mkdocstrings VS pydantic

Mkdocs has the mkdocstrings plugin, offering limited automated API documentation capabilities. It is however nowhere near as capable as Sphinx' autodoc, missing granularity in its configuration, limited intersphinx-like cross-referencing support, and essential features like documentation of inherited members, or the ability to manually describe objects if needed.

mkdocstrings works well, although it is not as powerful as the API documentation in Sphinx.

Author of Materia for MkDocs here. MkDocstrings [1] implements automatic generation of reference documentation from sources. It's language-agnostic, actively maintained and currently supports Python [2] and Crystal [3]. It also integrates nicely with Material for MkDocs.

[1]: https://mkdocstrings.github.io/

Some time has passed since I first introduced mkdocstrings here on reddit. If you don't know what mkdocstrings is: it's the equivalent of the autodoc Sphinx extension, but for MkDocs, a Markdown static site generator. It works differently though, and supports multiple languages by design (not only Python). Someone actually wrote a very good handler for the Crystal language, and another user on GitHub recently expressed their interest for writing one for Go.

Write proper docstrings as you go along (every time you write a new class/method/function you can document what it’s doing as you’ll know why and what from the pattern you chose). Using a tool like mkdocstrings makes maintaining documentation for larger projects automatic.

I learned of https://github.com/pawamoy/mkdocstrings recently. Is it similar ?

A user of mkdocstrings wrote a Crystal handler for their own use-case. They asked on the Gitter channel if we could allow to load external handlers, so they don't have to fork the project and install the fork, but rather just install their lightweight package containing just the handler.

First, note the method prefix_allowed_tokens_fn. This method applies a Pydantic model to constrain/guide how the LLM generates tokens. Next, see how that constrain can be applied to txtai's LLM pipeline.

utype is a concise alternative of pydantic with simplified parameters and usages, supporting both sync/async functions and generators parsing, and capable of using native logic operators to define logical types like AND/OR/NOT, also provides custom type parsing by register mechanism that supports libraries like pydantic, attrs and dataclasses

I like some of the changes from v1 to v2. But then you have something like this [0] removed from the library without proper documentation or replacement, resulting in ugly workarounds in the link that wont' work properly.

[0]: https://github.com/pydantic/pydantic/discussions/6337

The easiest implementation I found was to use a PyDantic class for my target schema — and use that as a parameter for the method call to “ChatCompletion.create()”. Here’s a fragment of the GinAI Python classes used.

Also, FastStream uses Pydantic to parse input JSON-encoded data into Python objects, making it easy to work with structured data in your applications, so you can serialize your input messages just using type annotations.

Pydantic Validation: Leverage Pydantic's validation capabilities to serialize and validate incoming messages

Not sure if it is exactly related, but there is an open issue on Github for that exact message.

Well the performance increase is so huge because pydantic1 is really really slow. And for using rust, I'd have expected more tbh…

I've been benchmarking pydantic v2 against typedload (which I write) and despite the rust, it still manages to be slower than pure python in some benchmarks.

The ones on the website are still about comparing to v1 because v2 was not out yet at the time of the last release.

pydantic's author will refuse to benchmark any library that is faster (https://github.com/pydantic/pydantic/pull/3264 https://github.com/pydantic/pydantic/pull/1525 https://github.com/pydantic/pydantic/pull/1810) and keep boasting about amazing performances.

On pypy, v2 beta was really really really slow.