new dictionary. https://github.com/Attumm/Maat
replies(2): >>27640197 #>>27640762 #
1. mulmboy ◴[] No.27640762[source]
How does Maat compare with pydantic? https://github.com/samuelcolvin/pydantic
replies(2): >>27640874 #>>27640983 #
2. rmnclmnt ◴[] No.27640874[source]
If only for conciseness, readability and speed, I'd take Pydantic over any day. Being able to express 80% of type checking using Python native type hints + dataclasses is just so intuitive!

And it's getting some wide adoption, for instance FastAPI which uses it for request validations.

replies(1): >>27641020 #
3. Attummm ◴[] No.27640983[source]
Pydantic is using classes and typehinting. The new dataclasses style. Currently Maat doens't have a parser for dataclasses, it could come in the future. Pydantic works great with typehinting.

Maat was created before dataclasses existed. For validation Maat offers the same. But it also allows for some really neat features such as validation on encrypted data. https://github.com/Attumm/Maat/blob/main/tests/test_validati...

Since validation is written as dictionaries its possible to store the validations in caching db such as Redis.

And since its simple its easy to extend for anyone use case. And there are no other dependencies.

Benchmarks of pydantic has Maat around twice as Pydantic.

replies(1): >>27646242 #
4. Attummm ◴[] No.27641020[source]
Engineering is about tradeoffs, even though both projects do validation.

The points you made are all very valid points.

At my employer we use both projects. If the data is very nested, or really large Maat is used.

replies(1): >>27641056 #
5. rmnclmnt ◴[] No.27641056{3}[source]
Mmmh interesting requirement! Indeed, defining very nested structure with Pydantic is one of its weaknesses.

And of course I agree 100% about tradoffs in engineering. However I usually advise against using 2 dependencies doing mainly the same thing if possible within the same project.

Anyway, good catch, thanks for enlightening me!

6. Attummm ◴[] No.27646242[source]
Unable to change my comment.

Benchmarks of pydantic has Maat around twice the speed of Pydantic