based regex

Author: KotlinIsland

.mypy/baseline.json

@@ -10,6 +10,7 @@
 - Type-guards narrow in the negative (#553)
 - Conditional types for asymmetric type-guards (#553)
 - Static conditions report an error (#553)
+- Regex groups are special-cased (#531)
 ### Enhancements
 - Show 'narrowed from' in `reveal_type` (#550)
 - `--color-output` is enabled by default (#531)

CHANGELOG.md

@@ -331,3 +331,18 @@ Checked Argument Names
     class B(A):
         @override
         def f(self, b: int): ...  # error: Signature of "f" incompatible with supertype "A"
+
+
+Regex Checks
+------------
+
+.. code-block:: python
+
+    re.compile("as(df")  #  error: missing ), unterminated subpattern at position 0  [regex]
+
+    if m := re.search("(a)?(b)", s):
+        reveal_type(m.groups())  # Revealed type is "(str | None, str)"
+
+    if m := re.search("(?P<foo>a)", s):
+        reveal_type(m.group("foo")
+        reveal_type(m.group("bar")  # error: no such group: 'bar'  [regex]

docs/source/based_features.rst

@@ -80,6 +80,17 @@ Check that typeguard definitions are valid [typeguard-subtype]
     def guard(x: str) -> x is int:  # A type-guard's type must be assignable to its parameter's type. (guard has type "int", parameter has type "str")  [typeguard-subtype]
         ...
 
+.. _code-regex:
+
+Check for regex issues [regex]
+------------------------------
+
+.. code-block:: python
+
+    re.compile("(foo).*)")  # error: unbalanced parenthesis at position 7  [regex]
+    m = re.match("(?P<name>\w+)", s)
+    assert m
+    m.group("nane")  # error: no such group: 'nane'  [regex]
 
 .. _code-bad-cast:
 

docs/source/error_code_list3.rst

@@ -14,13 +14,11 @@
 import sys
 import tempfile
 
-from mypy.util import safe
-
 
 def parse_commit_title(diff: str) -> str:
     m = re.search("\n    ([^ ].*)", diff)
     assert m is not None, "Could not parse diff"
-    return safe(m.group(1))
+    return m.group(1)
 
 
 def main() -> None:

misc/cherry-pick-typeshed.py

@@ -31,8 +31,6 @@
 import sys
 import tempfile
 
-from mypy.util import safe
-
 REVEAL_TYPE_START = "reveal_type("
 REVEAL_TYPE_END = ")"
 
@@ -53,11 +51,7 @@ def run_mypy(mypy_and_args: list[str], filename: str, tmp_name: str) -> str:
 
 def get_revealed_type(line: str, relevant_file: str, relevant_line: int) -> str | None:
     m = re.match(r'(.+?):(\d+): note: Revealed type is "(.*)"$', line)
-    if (
-        m
-        and int(safe(m.group(2))) == relevant_line
-        and os.path.samefile(relevant_file, safe(m.group(1)))
-    ):
+    if m and int(m.group(2)) == relevant_line and os.path.samefile(relevant_file, m.group(1)):
         return m.group(3)
     else:
         return None

misc/find_type.py

@@ -47,8 +47,6 @@
 from typing import Any, Dict, Final
 from typing_extensions import TypeAlias as _TypeAlias
 
-from mypy.util import safe
-
 CACHE_PATH: Final = ".incremental_checker_cache.json"
 MYPY_REPO_URL: Final = "https://github.com/python/mypy.git"
 MYPY_TARGET_FILE: Final = "mypy"
@@ -171,7 +169,7 @@ def filter_daemon_stats(output: str) -> tuple[str, dict[str, Any]]:
     for line in lines:
         m = re.match(r"(\w+)\s+:\s+(.*)", line)
         if m:
-            key, value = [safe(g) for g in m.groups()]
+            key, value = [g for g in m.groups()]
             stats[key] = value
         else:
             output_lines.append(line)

misc/incremental_checker.py

@@ -9,7 +9,6 @@
 from io import StringIO
 
 from mypy.errorcodes import error_codes
-from mypy.util import safe
 
 if sys.version_info >= (3, 11):
     import tomllib
@@ -29,6 +28,7 @@
     TextIO,
     Tuple,
     Union,
+    cast,
 )
 from typing_extensions import TypeAlias as _TypeAlias
 
@@ -45,7 +45,7 @@ def parse_version(v: str | float) -> tuple[int, int]:
     m = re.match(r"\A(\d)\.(\d+)\Z", str(v))
     if not m:
         raise argparse.ArgumentTypeError(f"Invalid python version '{v}' (expected format: 'x.y')")
-    major, minor = int(safe(m.group(1))), int(safe(m.group(2)))
+    major, minor = int(m.group(1)), int(m.group(2))
     if major == 2 and minor == 7:
         pass  # Error raised elsewhere
     elif major == 3:
@@ -68,7 +68,7 @@ def parse_version(v: str | float) -> tuple[int, int]:
 def try_split(v: str | Sequence[str], split_regex: str = "[,]") -> list[str]:
     """Split and trim a str or list of str into a list of str"""
     if isinstance(v, str):
-        return [p.strip() for p in re.split(split_regex, v)]
+        return [p.strip() for p in cast(List[str], re.split(split_regex, v))]
 
     return [p.strip() for p in v]
 

mypy/config_parser.py

@@ -66,6 +66,7 @@ def visit_deleted_type(self, t: DeletedType) -> ProperType:
 
     def visit_instance(self, t: Instance) -> ProperType:
         dup = Instance(t.type, t.args, last_known_value=t.last_known_value)
+        dup.metadata = t.metadata.copy()
         dup.invalid = t.invalid
         return self.copy_common(t, dup)
 

mypy/copytype.py

@@ -277,6 +277,8 @@ def __hash__(self) -> int:
 BAD_CAST: Final[ErrorCode] = ErrorCode(
     "bad-cast", "cast of non-overlapping types", "General", default_enabled=False
 )
+
+REGEX: Final = ErrorCode("regex", "Regex related errors", "General")
 REVEAL: Final = ErrorCode("reveal", "Reveal types at check time", "General")
 
 # Syntax errors are often blocking.