CI: Add test case for unwanted patterns

Author: MomIsBestFriend

ci/code_checks.sh

@@ -212,6 +212,10 @@ if [[ -z "$CHECK" || "$CHECK" == "patterns" ]]; then
     invgrep -R --include=*.{py,pyx} 'xrange' pandas
     RET=$(($RET + $?)) ; echo $MSG "DONE"
 
+    MSG='Check for use of not concatenated strings' ; echo $MSG
+    python $BASE_DIR/scripts/validate_string_concatenation.py pandas
+    RET=$(($RET + $?)) ; echo $MSG "DONE"
+
     MSG='Check that no file in the repo contains trailing whitespaces' ; echo $MSG
     INVGREP_APPEND=" <- trailing whitespaces found"
     invgrep -RI --exclude=\*.{svg,c,cpp,html,js} --exclude-dir=env "\s$" *

scripts/validate_string_concatenation.py

@@ -0,0 +1,131 @@
+#!/usr/bin/env python
+"""
+GH #30454
+
+Check where there is a string that needs to be concatenated.
+
+This is necessary after black formating,
+where for example black transforms this:
+
+>>> foo = (
+...         "bar "
+...         "baz"
+...     )
+
+into this:
+
+>>> foo = ("bar " "baz")
+
+Black is not considering this as an
+issue (see https://github.com/psf/black/issues/1051), so we are checking
+it here.
+"""
+
+import os
+import sys
+import token
+import tokenize
+
+FILE_EXTENSIONS_TO_CHECK = [".pxd", ".py", ".pyx", ".pyx.ini"]
+
+
+def main():
+    path = sys.argv[1]
+
+    if not os.path.exists(path):
+        raise ValueError("Please enter a valid path, to a file/directory.")
+
+    if os.path.isfile(path):
+        # Means that the given path is of a single file.
+        sys.exit(is_concatenated(path))
+
+    status_codes = set()
+    # Means that the given path is of a directory.
+    for subdir, _, files in os.walk(path):
+        for file_name in files:
+            ext = full_ext(os.path.join(subdir, file_name))
+            if ext in FILE_EXTENSIONS_TO_CHECK:
+                status_codes.add(is_concatenated(os.path.join(subdir, file_name)))
+
+    if 1 in status_codes:
+        sys.exit(1)
+
+    sys.exit(0)
+
+
+def full_ext(path):
+    """
+    Get the full file extention name.
+
+    Parameters
+    ----------
+    path : str
+        File path.
+
+    Returns
+    -------
+    str
+        Full extention of a file.
+
+    Notes
+    -----
+    This function is needed only because of file extentions like
+    ` .pxi.ini` for example.
+
+    Examples
+    -------
+
+    With one suffix:
+
+    >>> ext = full_ext('/full/path/to/file.py')
+    >>> ext
+        .py
+
+    Wuth two suffixes:
+
+    >>> ext = full_ext('/full/path/to/file.pxi.ini')
+    >>> ext
+        .pxi.ini
+    """
+    ext_list = [".{suffix}".format(suffix=suffix) for suffix in path.split(".")[1:]]
+    return "".join(ext_list)
+
+
+def is_concatenated(file_path):
+    """
+    Checking if the file containing strings that needs to be concatenated.
+
+    Parameters
+    ----------
+    file_path : str
+        File path pointing to a single file.
+
+    Returns
+    -------
+    int
+        Status code representing if the file needs a fix.
+        0 - All good.
+        1 - Needs to be fixed.
+    """
+    need_fix = False
+    with open(file_path, "r") as file_name:
+        toks = list(tokenize.generate_tokens(file_name.readline))
+        for i in range(len(toks) - 1):
+            tok = toks[i]
+            tok2 = toks[i + 1]
+            if tok[0] == token.STRING and tok[0] == tok2[0]:
+                need_fix = True
+                print(
+                    "{file_path}:{line_number}:\t{start} and {end}".format(
+                        file_path=file_path,
+                        line_number=tok[2][0],
+                        start=tok[1],
+                        end=tok2[1],
+                    )
+                )
+
+    return int(need_fix)
+
+
+if __name__ == "__main__":
+    main()