Allow parenthesized implicitly concatenated strings inside calls, to be more compatible with Black.

doc/data/messages/i/implicit-str-concat/details.rst

@@ -24,3 +24,16 @@ for string parameters passed on multiple lines in function calls:
         DeprecationWarning,
         stacklevel=3,
     )
+
+No message will be emitted, though, if you clarify the wanted concatenation with parentheses:
+
+.. code-block:: python
+
+    warnings.warn(
+        (
+            "rotate() is deprecated and will be removed in a future release. "
+            "Use the rotation() context manager instead."
+        ),
+        DeprecationWarning,
+        stacklevel=3,
+    )

doc/whatsnew/fragments/8552.false_positive

@@ -0,0 +1,3 @@
+Allow parenthesized implicitly concatenated strings when `check-str-concat-over-line-jumps` is enabled.
+
+Closes #8552.

pylint/checkers/strings.py

@@ -55,6 +55,11 @@
     "Rb",
     "RB",
 }
+_PAREN_IGNORE_TOKEN_TYPES = (
+    tokenize.NEWLINE,
+    tokenize.NL,
+    tokenize.COMMENT,
+)
 SINGLE_QUOTED_REGEX = re.compile(f"({'|'.join(_PREFIXES)})?'''")
 DOUBLE_QUOTED_REGEX = re.compile(f"({'|'.join(_PREFIXES)})?\"\"\"")
 QUOTE_DELIMITER_REGEX = re.compile(f"({'|'.join(_PREFIXES)})?(\"|')", re.DOTALL)
@@ -710,6 +715,7 @@ def __init__(self, linter: PyLinter) -> None:
             tuple[int, int], tuple[str, tokenize.TokenInfo | None]
         ] = {}
         """Token position -> (token value, next token)."""
+        self._parenthesized_string_tokens: dict[tuple[int, int], bool] = {}
 
     def process_module(self, node: nodes.Module) -> None:
         self._unicode_literals = "unicode_literals" in node.future_imports
@@ -738,10 +744,62 @@ def process_tokens(self, tokens: list[tokenize.TokenInfo]) -> None:
                     # to match with astroid `.col_offset`
                     start = (start[0], len(line[: start[1]].encode(encoding)))
                 self.string_tokens[start] = (str_eval(token), next_token)
+                is_parenthesized = self._is_initial_string_token(
+                    i, tokens
+                ) and self._is_parenthesized(i, tokens)
+                self._parenthesized_string_tokens[start] = is_parenthesized
 
         if self.linter.config.check_quote_consistency:
             self.check_for_consistent_string_delimiters(tokens)
 
+    def _is_initial_string_token(
+        self, index: int, tokens: Sequence[tokenize.TokenInfo]
+    ) -> bool:
+        # Must NOT be preceded by a string literal
+        prev_token = self._find_prev_token(index, tokens)
+        if prev_token and prev_token.type == tokenize.STRING:
+            return False
+        # Must be followed by a string literal token.
+        next_token = self._find_next_token(index, tokens)
+        return bool(next_token and next_token.type == tokenize.STRING)
+
+    def _is_parenthesized(self, index: int, tokens: list[tokenize.TokenInfo]) -> bool:
+        prev_token = self._find_prev_token(
+            index, tokens, ignore=(*_PAREN_IGNORE_TOKEN_TYPES, tokenize.STRING)
+        )
+        if not prev_token or prev_token.type != tokenize.OP or prev_token[1] != "(":
+            return False
+        next_token = self._find_next_token(
+            index, tokens, ignore=(*_PAREN_IGNORE_TOKEN_TYPES, tokenize.STRING)
+        )
+        return bool(
+            next_token and next_token.type == tokenize.OP and next_token[1] == ")"
+        )
+
+    def _find_prev_token(
+        self,
+        index: int,
+        tokens: Sequence[tokenize.TokenInfo],
+        *,
+        ignore: tuple[int, ...] = _PAREN_IGNORE_TOKEN_TYPES,
+    ) -> tokenize.TokenInfo | None:
+        i = index - 1
+        while i >= 0 and tokens[i].type in ignore:
+            i -= 1
+        return tokens[i] if i >= 0 else None
+
+    def _find_next_token(
+        self,
+        index: int,
+        tokens: Sequence[tokenize.TokenInfo],
+        *,
+        ignore: tuple[int, ...] = _PAREN_IGNORE_TOKEN_TYPES,
+    ) -> tokenize.TokenInfo | None:
+        i = index + 1
+        while i < len(tokens) and tokens[i].type in ignore:
+            i += 1
+        return tokens[i] if i < len(tokens) else None
+
     @only_required_for_messages("implicit-str-concat")
     def visit_call(self, node: nodes.Call) -> None:
         self.check_for_concatenated_strings(node.args, "call")
@@ -815,10 +873,18 @@ def check_for_concatenated_strings(
             matching_token, next_token = self.string_tokens[token_index]
             # We detect string concatenation: the AST Const is the
             # combination of 2 string tokens
-            if matching_token != elt.value and next_token is not None:
-                if next_token.type == tokenize.STRING and (
-                    next_token.start[0] == elt.lineno
-                    or self.linter.config.check_str_concat_over_line_jumps
+            if (
+                matching_token != elt.value
+                and next_token is not None
+                and next_token.type == tokenize.STRING
+            ):
+                if next_token.start[0] == elt.lineno or (
+                    self.linter.config.check_str_concat_over_line_jumps
+                    # Allow implicitly concatenated strings in parens.
+                    # See https://github.com/pylint-dev/pylint/issues/8552.
+                    and not self._parenthesized_string_tokens.get(
+                        (elt.lineno, elt.col_offset)
+                    )
                 ):
                     self.add_message(
                         "implicit-str-concat",

tests/functional/i/implicit/implicit_str_concat_multiline.py

@@ -3,9 +3,46 @@
 TEST_TUPLE = ('a', 'b'  # [implicit-str-concat]
               'c')
 
+# See https://github.com/pylint-dev/pylint/issues/8552.
+PARENTHESIZED_IS_OK = [
+    "a",
+    (
+        "Lorem ipsum dolor sit amet, consectetur adipiscing elit,"
+        " sed do eiusmod tempor incididunt ut labore et dolore "
+    ),
+]
+
+# Single argument without trailing comma is OK:
 print(
-    "Lorem ipsum dolor sit amet, consectetur adipiscing elit," # [implicit-str-concat]
+    "Lorem ipsum dolor sit amet, consectetur adipiscing elit,"
     " sed do eiusmod tempor incididunt ut labore et dolore "
     "magna aliqua. Ut enim ad minim veniam, quis nostrud "
     "exercitation ullamco laboris nisi ut aliquip ex ea "
 )
+
+# Implicit concatenated strings on the same line always raises:
+print(
+    "Lorem ipsum dolor sit amet, ""consectetur adipiscing elit,"  # [implicit-str-concat]
+    "Lorem ipsum dolor sit amet, consectetur adipiscing elit,"
+    " sed do eiusmod tempor incididunt ut labore et dolore "
+    "magna aliqua. Ut enim ad minim veniam, quis nostrud "
+    "exercitation ullamco laboris nisi ut aliquip ex ea "
+)
+
+# Explicitly wrapping in parens with a trailing comma is OK:
+print(
+    (
+        "Lorem ipsum dolor sit amet, consectetur adipiscing elit,"
+        " sed do eiusmod tempor incididunt ut labore et dolore "
+        "magna aliqua. Ut enim ad minim veniam, quis nostrud "
+        "exercitation ullamco laboris nisi ut aliquip ex ea "
+    ),
+)
+
+# But NOT OK when there is a trailing comma and NOT wrapped in parens:
+print(
+    "Lorem ipsum dolor sit amet, consectetur adipiscing elit," # [implicit-str-concat]
+    " sed do eiusmod tempor incididunt ut labore et dolore "
+    "magna aliqua. Ut enim ad minim veniam, quis nostrud "
+    "exercitation ullamco laboris nisi ut aliquip ex ea ",
+)

tests/functional/i/implicit/implicit_str_concat_multiline.txt

@@ -1,2 +1,3 @@
 implicit-str-concat:3:0:None:None::Implicit string concatenation found in tuple:HIGH
-implicit-str-concat:7:0:None:None::Implicit string concatenation found in call:HIGH
+implicit-str-concat:25:0:None:None::Implicit string concatenation found in call:HIGH
+implicit-str-concat:44:0:None:None::Implicit string concatenation found in call:HIGH