more type hints for io/formats/format.py #27512
Conversation
…e for in ("None")
…ent (expression has type "Union[bool, str]", variable has type "None")
… + ("List[str]" and "int")
…ent (expression has type "Optional[int]", variable has type "int")
…ent (expression has type "Optional[int]", variable has type "int")
… // ("None" and "int")
…ent (expression has type "Optional[int]", variable has type "None")
…fixed_width" has incompatible type "Union[str, int]"; expected "Optional[int]"
…patible type "Union[bool, List[str]]"; expected "Sized"
…int]" for "Union[Any, List[str]]"; expected type "int"
…ent (expression has type "List[str]", variable has type "str")
…ent (expression has type "List[Any]", variable has type "Tuple[Any, ...]")
… + ("None" and "int")
…fixed_width" has incompatible type "Union[str, int]"; expected "Optional[int]"
…ment (expression has type "Type[Timedelta64Formatter]", variable has type "Type[Datetime64Formatter]")
…ment (expression has type "Union[str, Callable[..., Any], EngFormatter, Type[str], None]", variable has type "Union[Callable[..., Any], partial[Any], None]")
…ment (expression has type "None", variable has type "partial[str]")
…mpatible with supertype "TableFormatter"
…= ("int" and "None")
simonjayhawkins
added
Output-Formatting
|
Thanks! Haven't looked in detail yet but FYI I think we'll want to get #27424 in as a precursor. There's some failures showing up in there in the io.formats file so may uncover some others here as well since that's a central module getting touched Haven't checked failures in that PR in detail yet so if you have time / interest in taking a look would be appreciated |
| @@ -1207,7 +1262,7 @@ def formatter(value): | |||
|
|
|||
| return formatter | |||
|
|
|||
| def get_result_as_array(self): | |||
| def get_result_as_array(self) -> Union[ndarray, List[str]]: | |||
There was a problem hiding this comment.
this is a Union? why just ndarray?
There was a problem hiding this comment.
it depends on self.fixed_width L1303
_trim_zeros_complex L1688 returns List[str]
_trim_zeros_float L1706 returns List[str]
otherwise L1309 returns ndarray
There was a problem hiding this comment.
grr really? this just sounds wrong, can you cahnge those 2 to return an ndarray(object) if needed; these should all have a consistent signature.
There was a problem hiding this comment.
I really don't think we should be refactoring at the same time as adding type hints.
it's ok when someone is adding new code or modifying existing code.
but to benefit from the network effect of typing, i want to get this merged and get on with typing call sites and typing the imported functions.
refactoring can be done in parallel.
There was a problem hiding this comment.
ok sure. its nice that the typing surfaces things like this. pls make an issue (or PR even better!)
| def _format_datetime64(x, tz=None, nat_rep="NaT"): | ||
| def _format_datetime64( | ||
| x: Union[NaTType, Timestamp], | ||
| tz: Optional[Union[tzfile, tzutc]] = None, |
There was a problem hiding this comment.
these are not just strings?
| def __init__(self, values, nat_rep="NaT", box=False, **kwargs): | ||
| def __init__( | ||
| self, | ||
| values: Union[ndarray, TimedeltaIndex], |
There was a problem hiding this comment.
so due the fact that _format_native_types in pandas/core/indexes/timedeltas.py uses Timedelta64Formatter directly, this should account for the TimedeltaIndex.
| formatter = self.formatter or _get_format_timedelta64( | ||
| self.values, nat_rep=self.nat_rep, box=self.box | ||
| ) | ||
| fmt_values = np.array([formatter(x) for x in self.values]) | ||
| return fmt_values | ||
|
|
||
|
|
||
| def _get_format_timedelta64(values, nat_rep="NaT", box=False): | ||
| def _get_format_timedelta64( | ||
| values: Union[ndarray, TimedeltaIndex, TimedeltaArray], |
There was a problem hiding this comment.
similiarly, _get_format_timedelta64 is imported and used by core\arrays\timedeltas.py and core\indexes\timedeltas.py
|
thanks @simonjayhawkins if you can address the questions (in this PR) in a followup would be great |
no problem. i'm hoping some issues will resolve themselves as more type hints are added to imported functions and the default return type of so i'm tending towards favoring more liberal use However, I think we may start to encounter more and more issues with the untyped _lib.pyx imports. mypy has good understanding of isinstance, but not of the have spent some time looking at stubbing and overloading, no joy yet. so we may need to add casts after every |
another pass to follow-on from #27418