My favorite Python "crime" is that a class that defines __rrshift__, instantiated and used as a right-hand-side, lets you have a pipe operator, regardless of the left-hand-side (as long as it doesn't define __rshift__).
It's reasonably type-safe, and there's no need to "close" your chain - every outputted value as you write the chain can have a primitive type.
It shines in notebooks and live coding, where you might want to type stream-of-thought in the same order of operations that you want to take place. Need to log where something might be going wrong? Tee it like you're on a command line!
Idiomatic? Absolutely not. Something to push to production? Not unless you like being stabbed with pitchforks. Actually useful for prototyping? 1000%.
Having spent a lot of time lurking on the frustratingly-slow-moving bikeshedding thread for the Javascript pipe operator [0], there's a great irony that a lot of people want a pipe operator because they don't want to deal with function composition in any way, other than just applying a series of operations to their data!
I think there's a big gap pedagogically here. Once a person understands functional programming, these kinds of composition shorthands make for very straightforward and intuitive code.
But, if you're just understanding basic Haskell/Clojure syntax, or stuck in the rabbit hole of "monad is a monoid" style introductions, a beginner could easily start to think: "This functional stuff is really making me need to think in reverse, to need to know my full pipeline before I type a single character, and even perhaps to need to write Lisp-like (g (f x)) style constructs that are quite the opposite of the way my data is flowing."
I'm quite partial to tutorials like Railway Oriented Programming [1] which start from a practical imperative problem, embrace the idea that data and code should feel like they flow in the same direction, and gradually guide the reader to understanding the power of the functional tools they can bring to bear!
If anything, I hope this hack sparks good conversations :)
Sadly many things define the __or__ operator, including dicts and sets which are common to find in pipelines. (https://peps.python.org/pep-0584/ was a happy day for everyone but me!)
In practice, rshift gives a lot more flexibility! And you’d rarely chain after a numeric value.
It's a sign of the design quality of a programming language when 2 arbitrary features A and B of that language can be combined and the combination will not explode in your face.
In python and C++ (and plenty of other languages) you constantly have the risk that 2 features don't combine. Both python and C++ are full of examples where you will learn the hard way: "ah yes, this doesn't work." Or "wow, this is really unexpected".
Well, there is also a question of attitude. Most of the Python programmers don't overload << or >> even though they technically can, while in C++ that's literally the way the standard library does I/O ― and I suspect it leaves an impression on people studying it as one of their first languages that no, it's fine to overload operators however quirkily you want. Overload "custom_string * 1251" to mean "convert string from Windows-1251 to UTF-8"? Sure, why not.
Yes. iostreams overriding << and >> was pretty much universally seen as a bad idea and they eventually abandoned it in C++20/23 with std::format and std::print.
It's usually a good idea for operators to have a specific meaning and not randomly change that meaning depending on the context. If you want to add new operators with new meanings, that's fine. Haskell does that. The downside is people find it really tempting and you end up with a gazillian difficult-to-google custom operators that you have to learn.
Personally, I have never liked the PEP 634 pattern matching. I write a lot of code in Python. 99% of the time when I could use pattern matching, I am going to use simple if statements or dictionaries. Most of the time, they are more straightforward and easier to read, especially for developers who are more familiar with traditional control flow.
Dictionaries with a limited key and value type definition are fine, but dictionaries as a blind storage type are a recipe for crashing in prod with type errors or key errors. Structural pattern matching exists to support type safety.
I'll argue that code is in fact not easy to read if reading it doesn't tell you what type an item is and what a given line of code using it even does at runtime.
What’s the problem using it as a switch statement if you care about typographic issues? I do this so I’d like to know if I missed something and this is a bad practice.
Could someone explain just what's so bad about this?
My best guess is that it adds complexity and makes code harder to read in a goto-style way where you can't reason locally about local things, but it feels like the author has a much more negative view ("crimes", "god no", "dark beating heart", the elmo gif).
Maybe I have too much of a "strongly typed language" view here, but I understood the utility of isinstance() as verifying that an object is, well, an instance of that class - so that subsequent code can safely interact with that object, call class-specific methods, rely on class-specific invariants, etc.
This also makes life directly easier for me as a programmer, because I know in what code files I have to look to understand the behavior of that object.
Even linters use it to that purpose, e.g. resolving call sites by looking at the last isinstance() statement to determine the type.
__subclasshook__ puts this at risk by letting a class lie about its instances.
A linter would pass this code without warnings, because it assumes that the if block is only entered if x is in fact an instance of Everything and therefore has the foo() method.
But what really happens is that the block is entered for any kind of object, and objects that don't happen to have a foo() method will throw an exception.
You _can_ write pathological code like the Everything example, but I can see this feature being helpful if used responsibly.
It essentially allows the user to check if a class implements an interface, without explicitly inheriting ABC or Protocol. It’s up to the user to ensure the body of the case doesn’t reference any methods or attributes not guaranteed by the subclass hook, but that’s not necessarily bad, just less safe.
> It essentially allows the user to check if a class implements an interface, without explicitly inheriting ABC or Protocol.
Protocols don't need to be explicit superclasses for compile time checks, or for runtime checks if they opt-in with @runtime_checkable, but Protocols are also much newer than __subclass_hook__.
I don't think so. I think the other code should just stop using isinstance checks and switch to some custom function. I personally think isinstance checks benefit from having its behavior simpler and less dynamic.
> check if a class implements an interface, without explicitly inheriting ABC or Protocol
This really doesn't sound like a feature that belongs in the language. Go do something custom if you really want it.
But the moment you use a third party library, you cannot use it “responsibly” because that library, too, might use it “responsibly”, and then, you can easily get spooky interaction at a distance, with bugs that are hard or even impossible to fix.
I took the memes as largely for comedic effect, only?
I do think there is a ton of indirection going on in the code that I would not immediately think to look for. As the post stated, could be a good reason for this in some things. But it would be the opposite of aiming for boring code, at that point.
TL;DR having a class that determines if some other class is a subclass of itself based off of arbitrary logic and then using that arbitrary logic to categorize other people's arbitrary classes at runtime is sociopathic.
Some of these examples are similar in effect to what you might do in other languages, where you define an 'interface' and then you check to see if this class follows that interface. For example, you could define an interface DistancePoint which has the fields x and y and a distance() method, and then say "If this object implements this interface, then go ahead and do X".
Other examples, though, are more along the lines of if you implemented an interface but instead of the interface constraints being 'this class has this method' the interface constraints are 'today is Tuesday'. That's an asinine concept, which is what makes this crimes and also hilarious.
You better not find out about Protocols in Python then. The behavior you describe is exactly how duck typing / "structural subtyping" works. Your class will be an instance of Iterable if you implement the right methods having never known the Iterable class exists.
I don't find using __subclasshook__ to implement structural subtyping that you can't express with Protocols/ABCs alone to be that much of a crime. You can do evil with it but I can perform evil with any language feature.
> You better not find out about Protocols in Python then. The behavior you describe is exactly how duck typing / "structural subtyping" works. Your class will be an instance of Iterable if you implement the right methods having never known the Iterable class exists.
Conforming to an interface is a widely accepted concept across many popular languages. __subclasshook__ magic is not. So there is a big difference in violating the principle of least surprise.
That said, I'd be curious to hear a legitimate example of using it to implement "structural subtyping that you can't express with Protocols/ABCs alone".
> That said, I'd be curious to hear a legitimate example of using it to implement "structural subtyping that you can't express with Protocols/ABCs alone".
ABCs with __subclasshook__ have been available since Python 2.6, providing a mechanism to inplement runtime-testable structural subtyping. Protocols and @runtime_checkable, which provide typechecking-time structural subtyping (Protocols) that can also be available at runtime (with @runtime_checkable) were added in Python 3.8, roughly 11 years later.
There may not be much reason to use __subclasshook__ in new code, but there's a pretty good reason it exists.
The real crime is the design of Python's pattern matching in the first place:
match status:
case 404:
return "Not found"
not_found = 404
match status:
case not_found:
return "Not found"
Everywhere else in the language, you can give a constant a name without changing the code's behaviour. But in this case, the two snippets are very different: the first checks for equality (`status == 404`) and the second performs an assignment (`not_found = status`).
At least functional languages tend to have block scope, so the latter snippet introduces a new variable that shadows `not_found` instead of mutating it.
No, at least in Erlang a variable is assigned once, you can then match against that variable as it can't be reassigned:
NotFound = 404,
case Status of
NotFound -> "Not Found";
_ -> "Other Status"
end.
That snippet will return "Other Status" for Status = 400. The Python equivalent of that snippet is a SyntaxError as the first case is a catch all and the rest is unreachable.
Destructuring yes but you can still argue it's poorly designed. Particularly unintuitive because matching on a nested name e.g. module.CONSTANT works for matching and doesn't destructure. It's just the use of an unnested name which does destructuring.
What Python needs is what Elixir has. A "pin" operator that forces the variable to be used as its value for matching, rather than destructuring.
Destructuring is a feature. Making it easy to confuse value capture and value reference was an error. Add single-namespace locals and you have a calamity.
It wouldn't be a problem is Python had block level variable scope. Having that destructuring be limited to the 'case' would be fine, but obliterating the variable outside of the match is very unexpected.
The very first example there shows a match/case block where almost every single case just runs "pass" and yet every single one has a side effect. It's very difficult to read at first, difficult to understand if you're new to the syntax, and is built entirely around side effects. This might be one of the worst PEPs I've ever seen just based on that example alone.
Fun fact: you can do the same thing with the current match/case, except that you have to put your logic in the body of the case so that it's obvious what's happening.
let num = Some(4);
match num {
Some(x) if x % 2 == 0 => println!("The number {x} is even"),
Some(x) => println!("The number {x} is odd"),
None => (),
}
Which x? There are two in your code, one for each time you introduce a pattern Some(x) and each x has scope which of course ends when that pattern is done with
Notice that the Python doesn't work this way, we didn't make a new variable but instead changed the existing one.
Also, the intent in the Python was a constant, in Rust we'd give this constant an uppercase name by convention, but regardless it's a constant and so of course matching against a constant does what you expect, it can't re-bind a constant, 404 is a constant and so is `const NOT_FOUND: u16 = 404;`
> Which x? There are two in your code, one for each time you introduce a pattern Some(x) and each x has scope which of course ends when that pattern is done with
if each x's scope ends at the end of each case doesn't that mean there's only one x?
> we didn't make a new variable but instead changed the existing one.
so because python doesn't have scopes except for function scopes it shouldn't ever have any new features that intersect with scope?
Being in line with the bad original design decision is another bad design decision, python developers should have a courage to admit these instances to benefit from better decisions in new peps. They didn't do it with pattern matching and now the language has another inferior implementation of a feature that, if implemented correctly, should have had clear block scopes, defined as expressions (as opposed to statements), and disallowed type-diverging branches. Java has designed it right, by the way, despite having a differently behaving switch statement in the language already.
> Being in line with the bad original design decision is another bad design decision
I disagree. Consistently going with the "bad" choice (in this case, leaking the variable to the outer scope) is better inconsistently swinging between 2 ways of doing things. Least astonishment!
"We've made this mistake before so for consistency we need to repeat it" is such a bad idea. Ideally you want a way to go back and fix things you got wrong, but, even if you can't do that (which is itself a defect and you should figure out how you can improve) you should improve as you move forward.
C++ has struggled with this, so that paper authors sometimes plead with the committee not to make their proposal needlessly worse in the name of "consistency" with existing bad features. This famously failed for std::span, which thus managed to be not only a real world footgun in a language which already has plenty of footguns but also a PR footgun - because for "consistency" the committee removed the safety from the safety feature and I believe in C++ 26 they will repair this so it's just pointless rather than actively worse...
> "We've made this mistake before so for consistency we need to repeat it" is such a bad idea.
no it's not. you literally have no other choice. you cannot add scopes after the fact - it is a breaking change. you cannot have some parts of the language have scopes and others not - it's so much worse for the user. think about human languages and which ones get quoted as being the hardest to learn (those with many exceptions). you're just wrong.
Or stated another way, adding block scoping necessitates Python 4. You could be cheeky and make it work in a backwards compatible way
by adding new opt-in syntax but it wouldn't fix match nor
be terribly ergonomic having
to declare it all the time.
Problem is, we already have a syntax for empty lists [], empty tuples (), and {} is taken for an empty dict. So having a syntax for an empty set actually makes sense to me
Python needed a breaking change for Unicode and a breaking change for exceptions and took it ages ago for a better future today - and it's still remembered as a huge PITA by everyone. I think you'll find everyone in the Python community disagreeing with you about a not-backwards-compatible Python 4.
Every minor release of Python is a breaking change. They deprecate stuff all the time, and remember the stdlib and the wider ecosystem has to also move in concert so the breaking changes cascade.
By major version I meant minor version, 3.13 -> 3.14 is a minor version in Python, but a major source of breaking changes, that is what I meant. There will be no Python 4
If Python actually incremented the major version every time they broke backwards compatibility, we'd be on something like Python 36 by now.
Almost every version they break existing code. This is why it's common for apps written in Python to depend on specific Python versions instead of just "anything above 3.x".
Are you suggesting to bump to Python 4 in order to be able to write `{}` instead of `set()` (or `{/}`) and simultaneously break all existing code using `{}` for dicts?
Breaking {} to be an empty set would be a HUGE breaking change, a _lot_ of code is already written where it is expected to be an empty dict. I don't think anyone in the Python committee would agree with breaking that
Making sense, and being good, is not necessary the same.
Yes, having a solution for this makes sense, but the proposed solutions are just not good. Sometimes one has to admit that not everything can be solved gracefully and just stop, hunting the whale.
Yes but they are not equivalent. dict and list are factories; {} and [] are reified when the code is touched and then never reinitialised again. This catches out beginners and LLMs alike:
They are equivalent. In function signatures (what your article is talking about), using dict() instead of {} will have the same effect. The only difference is that {} is a literal of an empty dict, and dict is a name bound to the builtin dict class. So you can reassign dict, but not {}, and if you use dict() instead of {}, then you have a name lookup before a call, so {} is a little more efficient.
I wrote the link and yes it does. Module evaluations reify {}, [], etc. once. That is why people keep making subtle bugs when they do `def foo(a=[]):` unaware that this will in fact not give you a brand new list on every function call.
Factory functions like list/tuple/set are function calls and are executed and avoid this problem. Hence why professional python devs default to `None` and check for that and _then_ initialise the list internally in the function body.
Adding {/} as empty set is great, sure; but that again is just another reified instance and the opposite of set() the function.
For reasons I don't think I understand, using the functions is "discouraged" because "someone might muck with how those functions work" and the python world, in it's perfect wisdom responded "Oh of course" instead of "That's so damn stupid, don't do that because it would be surprising to people who expect built in functions to do built in logic"
Next step is to have the subclass check pack all the code up, send it to ChatGPT, ask it if it thinks subjectively that class A should be a subclass of class B, and then run sentiment analysis on the resulting text to make the determination.
I don't think this is necessarily a "crime", I could see this being useful.
An example that uses a similar approach would be if you had a metaclass that overrode __instancecheck__ to return true if a string matched a regular expression. Then you could create (dynamically defined) classes that used that metaclass to use in match statements to march a string against multiple regexes. Unfortunately, I can't think of a good way to extract any capture groups.
ok, took me a bit to wrap my head around whats going on here, but actually this is neat, and probably useful.
Ive been hobbying in python for a decade plus at this point and it still ties my brain in knots sometimes with how it works at runtime. I do enjoy working with it though.
E: just to clarify if you're going to use this don't bury it deep in a multi file code structure, explicitly have it in the same file as you use it, otherwise people will get confused.
I've never understood why Python's pattern-matching isn't more general.
First, "case foo.bar" is a value match, but "case foo" is a name capture. Python could have defined "case .foo" to mean "look up foo as a variable the normal way" with zero ambiguity, but chose not to.
Second, there's no need to special-case some builtin types as matching whole values. You can write "case float(m): print(m)" and print the float that matched, but you can't write "case MyObject(obj): print(obj)" and print your object. Python could allow "..." or "None" or something in __match_args__ to mean "the whole object", but didn't.
After doing Erlang and Scala pattern matching, the whole Python implementation just feels really ugly and gross. They should have cribbed a lot more of how Scala does it.
> While potentially useful, it introduces strange-looking new syntax without making the pattern syntax any more expressive. Indeed, named constants can be made to work with the existing rules by converting them to Enum types, or enclosing them in their own namespace (considered by the authors to be one honking great idea)[...]
If needed, the leading-dot rule (or a similar variant) could be added back later with no backward-compatibility issues.
second: you can use case MyObject() as obj: print(obj)
I don't think I've written a match-case yet. Aside from not having a lot of use cases for it personally, I find that it's very strange-feeling syntax. It tries too hard to look right, with the consequence that it's sometimes quite hard to reason about.
> > While potentially useful, it introduces strange-looking new syntax without making the pattern syntax any more expressive. Indeed, named constants can be made to work with the existing rules by converting them to Enum types, or enclosing them in their own namespace (considered by the authors to be one honking great idea)[...]
Yeah, and I don't buy that for a microsecond.
A leading dot is not "strange" syntax: it mirrors relative imports. There's no workaround because it lets you use variables the same way you use them in any other part of the language. Having to distort your program by adding namespaces that exist only to work around an artificial pattern matching limitation is a bug, not a feature.
Also, it takes a lot of chutzpah for this PEP author to call a leading dot strange when his match/case introduces something that looks lexically like constructor invocation but is anything but.
The "as" thing works with primitive too, so why do we need int(m)? Either get rid of the syntax or make it general. Don't hard-code support for half a dozen stdlib types for some reason and make it impossible for user code to do the equivalent.
The Python pattern matching API is full of most stdlib antipatterns:
* It's irregular: matching prohibits things that the shape of the feature would suggest are possible because the PEP authors couldn't personally see a specific use case for those things. (What's the deal with prohibiting multiple _ but allowing as many __ as you want?)
* It privileges stdlib, as I mentioned above. Language features should not grant the standard library powers it doesn't extend to user code.
* The syntax feels bolted on. I get trying to reduce parser complexity and tool breakage by making pattern matching look like object construction, but it isn't, and the false cognate thing confuses every single person who tries to read a Python program. They could have used := or some other new syntax, but didn't, probably because of the need to build "consensus"
* The whole damn thing should have been an expression, like the if/then/else ternary, not a statement useless outside many lexical contexts in which one might want to make a decision. Why is it a statement? Probably because the PEP author didn't _personally_ have a need to pattern match in expression context.
And look: you can justify any of these technical decisions. You can a way to justify anything you might want to do. The end result, however, is a language facility that feels more cumbersome than it should and is applicable to fewer places than one might think.
> If needed, the leading-dot rule (or a similar variant) could be added back later with no backward-compatibility issues.
So what, after another decade of debate, consensus, and compromise, we'll end up with a .-prefix-rule but one that works only if the character after the dot is a lowercase letter that isn't a vowel.
PEP: "We decided not to do this because inspection of real-life potential use cases showed that in vast majority of cases destructuring is related to an if condition. Also many of those are grouped in a series of exclusive choices."
I find this philosophical stance off-putting. It's a good thing when users find ways to use your tools in ways you didn't imagine.
PEP: In most other languages pattern matching is represented by an expression, not statement. But making it an expression would be inconsistent with other syntactic choices in Python. All decision making logic is expressed almost exclusively in statements, so we decided to not deviate from this.
We've had conditional expressions for a long time.
After starting my new job and coming back to Python after many years I was happy to see that they had added `match` to the language. Then I was immediately disappointed as soon as I started using it as I ran into its weird limitations and quirks.
Why did they design it so poorly? The language would be better off without it in its current hamstrung form, as it only adds to the already complex syntax of the language.
> PEP: In most other languages pattern matching is represented by an expression, not statement. But making it an expression would be inconsistent with other syntactic choices in Python. All decision making logic is expressed almost exclusively in statements, so we decided to not deviate from this.
> We've had conditional expressions for a long time.
Also, maybe most other languages represent it as an expression because it's the sane thing to do? Python doing its own thing here isn't the win they think it is.
The Python core team has kind of run the language off the rails post 3.7 or 3.8 or so. There's been so much crap bolted on to the language for dubious reasons, and often times it comes with whole new sets of weird problems without really making life easier (async was a quintessential example of this in my mind). There's a lot of design choices core to the language itself that make it a poor choice for many tasks, but that never stops anyone from doing it anyways and bolting on lots of chincy "features" along the way.
> An irrefutable case block is a match-all case block. A match statement may have at most one irrefutable case block, and it must be last.
There is no reason to have this restriction except that some people as a matter of opinion think unreachable code is bad taste and the language grammar should make bad taste impossible to express. It's often useful to introduce such things as a temporary state during editing. For example,
def foo(x):
match x:
case _:
log.warning("XXX disabled for debugging")
return PLACEHOLDER
case int():
return bar()
case str():
return qux()
case _:
return "default"
Why should my temporary match-all be a SyntaxError???? Maybe it's a bug. Maybe my tools should warn me about it. But the language itself shouldn't enforce restrictions rooted in good taste instead of technical necessity.
I can, however, write this:
def foo(x):
match x:
case _ if True:
log.warning("XXX disabled for debugging")
return PLACEHOLDER
case int():
return bar()
case str():
return qux()
case _:
return "default"
Adding a dummy guard is a ridiculous workaround for a problem that shouldn't exist in the first place.
While the article is very entertaining, I'm not a fan of the pattern matching in Python.
I wish for some linter rule that can forbid the usage of pattern matching.
Should be easily doable with a semgrep rule, e.g.:
~> cat semgrep.yaml
rules:
- id: no-pattern-matching
pattern: |
match ...:
message: |
I'm not a fan of the pattern matching in Python
severity: ERROR
languages:
- python
...
~> cat test.py
#!/usr/bin/env python3
foo = 1
match foo:
case 1:
print("one")
...
~> semgrep --config semgrep.yaml test.py
no-pattern-matching
I'm not a fan of the pattern matching in Python
4┆ match foo:
5┆ case 1:
6┆ print("one")
You need to make that exclude match = ... since match can also be a variable name. This is because people used to write code like match = re.search(...)
If you're experienced enough with Python to say "I want to eliminate pattern matching from my codebase" you can surely construct that as a pre-commit check, no?
I've been doing Python full time for about 4 years now.
I always found that when pattern matching is used it always ends up increasing the cognitive load significantly. It shines for simple cases, and for precise object pattern sniping. But anything that acts on nested objects or more than 3 members with multiple case statements is pre-programmed for failure. What makes it worse, is that there seems to be a trend where other Python adepts will default to pattern matching over everything else, like a too good of a hammer.
My favorite Python "crime" is that a class that defines __rrshift__, instantiated and used as a right-hand-side, lets you have a pipe operator, regardless of the left-hand-side (as long as it doesn't define __rshift__).
It's reasonably type-safe, and there's no need to "close" your chain - every outputted value as you write the chain can have a primitive type.
It shines in notebooks and live coding, where you might want to type stream-of-thought in the same order of operations that you want to take place. Need to log where something might be going wrong? Tee it like you're on a command line!Idiomatic? Absolutely not. Something to push to production? Not unless you like being stabbed with pitchforks. Actually useful for prototyping? 1000%.
I spy a functional programmer lurking in this abuse of Python ;)
Looks a lot like function composition with the arguments flipped, which in Haskell is `>>>`. Neat!
But since you’re writing imperative code and binding the result to a variable, you could also compare to `>>=`.
(https://downloads.haskell.org/~ghc/7.6.2/docs/html/libraries...)
Having spent a lot of time lurking on the frustratingly-slow-moving bikeshedding thread for the Javascript pipe operator [0], there's a great irony that a lot of people want a pipe operator because they don't want to deal with function composition in any way, other than just applying a series of operations to their data!
I think there's a big gap pedagogically here. Once a person understands functional programming, these kinds of composition shorthands make for very straightforward and intuitive code.
But, if you're just understanding basic Haskell/Clojure syntax, or stuck in the rabbit hole of "monad is a monoid" style introductions, a beginner could easily start to think: "This functional stuff is really making me need to think in reverse, to need to know my full pipeline before I type a single character, and even perhaps to need to write Lisp-like (g (f x)) style constructs that are quite the opposite of the way my data is flowing."
I'm quite partial to tutorials like Railway Oriented Programming [1] which start from a practical imperative problem, embrace the idea that data and code should feel like they flow in the same direction, and gradually guide the reader to understanding the power of the functional tools they can bring to bear!
If anything, I hope this hack sparks good conversations :)
[0] https://github.com/tc39/proposal-pipeline-operator/issues/91 - 6 years and 793 comments!
[1] https://fsharpforfunandprofit.com/rop/
I suppose you could use the same trick with __ror__ and | (as long as the left-hand side doesn’t define __or__).
Sadly many things define the __or__ operator, including dicts and sets which are common to find in pipelines. (https://peps.python.org/pep-0584/ was a happy day for everyone but me!)
In practice, rshift gives a lot more flexibility! And you’d rarely chain after a numeric value.
It’s still useful in related situations. The following crime often finds its way into my $PYTHONSTARTUP:
and here the low precedence of | is useful.Or overloading division like https://scapy.net/
Oh god, it's c++ all over again!
Is that supposed to be a bad thing?
IMNSHO: Yes.
It's a sign of the design quality of a programming language when 2 arbitrary features A and B of that language can be combined and the combination will not explode in your face. In python and C++ (and plenty of other languages) you constantly have the risk that 2 features don't combine. Both python and C++ are full of examples where you will learn the hard way: "ah yes, this doesn't work." Or "wow, this is really unexpected".
Well, there is also a question of attitude. Most of the Python programmers don't overload << or >> even though they technically can, while in C++ that's literally the way the standard library does I/O ― and I suspect it leaves an impression on people studying it as one of their first languages that no, it's fine to overload operators however quirkily you want. Overload "custom_string * 1251" to mean "convert string from Windows-1251 to UTF-8"? Sure, why not.
I've seen >> being overloaded in several libraries/frameworks. From the top of my head:
Yes. iostreams overriding << and >> was pretty much universally seen as a bad idea and they eventually abandoned it in C++20/23 with std::format and std::print.
It's usually a good idea for operators to have a specific meaning and not randomly change that meaning depending on the context. If you want to add new operators with new meanings, that's fine. Haskell does that. The downside is people find it really tempting and you end up with a gazillian difficult-to-google custom operators that you have to learn.
Apache beam does it: https://beam.apache.org/documentation/sdks/python-streaming/
I wanted to wash my eyes the first time I saw it.
Personally, I have never liked the PEP 634 pattern matching. I write a lot of code in Python. 99% of the time when I could use pattern matching, I am going to use simple if statements or dictionaries. Most of the time, they are more straightforward and easier to read, especially for developers who are more familiar with traditional control flow.
Dictionaries with a limited key and value type definition are fine, but dictionaries as a blind storage type are a recipe for crashing in prod with type errors or key errors. Structural pattern matching exists to support type safety.
I'll argue that code is in fact not easy to read if reading it doesn't tell you what type an item is and what a given line of code using it even does at runtime.
You should use if statements if that's what you need. The match statement is for structural pattern matching mostly.
What’s the problem using it as a switch statement if you care about typographic issues? I do this so I’d like to know if I missed something and this is a bad practice.
Having user other languages after finishing university studies, Python does not spark joy.
Could someone explain just what's so bad about this?
My best guess is that it adds complexity and makes code harder to read in a goto-style way where you can't reason locally about local things, but it feels like the author has a much more negative view ("crimes", "god no", "dark beating heart", the elmo gif).
Maybe I have too much of a "strongly typed language" view here, but I understood the utility of isinstance() as verifying that an object is, well, an instance of that class - so that subsequent code can safely interact with that object, call class-specific methods, rely on class-specific invariants, etc.
This also makes life directly easier for me as a programmer, because I know in what code files I have to look to understand the behavior of that object.
Even linters use it to that purpose, e.g. resolving call sites by looking at the last isinstance() statement to determine the type.
__subclasshook__ puts this at risk by letting a class lie about its instances.
As an example, consider this class:
You can now write code like this: A linter would pass this code without warnings, because it assumes that the if block is only entered if x is in fact an instance of Everything and therefore has the foo() method.But what really happens is that the block is entered for any kind of object, and objects that don't happen to have a foo() method will throw an exception.
You _can_ write pathological code like the Everything example, but I can see this feature being helpful if used responsibly.
It essentially allows the user to check if a class implements an interface, without explicitly inheriting ABC or Protocol. It’s up to the user to ensure the body of the case doesn’t reference any methods or attributes not guaranteed by the subclass hook, but that’s not necessarily bad, just less safe.
All things have a place and time.
> It essentially allows the user to check if a class implements an interface, without explicitly inheriting ABC or Protocol.
Protocols don't need to be explicit superclasses for compile time checks, or for runtime checks if they opt-in with @runtime_checkable, but Protocols are also much newer than __subclass_hook__.
TIL, thanks!
(I love being wrong on HN, always learn something)
I don't think so. I think the other code should just stop using isinstance checks and switch to some custom function. I personally think isinstance checks benefit from having its behavior simpler and less dynamic.
> check if a class implements an interface, without explicitly inheriting ABC or Protocol
This really doesn't sound like a feature that belongs in the language. Go do something custom if you really want it.
But the moment you use a third party library, you cannot use it “responsibly” because that library, too, might use it “responsibly”, and then, you can easily get spooky interaction at a distance, with bugs that are hard or even impossible to fix.
A good example being stuff like isinstance(x, Iterable) and friends. Figuring out if something is iterable is a bit of a palaver otherwise.
I took the memes as largely for comedic effect, only?
I do think there is a ton of indirection going on in the code that I would not immediately think to look for. As the post stated, could be a good reason for this in some things. But it would be the opposite of aiming for boring code, at that point.
TL;DR having a class that determines if some other class is a subclass of itself based off of arbitrary logic and then using that arbitrary logic to categorize other people's arbitrary classes at runtime is sociopathic.
Some of these examples are similar in effect to what you might do in other languages, where you define an 'interface' and then you check to see if this class follows that interface. For example, you could define an interface DistancePoint which has the fields x and y and a distance() method, and then say "If this object implements this interface, then go ahead and do X".
Other examples, though, are more along the lines of if you implemented an interface but instead of the interface constraints being 'this class has this method' the interface constraints are 'today is Tuesday'. That's an asinine concept, which is what makes this crimes and also hilarious.
You better not find out about Protocols in Python then. The behavior you describe is exactly how duck typing / "structural subtyping" works. Your class will be an instance of Iterable if you implement the right methods having never known the Iterable class exists.
I don't find using __subclasshook__ to implement structural subtyping that you can't express with Protocols/ABCs alone to be that much of a crime. You can do evil with it but I can perform evil with any language feature.
> You better not find out about Protocols in Python then. The behavior you describe is exactly how duck typing / "structural subtyping" works. Your class will be an instance of Iterable if you implement the right methods having never known the Iterable class exists.
Conforming to an interface is a widely accepted concept across many popular languages. __subclasshook__ magic is not. So there is a big difference in violating the principle of least surprise.
That said, I'd be curious to hear a legitimate example of using it to implement "structural subtyping that you can't express with Protocols/ABCs alone".
> That said, I'd be curious to hear a legitimate example of using it to implement "structural subtyping that you can't express with Protocols/ABCs alone".
ABCs with __subclasshook__ have been available since Python 2.6, providing a mechanism to inplement runtime-testable structural subtyping. Protocols and @runtime_checkable, which provide typechecking-time structural subtyping (Protocols) that can also be available at runtime (with @runtime_checkable) were added in Python 3.8, roughly 11 years later.
There may not be much reason to use __subclasshook__ in new code, but there's a pretty good reason it exists.
> There may not be much reason to use __subclasshook__ in new code, but there's a pretty good reason it exists.
That's quite a different claim, and makes a lot of sense. Thanks for the history!
Side effects
The real crime is the design of Python's pattern matching in the first place:
Everywhere else in the language, you can give a constant a name without changing the code's behaviour. But in this case, the two snippets are very different: the first checks for equality (`status == 404`) and the second performs an assignment (`not_found = status`).https://x.com/brandon_rhodes/status/1360226108399099909
That's destructuring, a feature not a bug. Same as it works in any functional language - and tremendously useful once you get the hang of it
At least functional languages tend to have block scope, so the latter snippet introduces a new variable that shadows `not_found` instead of mutating it.
No, at least in Erlang a variable is assigned once, you can then match against that variable as it can't be reassigned:
That snippet will return "Other Status" for Status = 400. The Python equivalent of that snippet is a SyntaxError as the first case is a catch all and the rest is unreachable.Destructuring yes but you can still argue it's poorly designed. Particularly unintuitive because matching on a nested name e.g. module.CONSTANT works for matching and doesn't destructure. It's just the use of an unnested name which does destructuring.
What Python needs is what Elixir has. A "pin" operator that forces the variable to be used as its value for matching, rather than destructuring.
Destructuring is a feature. Making it easy to confuse value capture and value reference was an error. Add single-namespace locals and you have a calamity.
It wouldn't be a problem is Python had block level variable scope. Having that destructuring be limited to the 'case' would be fine, but obliterating the variable outside of the match is very unexpected.
And there was a much better proposal that got rejected in favor of what we got: https://peps.python.org/pep-0642/
The very first example there shows a match/case block where almost every single case just runs "pass" and yet every single one has a side effect. It's very difficult to read at first, difficult to understand if you're new to the syntax, and is built entirely around side effects. This might be one of the worst PEPs I've ever seen just based on that example alone.
Fun fact: you can do the same thing with the current match/case, except that you have to put your logic in the body of the case so that it's obvious what's happening.
because it's not a `switch` it's a `match` ie pattern matching...
Doesn't matter what it is, it shouldn't break fundamental rules of the language.
Ruby's `case`/`in` has the same problem.
> it shouldn't break fundamental rules of the language
it doesn't? you simply don't understand what a match statement is.
https://doc.rust-lang.org/book/ch19-03-pattern-syntax.html
notice that x is bound to 4.Which x? There are two in your code, one for each time you introduce a pattern Some(x) and each x has scope which of course ends when that pattern is done with
Notice that the Python doesn't work this way, we didn't make a new variable but instead changed the existing one.
Also, the intent in the Python was a constant, in Rust we'd give this constant an uppercase name by convention, but regardless it's a constant and so of course matching against a constant does what you expect, it can't re-bind a constant, 404 is a constant and so is `const NOT_FOUND: u16 = 404;`
> Which x? There are two in your code, one for each time you introduce a pattern Some(x) and each x has scope which of course ends when that pattern is done with
if each x's scope ends at the end of each case doesn't that mean there's only one x?
> we didn't make a new variable but instead changed the existing one.
so because python doesn't have scopes except for function scopes it shouldn't ever have any new features that intersect with scope?
Also this is pretty much in line for the rest of Python leaving variables around.
The match statement presented is equivalent to an assignment, you do have to know that, but then it's just regular Python.Being in line with the bad original design decision is another bad design decision, python developers should have a courage to admit these instances to benefit from better decisions in new peps. They didn't do it with pattern matching and now the language has another inferior implementation of a feature that, if implemented correctly, should have had clear block scopes, defined as expressions (as opposed to statements), and disallowed type-diverging branches. Java has designed it right, by the way, despite having a differently behaving switch statement in the language already.
> Being in line with the bad original design decision is another bad design decision
I disagree. Consistently going with the "bad" choice (in this case, leaking the variable to the outer scope) is better inconsistently swinging between 2 ways of doing things. Least astonishment!
"We've made this mistake before so for consistency we need to repeat it" is such a bad idea. Ideally you want a way to go back and fix things you got wrong, but, even if you can't do that (which is itself a defect and you should figure out how you can improve) you should improve as you move forward.
C++ has struggled with this, so that paper authors sometimes plead with the committee not to make their proposal needlessly worse in the name of "consistency" with existing bad features. This famously failed for std::span, which thus managed to be not only a real world footgun in a language which already has plenty of footguns but also a PR footgun - because for "consistency" the committee removed the safety from the safety feature and I believe in C++ 26 they will repair this so it's just pointless rather than actively worse...
> "We've made this mistake before so for consistency we need to repeat it" is such a bad idea.
no it's not. you literally have no other choice. you cannot add scopes after the fact - it is a breaking change. you cannot have some parts of the language have scopes and others not - it's so much worse for the user. think about human languages and which ones get quoted as being the hardest to learn (those with many exceptions). you're just wrong.
Or stated another way, adding block scoping necessitates Python 4. You could be cheeky and make it work in a backwards compatible way by adding new opt-in syntax but it wouldn't fix match nor be terribly ergonomic having to declare it all the time.
> you simply don't understand what a match statement is
It's "a DSL contrived to look like Python, and to be used inside of Python, but with very different semantics":
https://discuss.python.org/t/gauging-sentiment-on-pattern-ma...
you linked to a random detractors rant. i don't see what that has to do with whether a match statement binds a match?
More and more dubious things were designed in Python these days. A recent PEP purposes to use {/} as the empty set
Problem is, we already have a syntax for empty lists [], empty tuples (), and {} is taken for an empty dict. So having a syntax for an empty set actually makes sense to me
{:} should have been the empty dict, now there's no good solution
I agree that {:} would be a better empty expression for dicts, but that ship has already sailed. {/} looks like a good enough alternative
There is a way to make it work. Python has no problem breaking things across major versions.
Python needed a breaking change for Unicode and a breaking change for exceptions and took it ages ago for a better future today - and it's still remembered as a huge PITA by everyone. I think you'll find everyone in the Python community disagreeing with you about a not-backwards-compatible Python 4.
Every minor release of Python is a breaking change. They deprecate stuff all the time, and remember the stdlib and the wider ecosystem has to also move in concert so the breaking changes cascade.
By major version I meant minor version, 3.13 -> 3.14 is a minor version in Python, but a major source of breaking changes, that is what I meant. There will be no Python 4
If Python actually incremented the major version every time they broke backwards compatibility, we'd be on something like Python 36 by now.
Almost every version they break existing code. This is why it's common for apps written in Python to depend on specific Python versions instead of just "anything above 3.x".
Are you suggesting to bump to Python 4 in order to be able to write `{}` instead of `set()` (or `{/}`) and simultaneously break all existing code using `{}` for dicts?
Breaking {} to be an empty set would be a HUGE breaking change, a _lot_ of code is already written where it is expected to be an empty dict. I don't think anyone in the Python committee would agree with breaking that
Jesus can you imagine if they announced Python 4? :-D
Making sense, and being good, is not necessary the same.
Yes, having a solution for this makes sense, but the proposed solutions are just not good. Sometimes one has to admit that not everything can be solved gracefully and just stop, hunting the whale.
You can use “set()”. Introducing more weird special cases into the language is a bad direction for Python.
And you can use dict() for an empty dictionary, and list() for an empty list.
Yes but they are not equivalent. dict and list are factories; {} and [] are reified when the code is touched and then never reinitialised again. This catches out beginners and LLMs alike:
https://www.inspiredpython.com/article/watch-out-for-mutable...
That article is about how defaults for arguments are evaluated eagerly. It doesn't real have to do with dict vs {}.
However, using the literal syntax does seem to be more efficient. So that is an argument for having dedicated syntax for an empty set.
I am not replying to the article but to the poster.
I am talking about the article you linked to in your comment.
They are equivalent. In function signatures (what your article is talking about), using dict() instead of {} will have the same effect. The only difference is that {} is a literal of an empty dict, and dict is a name bound to the builtin dict class. So you can reassign dict, but not {}, and if you use dict() instead of {}, then you have a name lookup before a call, so {} is a little more efficient.
Right, but it instantiates it _once_ on module load! That is the point I am making; nothing else.
Your link doesn't support your argument.
I wrote the link and yes it does. Module evaluations reify {}, [], etc. once. That is why people keep making subtle bugs when they do `def foo(a=[]):` unaware that this will in fact not give you a brand new list on every function call.
Factory functions like list/tuple/set are function calls and are executed and avoid this problem. Hence why professional python devs default to `None` and check for that and _then_ initialise the list internally in the function body.
Adding {/} as empty set is great, sure; but that again is just another reified instance and the opposite of set() the function.
For reasons I don't think I understand, using the functions is "discouraged" because "someone might muck with how those functions work" and the python world, in it's perfect wisdom responded "Oh of course" instead of "That's so damn stupid, don't do that because it would be surprising to people who expect built in functions to do built in logic"
No no no, it's a great direction towards becoming the new Perl.
Idk that doesn’t sound so dubious to me. ∅ might be more approachable for the PHDs then set() ;)
we all love non ascii code (cough emoji variable names)
Discussed 3 years ago:
Crimes with Python's pattern matching
406 points on Aug 2, 2022. 120 comments
https://news.ycombinator.com/item?id=32314368
Barely a misdemeanor, all of the typechecks were deterministic
Next step is to have the subclass check pack all the code up, send it to ChatGPT, ask it if it thinks subjectively that class A should be a subclass of class B, and then run sentiment analysis on the resulting text to make the determination.
I don't think this is necessarily a "crime", I could see this being useful.
An example that uses a similar approach would be if you had a metaclass that overrode __instancecheck__ to return true if a string matched a regular expression. Then you could create (dynamically defined) classes that used that metaclass to use in match statements to march a string against multiple regexes. Unfortunately, I can't think of a good way to extract any capture groups.
ok, took me a bit to wrap my head around whats going on here, but actually this is neat, and probably useful.
Ive been hobbying in python for a decade plus at this point and it still ties my brain in knots sometimes with how it works at runtime. I do enjoy working with it though.
E: just to clarify if you're going to use this don't bury it deep in a multi file code structure, explicitly have it in the same file as you use it, otherwise people will get confused.
I've never understood why Python's pattern-matching isn't more general.
First, "case foo.bar" is a value match, but "case foo" is a name capture. Python could have defined "case .foo" to mean "look up foo as a variable the normal way" with zero ambiguity, but chose not to.
Second, there's no need to special-case some builtin types as matching whole values. You can write "case float(m): print(m)" and print the float that matched, but you can't write "case MyObject(obj): print(obj)" and print your object. Python could allow "..." or "None" or something in __match_args__ to mean "the whole object", but didn't.
After doing Erlang and Scala pattern matching, the whole Python implementation just feels really ugly and gross. They should have cribbed a lot more of how Scala does it.
I've given up on matching as I'm tired of running into its limitations.
That said, I don't think OP's antics are a crime. That SyntaxError though, that might be a crime.
And a class-generating callable class would get around Python caching the results of __subclasshook__.
Now I'm mad I didn't remember the word "antics". It's so much more evocative than "crimes"!
case .foo is explicitly mentioned in https://peps.python.org/pep-0622/ :
> While potentially useful, it introduces strange-looking new syntax without making the pattern syntax any more expressive. Indeed, named constants can be made to work with the existing rules by converting them to Enum types, or enclosing them in their own namespace (considered by the authors to be one honking great idea)[...] If needed, the leading-dot rule (or a similar variant) could be added back later with no backward-compatibility issues.
second: you can use case MyObject() as obj: print(obj)
I don't think I've written a match-case yet. Aside from not having a lot of use cases for it personally, I find that it's very strange-feeling syntax. It tries too hard to look right, with the consequence that it's sometimes quite hard to reason about.
I just use match case as a traditional switch statement.
> > While potentially useful, it introduces strange-looking new syntax without making the pattern syntax any more expressive. Indeed, named constants can be made to work with the existing rules by converting them to Enum types, or enclosing them in their own namespace (considered by the authors to be one honking great idea)[...]
Yeah, and I don't buy that for a microsecond.
A leading dot is not "strange" syntax: it mirrors relative imports. There's no workaround because it lets you use variables the same way you use them in any other part of the language. Having to distort your program by adding namespaces that exist only to work around an artificial pattern matching limitation is a bug, not a feature.
Also, it takes a lot of chutzpah for this PEP author to call a leading dot strange when his match/case introduces something that looks lexically like constructor invocation but is anything but.
The "as" thing works with primitive too, so why do we need int(m)? Either get rid of the syntax or make it general. Don't hard-code support for half a dozen stdlib types for some reason and make it impossible for user code to do the equivalent.
The Python pattern matching API is full of most stdlib antipatterns:
* It's irregular: matching prohibits things that the shape of the feature would suggest are possible because the PEP authors couldn't personally see a specific use case for those things. (What's the deal with prohibiting multiple _ but allowing as many __ as you want?)
* It privileges stdlib, as I mentioned above. Language features should not grant the standard library powers it doesn't extend to user code.
* The syntax feels bolted on. I get trying to reduce parser complexity and tool breakage by making pattern matching look like object construction, but it isn't, and the false cognate thing confuses every single person who tries to read a Python program. They could have used := or some other new syntax, but didn't, probably because of the need to build "consensus"
* The whole damn thing should have been an expression, like the if/then/else ternary, not a statement useless outside many lexical contexts in which one might want to make a decision. Why is it a statement? Probably because the PEP author didn't _personally_ have a need to pattern match in expression context.
And look: you can justify any of these technical decisions. You can a way to justify anything you might want to do. The end result, however, is a language facility that feels more cumbersome than it should and is applicable to fewer places than one might think.
Here's how to do it right: https://www.gnu.org/software/emacs/manual/html_node/elisp/pc...
> If needed, the leading-dot rule (or a similar variant) could be added back later with no backward-compatibility issues.
So what, after another decade of debate, consensus, and compromise, we'll end up with a .-prefix-rule but one that works only if the character after the dot is a lowercase letter that isn't a vowel.
PEP: "We decided not to do this because inspection of real-life potential use cases showed that in vast majority of cases destructuring is related to an if condition. Also many of those are grouped in a series of exclusive choices."
I find this philosophical stance off-putting. It's a good thing when users find ways to use your tools in ways you didn't imagine.
PEP: In most other languages pattern matching is represented by an expression, not statement. But making it an expression would be inconsistent with other syntactic choices in Python. All decision making logic is expressed almost exclusively in statements, so we decided to not deviate from this.
We've had conditional expressions for a long time.
Agreed.
After starting my new job and coming back to Python after many years I was happy to see that they had added `match` to the language. Then I was immediately disappointed as soon as I started using it as I ran into its weird limitations and quirks.
Why did they design it so poorly? The language would be better off without it in its current hamstrung form, as it only adds to the already complex syntax of the language.
> PEP: In most other languages pattern matching is represented by an expression, not statement. But making it an expression would be inconsistent with other syntactic choices in Python. All decision making logic is expressed almost exclusively in statements, so we decided to not deviate from this.
> We've had conditional expressions for a long time.
Also, maybe most other languages represent it as an expression because it's the sane thing to do? Python doing its own thing here isn't the win they think it is.
The Python core team has kind of run the language off the rails post 3.7 or 3.8 or so. There's been so much crap bolted on to the language for dubious reasons, and often times it comes with whole new sets of weird problems without really making life easier (async was a quintessential example of this in my mind). There's a lot of design choices core to the language itself that make it a poor choice for many tasks, but that never stops anyone from doing it anyways and bolting on lots of chincy "features" along the way.
> (What's the deal with prohibiting multiple _ but allowing as many __ as you want?)
What do you mean "prohibiting multiple _"? As in this pattern:
That works fine.> An irrefutable case block is a match-all case block. A match statement may have at most one irrefutable case block, and it must be last.
There is no reason to have this restriction except that some people as a matter of opinion think unreachable code is bad taste and the language grammar should make bad taste impossible to express. It's often useful to introduce such things as a temporary state during editing. For example,
Why should my temporary match-all be a SyntaxError???? Maybe it's a bug. Maybe my tools should warn me about it. But the language itself shouldn't enforce restrictions rooted in good taste instead of technical necessity.I can, however, write this:
Adding a dummy guard is a ridiculous workaround for a problem that shouldn't exist in the first place.I don't disagree, it should be a warning but not an error. Thanks for clarifying, your original remark was ambiguous there.
While the article is very entertaining, I'm not a fan of the pattern matching in Python. I wish for some linter rule that can forbid the usage of pattern matching.
Can you explain why? Genuinely curious as a lover of case/match. My only complaint is that it is not general enough.
Double indentation
So? Other languages with pattern match similarly have such double indentation. C-style switch with unintended cases is weird.
Should be easily doable with a semgrep rule, e.g.:
... ... (exits non-0)You need to make that exclude match = ... since match can also be a variable name. This is because people used to write code like match = re.search(...)
The existing pattern suggested above, "match ...:", will not match 'match = ...'.
Presumably the reason the parent comment suggested semgrep, not just a grep, is because they're aware that naive substring matching would be wrong.
You could use the playground to check your understanding before implying someone is an idiot.
https://semgrep.dev/playground/new
If you're experienced enough with Python to say "I want to eliminate pattern matching from my codebase" you can surely construct that as a pre-commit check, no?
I've been doing Python full time for about 4 years now. I always found that when pattern matching is used it always ends up increasing the cognitive load significantly. It shines for simple cases, and for precise object pattern sniping. But anything that acts on nested objects or more than 3 members with multiple case statements is pre-programmed for failure. What makes it worse, is that there seems to be a trend where other Python adepts will default to pattern matching over everything else, like a too good of a hammer.
[dead]