INTERNAL ERROR

[leandropls]

I've just got one of these:

INTERNAL ERROR -- please report a bug at https://github.com/python/mypy/issues version: 0.510-dev-88fe21d5e570160e790384d62c75ee1234e6a597
Traceback (most recent call last):
  File "/Users/leandro/env/acg-3.6.1/bin/mypy", line 6, in <module>
    main(__file__)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/main.py", line 46, in main
    res = type_check_only(sources, bin_dir, options)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/main.py", line 93, in type_check_only
    options=options)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/build.py", line 188, in build
    graph = dispatch(sources, manager)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/build.py", line 1570, in dispatch
    process_graph(graph, manager)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/build.py", line 1813, in process_graph
    process_stale_scc(graph, scc, manager)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/build.py", line 1912, in process_stale_scc
    graph[id].type_check_first_pass()
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/build.py", line 1485, in type_check_first_pass
    self.type_checker.check_first_pass()
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 177, in check_first_pass
    self.accept(d)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 262, in accept
    stmt.accept(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/nodes.py", line 748, in accept
    return visitor.visit_class_def(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 1077, in visit_class_def
    self.accept(defn.defs)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 262, in accept
    stmt.accept(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/nodes.py", line 809, in accept
    return visitor.visit_block(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 1170, in visit_block
    self.accept(s)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 262, in accept
    stmt.accept(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/nodes.py", line 630, in accept
    return visitor.visit_decorator(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 2207, in visit_decorator
    e.func.accept(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/nodes.py", line 564, in accept
    return visitor.visit_func_def(self)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 510, in visit_func_def
    self.check_method_override(defn)
  File "/Users/leandro/env/acg-3.6.1/lib/python3.6/site-packages/mypy/checker.py", line 942, in check_method_override
    for base in defn.info.mro[1:]:
TypeError: 'NoneType' object is not subscriptable

[refi64]

...example code? It's kind of hard to debug a problem otherwise...

[ilevkivskyi]

Could you please tell us what are the flags you are using? In particular, do you --incremental flag?

[leandropls]

@kirbyfan64,

from typing import NamedTuple, TypeVar, Type

T = TypeVar('T')


class MyClass(NamedTuple):
    first: float
    second: float

    @classmethod
    def fromnumber(cls: Type[T], value: float) -> T:
        return cls(value / 2, value / 2)

@ilevkivskyi, no special flags needed.

[refi64]

This doesn't crash under master:

ryan@DevPC-LX ~/stuff/mypy master $ mypy nt.py
nt.py:12: error: Too many arguments for "object"
ryan@DevPC-LX ~/stuff/mypy master $ 

[leandropls]

Hasn't this been settled here?
python/typing#352

[leandropls]

Oh, sorry, I was responding to a comment that doesn't really exist anymore.

[gvanrossum]

If the crash is fixed in master, close the issue please.

[leandropls]

Is this really supposed to be an error? Even if it doesn't crash. It seems fr the issue above, that the code should be valid.

[ilevkivskyi]

@leandropls

There is a proposal (proto-PEP "Data classes") to add a support for such use cases (both runtime and static). NamedTuple should be kept rather minimalistic.