Convert lint-include-guards.sh to python

Specify encoding when reading header files, add docstring

Update test/lint/lint-include-guards.py  include guard count logic

Co-authored-by: Kevin Musgrave <tkm45@cornell.edu>

Update test/lint/lint-include-guards.py by removing whitespace

2 files changed, 100 insertions, 30 deletions

diff --git a/test/lint/lint-include-guards.py b/test/lint/lint-include-guards.py
new file mode 100755
index 0000000000..86284517d5
--- /dev/null
+++ b/test/lint/lint-include-guards.py
@@ -0,0 +1,100 @@
+#!/usr/bin/env python3
+#
+# Copyright (c) 2018-2022 The Bitcoin Core developers
+# Distributed under the MIT software license, see the accompanying
+# file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+"""
+Check include guards.
+"""
+
+import re
+import sys
+from subprocess import check_output
+from typing import List
+
+
+HEADER_ID_PREFIX = 'BITCOIN_'
+HEADER_ID_SUFFIX = '_H'
+
+EXCLUDE_FILES_WITH_PREFIX = ['src/crypto/ctaes',
+                             'src/leveldb',
+                             'src/crc32c',
+                             'src/secp256k1',
+                             'src/minisketch',
+                             'src/univalue',
+                             'src/tinyformat.h',
+                             'src/bench/nanobench.h',
+                             'src/test/fuzz/FuzzedDataProvider.h']
+
+
+def _get_header_file_lst() -> List[str]:
+    """ Helper function to get a list of header filepaths to be
+        checked for include guards.
+    """
+    git_cmd_lst = ['git', 'ls-files', '--', '*.h']
+    header_file_lst = check_output(
+        git_cmd_lst).decode('utf-8').splitlines()
+
+    header_file_lst = [hf for hf in header_file_lst
+                       if not any(ef in hf for ef
+                                  in EXCLUDE_FILES_WITH_PREFIX)]
+
+    return header_file_lst
+
+
+def _get_header_id(header_file: str) -> str:
+    """ Helper function to get the header id from a header file
+        string.
+
+        eg: 'src/wallet/walletdb.h' -> 'BITCOIN_WALLET_WALLETDB_H'
+
+    Args:
+        header_file: Filepath to header file.
+
+    Returns:
+        The header id.
+    """
+    header_id_base = header_file.split('/')[1:]
+    header_id_base = '_'.join(header_id_base)
+    header_id_base = header_id_base.replace('.h', '').replace('-', '_')
+    header_id_base = header_id_base.upper()
+
+    header_id = f'{HEADER_ID_PREFIX}{header_id_base}{HEADER_ID_SUFFIX}'
+
+    return header_id
+
+
+def main():
+    exit_code = 0
+
+    header_file_lst = _get_header_file_lst()
+    for header_file in header_file_lst:
+        header_id = _get_header_id(header_file)
+
+        regex_pattern = f'^#(ifndef|define|endif //) {header_id}'
+
+        with open(header_file, 'r', encoding='utf-8') as f:
+            header_file_contents = f.readlines()
+
+        count = 0
+        for header_file_contents_string in header_file_contents:
+            include_guard_lst = re.findall(
+                regex_pattern, header_file_contents_string)
+
+            count += len(include_guard_lst)
+
+        if count != 3:
+            print(f'{header_file} seems to be missing the expected '
+                  'include guard:')
+            print(f'  #ifndef {header_id}')
+            print(f'  #define {header_id}')
+            print('  ...')
+            print(f'  #endif // {header_id}\n')
+            exit_code = 1
+
+    sys.exit(exit_code)
+
+
+if __name__ == '__main__':
+    main()
diff --git a/test/lint/lint-include-guards.sh b/test/lint/lint-include-guards.sh
deleted file mode 100755
index f14218aa74..0000000000
--- a/test/lint/lint-include-guards.sh
+++ /dev/null
@@ -1,30 +0,0 @@
-#!/usr/bin/env bash
-#
-# Copyright (c) 2018-2021 The Bitcoin Core developers
-# Distributed under the MIT software license, see the accompanying
-# file COPYING or http://www.opensource.org/licenses/mit-license.php.
-#
-# Check include guards.
-
-export LC_ALL=C
-HEADER_ID_PREFIX="BITCOIN_"
-HEADER_ID_SUFFIX="_H"
-
-REGEXP_EXCLUDE_FILES_WITH_PREFIX="src/(crypto/ctaes/|leveldb/|crc32c/|secp256k1/|minisketch/|test/fuzz/FuzzedDataProvider.h|tinyformat.h|bench/nanobench.h|univalue/)"
-
-EXIT_CODE=0
-for HEADER_FILE in $(git ls-files -- "*.h" | grep -vE "^${REGEXP_EXCLUDE_FILES_WITH_PREFIX}")
-do
-    HEADER_ID_BASE=$(cut -f2- -d/ <<< "${HEADER_FILE}" | sed "s/\.h$//g" | tr / _ | tr - _ | tr "[:lower:]" "[:upper:]")
-    HEADER_ID="${HEADER_ID_PREFIX}${HEADER_ID_BASE}${HEADER_ID_SUFFIX}"
-    if [[ $(grep --count --extended-regexp "^#(ifndef|define|endif //) ${HEADER_ID}" "${HEADER_FILE}") != 3 ]]; then
-        echo "${HEADER_FILE} seems to be missing the expected include guard:"
-        echo "  #ifndef ${HEADER_ID}"
-        echo "  #define ${HEADER_ID}"
-        echo "  ..."
-        echo "  #endif // ${HEADER_ID}"
-        echo
-        EXIT_CODE=1
-    fi
-done
-exit ${EXIT_CODE}