TYP: disallow comment-based annotation syntax (#29741)

Author: simonjayhawkins

ci/code_checks.sh

@@ -194,6 +194,10 @@ if [[ -z "$CHECK" || "$CHECK" == "patterns" ]]; then
     invgrep -R --include="*.py" --include="*.pyx" -E 'class.*:\n\n( )+"""' .
     RET=$(($RET + $?)) ; echo $MSG "DONE"
 
+    MSG='Check for use of comment-based annotation syntax' ; echo $MSG
+    invgrep -R --include="*.py" -P '# type: (?!ignore)' pandas
+    RET=$(($RET + $?)) ; echo $MSG "DONE"
+
     MSG='Check that no file in the repo contains trailing whitespaces' ; echo $MSG
     set -o pipefail
     if [[ "$AZURE" == "true" ]]; then

doc/source/development/contributing.rst

@@ -804,27 +804,27 @@ Types imports should follow the ``from typing import ...`` convention. So rather
 
    import typing
 
-   primes = []  # type: typing.List[int]
+   primes: typing.List[int] = []
 
 You should write
 
 .. code-block:: python
 
    from typing import List, Optional, Union
 
-   primes = []  # type: List[int]
+   primes: List[int] = []
 
 ``Optional`` should be used where applicable, so instead of
 
 .. code-block:: python
 
-   maybe_primes = []  # type: List[Union[int, None]]
+   maybe_primes: List[Union[int, None]] = []
 
 You should write
 
 .. code-block:: python
 
-   maybe_primes = []  # type: List[Optional[int]]
+   maybe_primes: List[Optional[int]] = []
 
 In some cases in the code base classes may define class variables that shadow builtins. This causes an issue as described in `Mypy 1775 <https://github.com/python/mypy/issues/1775#issuecomment-310969854>`_. The defensive solution here is to create an unambiguous alias of the builtin and use that without your annotation. For example, if you come across a definition like
 
@@ -840,7 +840,7 @@ The appropriate way to annotate this would be as follows
    str_type = str
 
    class SomeClass2:
-       str = None  # type: str_type
+       str: str_type = None
 
 In some cases you may be tempted to use ``cast`` from the typing module when you know better than the analyzer. This occurs particularly when using custom inference functions. For example
 

pandas/_config/config.py

@@ -58,16 +58,16 @@
 RegisteredOption = namedtuple("RegisteredOption", "key defval doc validator cb")
 
 # holds deprecated option metdata
-_deprecated_options = {}  # type: Dict[str, DeprecatedOption]
+_deprecated_options: Dict[str, DeprecatedOption] = {}
 
 # holds registered option metdata
-_registered_options = {}  # type: Dict[str, RegisteredOption]
+_registered_options: Dict[str, RegisteredOption] = {}
 
 # holds the current values for registered options
-_global_config = {}  # type: Dict[str, str]
+_global_config: Dict[str, str] = {}
 
 # keys which have a special meaning
-_reserved_keys = ["all"]  # type: List[str]
+_reserved_keys: List[str] = ["all"]
 
 
 class OptionError(AttributeError, KeyError):

pandas/_version.py

@@ -47,7 +47,7 @@ class NotThisMethod(Exception):
     pass
 
 
-HANDLERS = {}  # type: Dict[str, Dict[str, Callable]]
+HANDLERS: Dict[str, Dict[str, Callable]] = {}
 
 
 def register_vcs_handler(vcs: str, method: str) -> Callable:  # decorator

pandas/compat/numpy/function.py

@@ -106,7 +106,7 @@ def validate_argmax_with_skipna(skipna, args, kwargs):
     return skipna
 
 
-ARGSORT_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[Union[int, str]]]
+ARGSORT_DEFAULTS: "OrderedDict[str, Optional[Union[int, str]]]" = OrderedDict()
 ARGSORT_DEFAULTS["axis"] = -1
 ARGSORT_DEFAULTS["kind"] = "quicksort"
 ARGSORT_DEFAULTS["order"] = None
@@ -122,7 +122,7 @@ def validate_argmax_with_skipna(skipna, args, kwargs):
 
 # two different signatures of argsort, this second validation
 # for when the `kind` param is supported
-ARGSORT_DEFAULTS_KIND = OrderedDict()  # type: OrderedDict[str, Optional[int]]
+ARGSORT_DEFAULTS_KIND: "OrderedDict[str, Optional[int]]" = OrderedDict()
 ARGSORT_DEFAULTS_KIND["axis"] = -1
 ARGSORT_DEFAULTS_KIND["order"] = None
 validate_argsort_kind = CompatValidator(
@@ -169,14 +169,14 @@ def validate_clip_with_axis(axis, args, kwargs):
     return axis
 
 
-COMPRESS_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Any]
+COMPRESS_DEFAULTS: "OrderedDict[str, Any]" = OrderedDict()
 COMPRESS_DEFAULTS["axis"] = None
 COMPRESS_DEFAULTS["out"] = None
 validate_compress = CompatValidator(
     COMPRESS_DEFAULTS, fname="compress", method="both", max_fname_arg_count=1
 )
 
-CUM_FUNC_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Any]
+CUM_FUNC_DEFAULTS: "OrderedDict[str, Any]" = OrderedDict()
 CUM_FUNC_DEFAULTS["dtype"] = None
 CUM_FUNC_DEFAULTS["out"] = None
 validate_cum_func = CompatValidator(
@@ -202,7 +202,7 @@ def validate_cum_func_with_skipna(skipna, args, kwargs, name):
     return skipna
 
 
-ALLANY_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[bool]]
+ALLANY_DEFAULTS: "OrderedDict[str, Optional[bool]]" = OrderedDict()
 ALLANY_DEFAULTS["dtype"] = None
 ALLANY_DEFAULTS["out"] = None
 ALLANY_DEFAULTS["keepdims"] = False
@@ -224,28 +224,28 @@ def validate_cum_func_with_skipna(skipna, args, kwargs, name):
     MINMAX_DEFAULTS, fname="max", method="both", max_fname_arg_count=1
 )
 
-RESHAPE_DEFAULTS = dict(order="C")  # type: Dict[str, str]
+RESHAPE_DEFAULTS: Dict[str, str] = dict(order="C")
 validate_reshape = CompatValidator(
     RESHAPE_DEFAULTS, fname="reshape", method="both", max_fname_arg_count=1
 )
 
-REPEAT_DEFAULTS = dict(axis=None)  # type: Dict[str, Any]
+REPEAT_DEFAULTS: Dict[str, Any] = dict(axis=None)
 validate_repeat = CompatValidator(
     REPEAT_DEFAULTS, fname="repeat", method="both", max_fname_arg_count=1
 )
 
-ROUND_DEFAULTS = dict(out=None)  # type: Dict[str, Any]
+ROUND_DEFAULTS: Dict[str, Any] = dict(out=None)
 validate_round = CompatValidator(
     ROUND_DEFAULTS, fname="round", method="both", max_fname_arg_count=1
 )
 
-SORT_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[Union[int, str]]]
+SORT_DEFAULTS: "OrderedDict[str, Optional[Union[int, str]]]" = OrderedDict()
 SORT_DEFAULTS["axis"] = -1
 SORT_DEFAULTS["kind"] = "quicksort"
 SORT_DEFAULTS["order"] = None
 validate_sort = CompatValidator(SORT_DEFAULTS, fname="sort", method="kwargs")
 
-STAT_FUNC_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[Any]]
+STAT_FUNC_DEFAULTS: "OrderedDict[str, Optional[Any]]" = OrderedDict()
 STAT_FUNC_DEFAULTS["dtype"] = None
 STAT_FUNC_DEFAULTS["out"] = None
 
@@ -273,13 +273,13 @@ def validate_cum_func_with_skipna(skipna, args, kwargs, name):
     MEDIAN_DEFAULTS, fname="median", method="both", max_fname_arg_count=1
 )
 
-STAT_DDOF_FUNC_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[bool]]
+STAT_DDOF_FUNC_DEFAULTS: "OrderedDict[str, Optional[bool]]" = OrderedDict()
 STAT_DDOF_FUNC_DEFAULTS["dtype"] = None
 STAT_DDOF_FUNC_DEFAULTS["out"] = None
 STAT_DDOF_FUNC_DEFAULTS["keepdims"] = False
 validate_stat_ddof_func = CompatValidator(STAT_DDOF_FUNC_DEFAULTS, method="kwargs")
 
-TAKE_DEFAULTS = OrderedDict()  # type: OrderedDict[str, Optional[str]]
+TAKE_DEFAULTS: "OrderedDict[str, Optional[str]]" = OrderedDict()
 TAKE_DEFAULTS["out"] = None
 TAKE_DEFAULTS["mode"] = "raise"
 validate_take = CompatValidator(TAKE_DEFAULTS, fname="take", method="kwargs")

pandas/core/accessor.py

@@ -11,8 +11,8 @@
 
 
 class DirNamesMixin:
-    _accessors = set()  # type: Set[str]
-    _deprecations = frozenset()  # type: FrozenSet[str]
+    _accessors: Set[str] = set()
+    _deprecations: FrozenSet[str] = frozenset()
 
     def _dir_deletions(self):
         """

pandas/core/algorithms.py

@@ -50,7 +50,7 @@
 from pandas.core.construction import array, extract_array
 from pandas.core.indexers import validate_indices
 
-_shared_docs = {}  # type: Dict[str, str]
+_shared_docs: Dict[str, str] = {}
 
 
 # --------------- #

pandas/core/apply.py

@@ -34,8 +34,9 @@ def frame_apply(
     """ construct and return a row or column based frame apply object """
 
     axis = obj._get_axis_number(axis)
+    klass: Type[FrameApply]
     if axis == 0:
-        klass = FrameRowApply  # type: Type[FrameApply]
+        klass = FrameRowApply
     elif axis == 1:
         klass = FrameColumnApply
 

pandas/core/arrays/base.py

@@ -29,7 +29,7 @@
 
 _not_implemented_message = "{} does not implement {}."
 
-_extension_array_shared_docs = dict()  # type: Dict[str, str]
+_extension_array_shared_docs: Dict[str, str] = dict()
 
 
 def try_cast_to_ea(cls_or_instance, obj, dtype=None):

pandas/core/arrays/datetimelike.py

@@ -51,7 +51,7 @@
 
 
 class AttributesMixin:
-    _data = None  # type: np.ndarray
+    _data: np.ndarray
 
     @classmethod
     def _simple_new(cls, values, **kwargs):

pandas/core/arrays/datetimes.py

@@ -320,7 +320,7 @@ class DatetimeArray(dtl.DatetimeLikeArrayMixin, dtl.TimelikeOps, dtl.DatelikeOps
     # -----------------------------------------------------------------
     # Constructors
 
-    _dtype = None  # type: Union[np.dtype, DatetimeTZDtype]
+    _dtype: Union[np.dtype, DatetimeTZDtype]
     _freq = None
 
     def __init__(self, values, dtype=_NS_DTYPE, freq=None, copy=False):

pandas/core/arrays/integer.py

@@ -40,9 +40,9 @@ class _IntegerDtype(ExtensionDtype):
     The attributes name & type are set when these subclasses are created.
     """
 
-    name = None  # type: str
+    name: str
     base = None
-    type = None  # type: Type
+    type: Type
     na_value = np.nan
 
     def __repr__(self) -> str:

pandas/core/arrays/period.py

@@ -161,7 +161,7 @@ class PeriodArray(dtl.DatetimeLikeArrayMixin, dtl.DatelikeOps):
     _scalar_type = Period
 
     # Names others delegate to us
-    _other_ops = []  # type: List[str]
+    _other_ops: List[str] = []
     _bool_ops = ["is_leap_year"]
     _object_ops = ["start_time", "end_time", "freq"]
     _field_ops = [
@@ -894,9 +894,9 @@ def period_array(
 
     data = np.asarray(data)
 
+    dtype: Optional[PeriodDtype]
     if freq:
-        # typed Optional here because the else block below assigns None
-        dtype = PeriodDtype(freq)  # type: Optional[PeriodDtype]
+        dtype = PeriodDtype(freq)
     else:
         dtype = None
 

pandas/core/arrays/timedeltas.py

@@ -161,8 +161,8 @@ class TimedeltaArray(dtl.DatetimeLikeArrayMixin, dtl.TimelikeOps):
     _scalar_type = Timedelta
     __array_priority__ = 1000
     # define my properties & methods for delegation
-    _other_ops = []  # type: List[str]
-    _bool_ops = []  # type: List[str]
+    _other_ops: List[str] = []
+    _bool_ops: List[str] = []
     _object_ops = ["freq"]
     _field_ops = ["days", "seconds", "microseconds", "nanoseconds"]
     _datetimelike_ops = _field_ops + _object_ops + _bool_ops

pandas/core/base.py

@@ -36,7 +36,7 @@
 from pandas.core.arrays import ExtensionArray
 import pandas.core.nanops as nanops
 
-_shared_docs = dict()  # type: Dict[str, str]
+_shared_docs: Dict[str, str] = dict()
 _indexops_doc_kwargs = dict(
     klass="IndexOpsMixin",
     inplace="",
@@ -603,7 +603,7 @@ def _is_builtin_func(self, arg):
 
 
 class ShallowMixin:
-    _attributes = []  # type: List[str]
+    _attributes: List[str] = []
 
     def _shallow_copy(self, obj=None, **kwargs):
         """
@@ -627,7 +627,7 @@ class IndexOpsMixin:
 
     # ndarray compatibility
     __array_priority__ = 1000
-    _deprecations = frozenset(
+    _deprecations: FrozenSet[str] = frozenset(
         [
             "tolist",  # tolist is not deprecated, just suppressed in the __dir__
             "base",
@@ -637,7 +637,7 @@ class IndexOpsMixin:
             "flags",
             "strides",
         ]
-    )  # type: FrozenSet[str]
+    )
 
     def transpose(self, *args, **kwargs):
         """

pandas/core/computation/expr.py

@@ -378,7 +378,7 @@ class BaseExprVisitor(ast.NodeVisitor):
     preparser : callable
     """
 
-    const_type = Constant  # type: Type[Term]
+    const_type: Type[Term] = Constant
     term_type = Term
 
     binary_ops = _cmp_ops_syms + _bool_ops_syms + _arith_ops_syms

pandas/core/dtypes/base.py

@@ -81,7 +81,7 @@ def __from_arrow__(
     provided for registering virtual subclasses.
     """
 
-    _metadata = ()  # type: Tuple[str, ...]
+    _metadata: Tuple[str, ...] = ()
 
     def __str__(self) -> str:
         return self.name