chore: Update mypy config (#134)

* chore: Update mypy config

* chore: rm unused type ignores

* chore: rm redundant cast

* chore: ignore unreachable false positive

* chore: fix any-return

* chore: fix untyped-call

* chore: fix no-any-return

* doc: clarify cast

Author: msto

ci/mypy.ini

@@ -1,6 +1,16 @@
 [mypy]
-strict_optional = False
-ignore_missing_imports = True
-disallow_untyped_decorators = False
-follow_imports = silent
-disallow_untyped_defs = True
+python_version              = 3.8
+strict_optional             = false
+check_untyped_defs          = true
+disallow_incomplete_defs    = true
+disallow_untyped_calls      = true
+disallow_untyped_decorators = true
+disallow_untyped_defs       = true
+no_implicit_optional        = true
+warn_no_return              = true
+warn_redundant_casts        = true
+warn_return_any             = true
+warn_unreachable            = true
+warn_unused_configs         = true
+warn_unused_ignores         = true
+

fgpyo/fastx/__init__.py

@@ -34,7 +34,6 @@
 from typing import Tuple
 from typing import Type
 from typing import Union
-from typing import cast
 
 from pysam import FastxFile
 from pysam import FastxRecord
@@ -77,7 +76,6 @@ def __next__(self) -> Tuple[FastxRecord, ...]:
                 )
             )
         else:
-            records = cast(Tuple[FastxRecord, ...], records)
             record_names: Set[str] = {self._name_minus_ordinal(record.name) for record in records}
             if len(record_names) != 1:
                 raise ValueError(f"FASTX record names do not all match: {record_names}")

fgpyo/sam/__init__.py

@@ -253,7 +253,7 @@ def _pysam_open(
         :class:`~pysam.AlignmentFile`; may not include "mode".
     """
 
-    if isinstance(path, (str, Path)):  # type: ignore
+    if isinstance(path, (str, Path)):
         if str(path) in _STDIN_PATHS and open_for_reading:
             path = sys.stdin
         elif str(path) in _STDOUT_PATHS and not open_for_reading:
@@ -505,23 +505,23 @@ def from_cigarstring(cls, cigarstring: str) -> "Cigar":
         i = 0
         while i < cigarstring_length:
             if not cigarstring[i].isdigit():
-                raise cls._pretty_cigarstring_exception(cigarstring, i)  # type: ignore
+                raise cls._pretty_cigarstring_exception(cigarstring, i)
             length = int(cigarstring[i])
             i += 1
             while i < cigarstring_length and cigarstring[i].isdigit():
                 length = (length * 10) + int(cigarstring[i])
                 i += 1
             if i == cigarstring_length:
-                raise cls._pretty_cigarstring_exception(cigarstring, i)  # type: ignore
+                raise cls._pretty_cigarstring_exception(cigarstring, i)
             try:
                 operator = CigarOp.from_character(cigarstring[i])
                 elements.append(CigarElement(length, operator))
             except KeyError as ex:
                 # cigar operator was not valid
-                raise cls._pretty_cigarstring_exception(cigarstring, i) from ex  # type: ignore
+                raise cls._pretty_cigarstring_exception(cigarstring, i) from ex
             except IndexError as ex:
                 # missing cigar operator (i == len(cigarstring))
-                raise cls._pretty_cigarstring_exception(cigarstring, i) from ex  # type: ignore
+                raise cls._pretty_cigarstring_exception(cigarstring, i) from ex
             i += 1
         return Cigar(tuple(elements))
 

fgpyo/sam/builder.py

@@ -22,6 +22,7 @@
 from typing import List
 from typing import Optional
 from typing import Tuple
+from typing import cast
 
 import pysam
 from pysam import AlignedSegment
@@ -154,7 +155,7 @@ def _next_name(self) -> str:
 
     def _bases(self, length: int) -> str:
         """Returns a random string of bases of the length requested."""
-        return "".join(self._random.choices("ACGT", k=length))  # type: ignore
+        return "".join(self._random.choices("ACGT", k=length))
 
     def _new_rec(
         self,
@@ -334,13 +335,17 @@ def _set_mate_info(self, r1: pysam.AlignedSegment, r2: pysam.AlignedSegment) ->
 
     def rg(self) -> Dict[str, Any]:
         """Returns the single read group that is defined in the header."""
-        rgs = self._header["RG"]
+        # The `RG` field contains a list of read group mappings
+        # e.g. `[{"ID": "rg1", "PL": "ILLUMINA"}]`
+        rgs = cast(List[Dict[str, Any]], self._header["RG"])
         assert len(rgs) == 1, "Header did not contain exactly one read group!"
         return rgs[0]
 
     def rg_id(self) -> str:
         """Returns the ID of the single read group that is defined in the header."""
-        return self.rg()["ID"]
+        # The read group mapping has mixed types of values (e.g. "PI" is numeric), but the "ID"
+        # field is always a string.
+        return cast(str, self.rg()["ID"])
 
     def add_pair(
         self,
@@ -585,7 +590,7 @@ def to_path(
                 file_handle = fp.file
 
             with sam.writer(
-                file_handle, header=self._samheader, file_type=sam.SamFileType.BAM  # type: ignore
+                file_handle, header=self._samheader, file_type=sam.SamFileType.BAM
             ) as writer:
                 for rec in self._records:
                     if pred(rec):

fgpyo/sam/tests/test_sam.py

@@ -227,7 +227,11 @@ def test_pretty_cigarstring_exception() -> None:
     expected = "10M5[U]4M"
     with pytest.raises(CigarParsingException, match=r".*Malformed cigar") as ex:
         raise Cigar._pretty_cigarstring_exception(cigar, index)
-    assert expected in str(ex)
+
+    # Known issue, `mypy` thinks the `raise` above makes the following unreachable
+    # https://github.com/python/mypy/issues/8985
+    # https://github.com/python/mypy/issues/8766
+    assert expected in str(ex)  # type: ignore[unreachable]
 
     expected = cigar + "[]"
     with pytest.raises(CigarParsingException, match=r".*Malformed cigar") as ex:

fgpyo/util/inspect.py

@@ -44,14 +44,14 @@
     from attr import fields as get_attr_fields
     from attr import fields_dict as get_attr_fields_dict
 
-    Attribute: TypeAlias = attr.Attribute  # type: ignore[name-defined, no-redef]
+    Attribute: TypeAlias = attr.Attribute  # type: ignore[name-defined]
     # dataclasses and attr have internal tokens for missing values, join into a set so that we can
     # check if a value is missing without knowing the type of backing class
     MISSING = frozenset({DATACLASSES_MISSING, attr.NOTHING})
 except ImportError:  # pragma: no cover
     _use_attr = False
     attr = None
-    Attribute: TypeAlias = TypeVar("Attribute", bound=object)  # type: ignore[misc, assignment, no-redef]  # noqa: E501
+    Attribute: TypeAlias = TypeVar("Attribute", bound=object)  # type: ignore[misc, no-redef]  # noqa: E501
 
     # define empty placeholders for getting attr fields as a tuple or dict. They will never be
     # called because the import failed; but they're here to ensure that the function is defined in
@@ -85,7 +85,7 @@ class AttrsInstance(Protocol):  # type: ignore[no-redef]
         __attrs_attrs__: ClassVar[Any]
 
 
-def is_attr_class(cls: type) -> bool:  # type: ignore[arg-type]
+def is_attr_class(cls: type) -> bool:
     """Return True if the class is an attr class, and False otherwise"""
     return hasattr(cls, "__attrs_attrs__")
 
@@ -396,7 +396,7 @@ def get_fields_dict(
 ) -> Mapping[str, FieldType]:
     """Get the fields dict from either a dataclasses or attr dataclass (or instance)"""
     if is_dataclasses_class(cls):
-        return _get_dataclasses_fields_dict(cls)  # type: ignore[arg-type]
+        return _get_dataclasses_fields_dict(cls)
     elif is_attr_class(cls):  # type: ignore[arg-type]
         return get_attr_fields_dict(cls)  # type: ignore[arg-type]
     else:
@@ -408,9 +408,9 @@ def get_fields(
 ) -> Tuple[FieldType, ...]:
     """Get the fields tuple from either a dataclasses or attr dataclass (or instance)"""
     if is_dataclasses_class(cls):
-        return get_dataclasses_fields(cls)  # type: ignore[arg-type]
+        return get_dataclasses_fields(cls)
     elif is_attr_class(cls):  # type: ignore[arg-type]
-        return get_attr_fields(cls)  # type: ignore[arg-type]
+        return get_attr_fields(cls)  # type: ignore[arg-type, no-any-return]
     else:
         raise TypeError("cls must a dataclasses or attr class")
 

fgpyo/util/logging.py

@@ -105,6 +105,7 @@ def __init__(
         verb: str = "Read",
         unit: int = 100000,
     ) -> None:
+        self.printer: Callable[[str], Any]
         if isinstance(printer, Logger):
             self.printer = lambda s: printer.info(s)
         else:
@@ -187,6 +188,8 @@ def _log(
 
         self.printer(f"{self.verb} {self.count:,d} {self.noun}: {coordinate}")
 
+        return None
+
     def log_last(
         self,
     ) -> bool:

fgpyo/util/types.py

@@ -9,6 +9,7 @@
 from typing import Type
 from typing import TypeVar
 from typing import Union
+from typing import cast
 
 UnionType = TypeVar("UnionType", bound="Union")
 EnumType = TypeVar("EnumType", bound="Enum")
@@ -119,7 +120,9 @@ def _make_literal_parser_worker(
     for arg, p in zip(typing.get_args(literal), parsers):
         try:
             if p(value) == arg:
-                return arg
+                # typing.get_args returns `Any` because there's no guarantee on the input type, but
+                # if we're passing it a literal then the returned args will always be `LiteralType`
+                return cast(LiteralType, arg)
         except ValueError:
             pass
     raise InspectException(