// the find
samuelcolvin/dirty-equals
Doing dirty (but extremely useful) things with equals.
dirty-equals lets you write test assertions by overloading __eq__ on special matcher objects, so you can compare an entire API response dict in one assert statement instead of unpacking individual fields. It's from Samuel Colvin (pydantic author), which is a reasonable credibility signal. Best fit for Python developers writing integration or API tests where you care about shape and approximate values, not exact bytes.
The dict-matching pattern is genuinely useful — asserting a whole response in one block reads far better than five separate field checks. IsNow with a delta tolerance solves a real and annoying problem with timestamp assertions. Boolean composition with | and & is a nice touch that avoids nested any_of helper functions. The library is narrow and focused; it doesn't try to replace pytest or hypothesis, just fills the gap between 'exact equality' and 'write your own helper'.
Under 1000 stars after what appears to be years of existence suggests it never found wide adoption outside pydantic ecosystem users. The __eq__ abuse is clever but can confuse readers who don't know the library — `assert value == IsPositiveInt` is surprising enough to stop someone cold during a code review. No async-specific matchers for things like coroutine results or asyncio-native patterns. At ~50 forks and a small contributor base, you're largely betting on Colvin's continued interest to keep it maintained.