Fixes for datapythonista's review

Author: MomIsBestFriend

ci/code_checks.sh

@@ -100,8 +100,13 @@ if [[ -z "$CHECK" || "$CHECK" == "lint" ]]; then
     cpplint --quiet --extensions=c,h --headers=h --recursive --filter=-readability/casting,-runtime/int,-build/include_subdir pandas/_libs/src/*.h pandas/_libs/src/parser pandas/_libs/ujson pandas/_libs/tslibs/src/datetime pandas/_libs/*.cpp
     RET=$(($RET + $?)) ; echo $MSG "DONE"
 
+    # Get the output format for azure/githubaction from datapythonista
     MSG='Check for use of not concatenated strings' ; echo $MSG
-    $BASE_DIR/scripts/validate_string_concatenation.py pandas
+    if [[ "$GITHUB_ACTIONS" == "true" ]]; then
+        $BASE_DIR/scripts/validate_string_concatenation.py --format="[error]{source_path}:{line_number}:\t BETWEEN {start} AND {end}" .
+    else
+        $BASE_DIR/scripts/validate_string_concatenation.py .
+    fi
     RET=$(($RET + $?)) ; echo $MSG "DONE"
 
     echo "isort --version-number"

scripts/validate_string_concatenation.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python3
+#!/usr/bin/env python
 """
 GH #30454
 
@@ -12,7 +12,6 @@
 ...         "baz"
 ...     )
 
-
 into this:
 
 >>> foo = ("bar " "baz")
@@ -22,65 +21,108 @@
 so we are checking it here.
 """
 
+import argparse
 import os
 import sys
 import token
 import tokenize
-from typing import FrozenSet, Generator, List
+from typing import Dict, Generator, List
+
+FILE_EXTENSIONS_TO_CHECK = (".py", ".pyx", ".pyx.ini", ".pxd")
+
+
+def main(source_path: str = ".", output_format: str) -> bool:
+    """
+    Main entry point of the script.
+
+    Parameters
+    ----------
+    source_path : str, default '.'
+        Source path representing path to a file/directory.
+    output_format : str
+        Output format of the script.
+
+    Returns
+    -------
+    bool
+        True if found any strings that needs to be concatenated.
+
+    Raises
+    ------
+    ValueError
+        If the `source_path` is not pointing to existing file/directory.
+    """
+    if not os.path.exists(source_path):
+        raise ValueError(
+            "Please enter a valid path, pointing to a valid file/directory."
+        )
 
-FILE_EXTENSIONS_TO_CHECK: FrozenSet[str] = frozenset(
-    (".pxd", ".py", ".pyx", ".pyx.ini")
-)
+    is_failed: bool = False
 
+    if os.path.isfile(source_path):
+        for values in strings_to_concatenate(source_path):
+            is_failed = True
+            print(output_format.format(**values))
 
-def strings_to_concatenate(file_path: str) -> Generator[str, None, None]:
+    for subdir, _, files in os.walk(source_path):
+        for file_name in files:
+            if any(
+                file_name.endswith(extension) for extension in FILE_EXTENSIONS_TO_CHECK
+            ):
+                for values in strings_to_concatenate(os.path.join(subdir, file_name)):
+                    is_failed = True
+                    print(output_format.format(**values))
+    return is_failed
+
+
+def strings_to_concatenate(source_path: str) -> Generator[Dict[str, str], None, None]:
     """
     Yielding the strings that needs to be concatenated in a given file.
 
     Parameters
     ----------
-    file_path : str
+    source_path : str
         File path pointing to a single file.
 
     Yields
     ------
-    str
-        Message containing info about the string that needs to be concatenated.
+    dict of {str: str}
+        Containing:
+            source_path
+                Source file path.
+            line_number
+                Line number of unconcatenated string.
+            start
+                Starting string of unconcatenated string.
+            end
+                Ending string of unconcatenated string.
     """
-    with open(file_path, "r") as file_name:
+    with open(source_path, "r") as file_name:
         tokens: List = list(tokenize.generate_tokens(file_name.readline))
 
     for current_token, next_token in zip(tokens, tokens[1:]):
         if current_token[0] == next_token[0] == token.STRING:
-            line_number = current_token[2][0]
-            start = current_token[1]
-            end = next_token[1]
-            yield f"{file_path}:{line_number}:\t between {start} and {end}\n"
+            yield {
+                "source_path": source_path,
+                "line_number": current_token[2][0],
+                "start": current_token[1],
+                "end": next_token[1],
+            }
 
 
 if __name__ == "__main__":
-    path: str = sys.argv[1]
+    parser = argparse.ArgumentParser(description="Validate concatenated strings")
 
-    if not os.path.exists(path):
-        raise ValueError("Please enter a valid path, to a file/directory.")
+    parser.add_argument(
+        "path", nargs="?", default=".", help="Source path of file/directory to check."
+    )
+    parser.add_argument(
+        "--format",
+        "-f",
+        default="{source_path}:{line_number}:\t BETWEEN {start} AND {end}",
+        help="Output format of the unconcatenated strings.",
+    )
 
-    failed: bool = False
-
-    if os.path.isfile(path):
-        for msg in strings_to_concatenate(path):
-            if msg:
-                failed = True
-                print(msg)
-
-    for subdir, _, files in os.walk(path):
-        for file_name in files:
-            if any(
-                file_name.endswith(extension) for extension in FILE_EXTENSIONS_TO_CHECK
-            ):
-                file_extension = os.path.join(subdir, file_name)
+    args = parser.parse_args()
 
-                for msg in strings_to_concatenate(os.path.join(subdir, file_name)):
-                    if msg:
-                        failed = True
-                        print(msg)
-    sys.exit(failed)
+    sys.exit(main(source_path=args.path, output_format=args.format))