(WIP) CI: Check that private functions are not used across modules

ci/code_checks.sh

@@ -232,6 +232,14 @@ if [[ -z "$CHECK" || "$CHECK" == "patterns" ]]; then
     invgrep -R --include=*.{py,pyx} 'xrange' pandas
     RET=$(($RET + $?)) ; echo $MSG "DONE"
 
+    MSG='Check for use of private functions across modules' ; echo $MSG
+    if [[ "$GITHUB_ACTIONS" == "true" ]]; then
+        $BASE_DIR/scripts/validate_unwanted_patterns.py --validation-type="private_function_across_module" --format="##[error]{source_path}:{line_number}:{msg}" pandas/core/
+    else
+        $BASE_DIR/scripts/validate_unwanted_patterns.py --validation-type="private_function_across_module" pandas/core
+    fi
+    RET=$(($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_unwanted_patterns.py

@@ -16,11 +16,18 @@
 import sys
 import token
 import tokenize
-from typing import IO, Callable, Iterable, List, Tuple
+from typing import IO, Callable, Iterable, List, Set, Tuple
 
 FILE_EXTENSIONS_TO_CHECK: Tuple[str, ...] = (".py", ".pyx", ".pxi.ini", ".pxd")
 PATHS_TO_IGNORE: Tuple[str, ...] = ("asv_bench/env",)
 
+PRIVATE_IMPORTS_TO_IGNORE: Set[str] = {
+    "_extension_array_shared_docs",
+    "_index_shared_docs",
+    "_merge_doc",
+    "_shared_docs",
+}
+
 
 def _get_literal_string_prefix_len(token_string: str) -> int:
     """
@@ -115,6 +122,38 @@ def bare_pytest_raises(file_obj: IO[str]) -> Iterable[Tuple[int, str]]:
                 )
 
 
+def private_function_across_module(file_obj: IO[str]) -> Iterable[Tuple[int, str]]:
+    """
+    Checking that a private function is not used across modules.
+
+    Parameters
+    ----------
+    file_obj : IO
+        File-like object containing the Python code to validate.
+
+    Yields
+    ------
+    line_number : int
+        Line number of import statement, that imports the private function.
+    msg : str
+        Explenation of the error.
+    """
+    contents = file_obj.read()
+    tree = ast.parse(contents)
+
+    for node in ast.walk(tree):
+        if not (isinstance(node, ast.Import) or isinstance(node, ast.ImportFrom)):
+            continue
+
+        for module in node.names:
+            module_name = module.name.split(".")[-1]
+            if module_name in PRIVATE_IMPORTS_TO_IGNORE:
+                continue
+
+            if module_name.startswith("_"):
+                yield (node.lineno, f"Import of internal function {repr(module_name)}")
+
+
 def strings_to_concatenate(file_obj: IO[str]) -> Iterable[Tuple[int, str]]:
     """
     This test case is necessary after 'Black' (https://github.com/psf/black),
@@ -358,6 +397,7 @@ def main(
 if __name__ == "__main__":
     available_validation_types: List[str] = [
         "bare_pytest_raises",
+        "private_function_across_module",
         "strings_to_concatenate",
         "strings_with_wrong_placed_whitespace",
     ]