diff options
Diffstat (limited to 'test/lint')
24 files changed, 1301 insertions, 865 deletions
diff --git a/test/lint/lint-circular-dependencies.py b/test/lint/lint-circular-dependencies.py new file mode 100755 index 0000000000..24163ec787 --- /dev/null +++ b/test/lint/lint-circular-dependencies.py @@ -0,0 +1,86 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 2020-2022 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 for circular dependencies + +import glob +import os +import re +import subprocess +import sys + +EXPECTED_CIRCULAR_DEPENDENCIES = ( + "chainparamsbase -> util/system -> chainparamsbase", + "node/blockstorage -> validation -> node/blockstorage", + "index/blockfilterindex -> node/blockstorage -> validation -> index/blockfilterindex", + "index/base -> validation -> index/blockfilterindex -> index/base", + "index/coinstatsindex -> node/coinstats -> index/coinstatsindex", + "policy/fees -> txmempool -> policy/fees", + "qt/addresstablemodel -> qt/walletmodel -> qt/addresstablemodel", + "qt/recentrequeststablemodel -> qt/walletmodel -> qt/recentrequeststablemodel", + "qt/sendcoinsdialog -> qt/walletmodel -> qt/sendcoinsdialog", + "qt/transactiontablemodel -> qt/walletmodel -> qt/transactiontablemodel", + "wallet/fees -> wallet/wallet -> wallet/fees", + "wallet/wallet -> wallet/walletdb -> wallet/wallet", + "node/coinstats -> validation -> node/coinstats", +) + +CODE_DIR = "src" + + +def main(): + circular_dependencies = [] + exit_code = 0 + os.chdir( + CODE_DIR + ) # We change dir before globbing since glob.glob's root_dir option is only available in Python 3.10 + + # Using glob.glob since subprocess.run's globbing won't work without shell=True + files = [] + for path in ["*", "*/*", "*/*/*"]: + for extension in ["h", "cpp"]: + files.extend(glob.glob(f"{path}.{extension}")) + + command = ["python3", "../contrib/devtools/circular-dependencies.py", *files] + dependencies_output = subprocess.run( + command, + stdout=subprocess.PIPE, + universal_newlines=True, + ) + + for dependency_str in dependencies_output.stdout.rstrip().split("\n"): + circular_dependencies.append( + re.sub("^Circular dependency: ", "", dependency_str) + ) + + # Check for an unexpected dependencies + for dependency in circular_dependencies: + if dependency not in EXPECTED_CIRCULAR_DEPENDENCIES: + exit_code = 1 + print( + f'A new circular dependency in the form of "{dependency}" appears to have been introduced.\n', + file=sys.stderr, + ) + + # Check for missing expected dependencies + for expected_dependency in EXPECTED_CIRCULAR_DEPENDENCIES: + if expected_dependency not in circular_dependencies: + exit_code = 1 + print( + f'Good job! The circular dependency "{expected_dependency}" is no longer present.', + ) + print( + f"Please remove it from EXPECTED_CIRCULAR_DEPENDENCIES in {__file__}", + ) + print( + "to make sure this circular dependency is not accidentally reintroduced.\n", + ) + + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-circular-dependencies.sh b/test/lint/lint-circular-dependencies.sh deleted file mode 100755 index 69185090d1..0000000000 --- a/test/lint/lint-circular-dependencies.sh +++ /dev/null @@ -1,65 +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 for circular dependencies - -export LC_ALL=C - -EXPECTED_CIRCULAR_DEPENDENCIES=( - "chainparamsbase -> util/system -> chainparamsbase" - "node/blockstorage -> validation -> node/blockstorage" - "index/blockfilterindex -> node/blockstorage -> validation -> index/blockfilterindex" - "index/base -> validation -> index/blockfilterindex -> index/base" - "index/coinstatsindex -> node/coinstats -> index/coinstatsindex" - "policy/fees -> txmempool -> policy/fees" - "qt/addresstablemodel -> qt/walletmodel -> qt/addresstablemodel" - "qt/recentrequeststablemodel -> qt/walletmodel -> qt/recentrequeststablemodel" - "qt/sendcoinsdialog -> qt/walletmodel -> qt/sendcoinsdialog" - "qt/transactiontablemodel -> qt/walletmodel -> qt/transactiontablemodel" - "wallet/fees -> wallet/wallet -> wallet/fees" - "wallet/wallet -> wallet/walletdb -> wallet/wallet" - "node/coinstats -> validation -> node/coinstats" -) - -EXIT_CODE=0 - -CIRCULAR_DEPENDENCIES=() - -IFS=$'\n' -for CIRC in $(cd src && ../contrib/devtools/circular-dependencies.py {*,*/*,*/*/*}.{h,cpp} | sed -e 's/^Circular dependency: //'); do - CIRCULAR_DEPENDENCIES+=( "$CIRC" ) - IS_EXPECTED_CIRC=0 - for EXPECTED_CIRC in "${EXPECTED_CIRCULAR_DEPENDENCIES[@]}"; do - if [[ "${CIRC}" == "${EXPECTED_CIRC}" ]]; then - IS_EXPECTED_CIRC=1 - break - fi - done - if [[ ${IS_EXPECTED_CIRC} == 0 ]]; then - echo "A new circular dependency in the form of \"${CIRC}\" appears to have been introduced." - echo - EXIT_CODE=1 - fi -done - -for EXPECTED_CIRC in "${EXPECTED_CIRCULAR_DEPENDENCIES[@]}"; do - IS_PRESENT_EXPECTED_CIRC=0 - for CIRC in "${CIRCULAR_DEPENDENCIES[@]}"; do - if [[ "${CIRC}" == "${EXPECTED_CIRC}" ]]; then - IS_PRESENT_EXPECTED_CIRC=1 - break - fi - done - if [[ ${IS_PRESENT_EXPECTED_CIRC} == 0 ]]; then - echo "Good job! The circular dependency \"${EXPECTED_CIRC}\" is no longer present." - echo "Please remove it from EXPECTED_CIRCULAR_DEPENDENCIES in $0" - echo "to make sure this circular dependency is not accidentally reintroduced." - echo - EXIT_CODE=1 - fi -done - -exit ${EXIT_CODE} diff --git a/test/lint/lint-format-strings.py b/test/lint/lint-format-strings.py new file mode 100755 index 0000000000..5a36da11fd --- /dev/null +++ b/test/lint/lint-format-strings.py @@ -0,0 +1,98 @@ +#!/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. +# + +""" +Lint format strings: This program checks that the number of arguments passed +to a variadic format string function matches the number of format specifiers +in the format string. +""" + +import subprocess +import re +import sys + +FUNCTION_NAMES_AND_NUMBER_OF_LEADING_ARGUMENTS = [ + 'FatalError,0', + 'fprintf,1', + 'tfm::format,1', # Assuming tfm::::format(std::ostream&, ... + 'LogConnectFailure,1', + 'LogPrint,1', + 'LogPrintf,0', + 'printf,0', + 'snprintf,2', + 'sprintf,1', + 'strprintf,0', + 'vfprintf,1', + 'vprintf,1', + 'vsnprintf,1', + 'vsprintf,1', + 'WalletLogPrintf,0', +] +RUN_LINT_FILE = 'test/lint/run-lint-format-strings.py' + +def check_doctest(): + command = [ + 'python3', + '-m', + 'doctest', + RUN_LINT_FILE, + ] + try: + subprocess.run(command, check = True) + except subprocess.CalledProcessError: + sys.exit(1) + +def get_matching_files(function_name): + command = [ + 'git', + 'grep', + '--full-name', + '-l', + function_name, + '--', + '*.c', + '*.cpp', + '*.h', + ] + try: + return subprocess.check_output(command, stderr = subprocess.STDOUT).decode('utf-8').splitlines() + except subprocess.CalledProcessError as e: + if e.returncode > 1: # return code is 1 when match is empty + print(e.output.decode('utf-8'), end='') + sys.exit(1) + return [] + +def main(): + exit_code = 0 + check_doctest() + for s in FUNCTION_NAMES_AND_NUMBER_OF_LEADING_ARGUMENTS: + function_name, skip_arguments = s.split(',') + matching_files = get_matching_files(function_name) + + matching_files_filtered = [] + for matching_file in matching_files: + if not re.search('^src/(leveldb|secp256k1|minisketch|tinyformat|univalue|test/fuzz/strprintf.cpp)', matching_file): + matching_files_filtered.append(matching_file) + matching_files_filtered.sort() + + run_lint_args = [ + RUN_LINT_FILE, + '--skip-arguments', + skip_arguments, + function_name, + ] + run_lint_args.extend(matching_files_filtered) + + try: + subprocess.run(run_lint_args, check = True) + except subprocess.CalledProcessError: + exit_code = 1 + + sys.exit(exit_code) + +if __name__ == '__main__': + main() diff --git a/test/lint/lint-format-strings.sh b/test/lint/lint-format-strings.sh deleted file mode 100755 index 73730f16b3..0000000000 --- a/test/lint/lint-format-strings.sh +++ /dev/null @@ -1,44 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2018-2020 The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -# -# Lint format strings: This program checks that the number of arguments passed -# to a variadic format string function matches the number of format specifiers -# in the format string. - -export LC_ALL=C - -FUNCTION_NAMES_AND_NUMBER_OF_LEADING_ARGUMENTS=( - "FatalError,0" - "fprintf,1" - "tfm::format,1" # Assuming tfm::::format(std::ostream&, ... - "LogConnectFailure,1" - "LogPrint,1" - "LogPrintf,0" - "printf,0" - "snprintf,2" - "sprintf,1" - "strprintf,0" - "vfprintf,1" - "vprintf,1" - "vsnprintf,1" - "vsprintf,1" - "WalletLogPrintf,0" -) - -EXIT_CODE=0 -if ! python3 -m doctest "test/lint/run-lint-format-strings.py"; then - EXIT_CODE=1 -fi -for S in "${FUNCTION_NAMES_AND_NUMBER_OF_LEADING_ARGUMENTS[@]}"; do - IFS="," read -r FUNCTION_NAME SKIP_ARGUMENTS <<< "${S}" - for MATCHING_FILE in $(git grep --full-name -l "${FUNCTION_NAME}" -- "*.c" "*.cpp" "*.h" | sort | grep -vE "^src/(leveldb|secp256k1|minisketch|tinyformat|univalue|test/fuzz/strprintf.cpp)"); do - MATCHING_FILES+=("${MATCHING_FILE}") - done - if ! "test/lint/run-lint-format-strings.py" --skip-arguments "${SKIP_ARGUMENTS}" "${FUNCTION_NAME}" "${MATCHING_FILES[@]}"; then - EXIT_CODE=1 - fi -done -exit ${EXIT_CODE} diff --git a/test/lint/lint-git-commit-check.py b/test/lint/lint-git-commit-check.py new file mode 100755 index 0000000000..a1d03370e8 --- /dev/null +++ b/test/lint/lint-git-commit-check.py @@ -0,0 +1,63 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 2020-2022 The Bitcoin Core developers +# Distributed under the MIT software license, see the accompanying +# file COPYING or http://www.opensource.org/licenses/mit-license.php. +# +# Linter to check that commit messages have a new line before the body +# or no body at all + +import argparse +import os +import sys + +from subprocess import check_output + + +def parse_args(): + """Parse command line arguments.""" + parser = argparse.ArgumentParser( + description=""" + Linter to check that commit messages have a new line before + the body or no body at all. + """, + epilog=f""" + You can manually set the commit-range with the COMMIT_RANGE + environment variable (e.g. "COMMIT_RANGE='47ba2c3...ee50c9e' + {sys.argv[0]}"). Defaults to current merge base when neither + prev-commits nor the environment variable is set. + """) + + parser.add_argument("--prev-commits", "-p", required=False, help="The previous n commits to check") + + return parser.parse_args() + + +def main(): + args = parse_args() + exit_code = 0 + + if not os.getenv("COMMIT_RANGE"): + if args.prev_commits: + commit_range = "HEAD~" + args.prev_commits + "...HEAD" + else: + # This assumes that the target branch of the pull request will be master. + merge_base = check_output(["git", "merge-base", "HEAD", "master"], universal_newlines=True, encoding="utf8").rstrip("\n") + commit_range = merge_base + "..HEAD" + else: + commit_range = os.getenv("COMMIT_RANGE") + + commit_hashes = check_output(["git", "log", commit_range, "--format=%H"], universal_newlines=True, encoding="utf8").splitlines() + + for hash in commit_hashes: + commit_info = check_output(["git", "log", "--format=%B", "-n", "1", hash], universal_newlines=True, encoding="utf8").splitlines() + if len(commit_info) >= 2: + if commit_info[1]: + print(f"The subject line of commit hash {hash} is followed by a non-empty line. Subject lines should always be followed by a blank line.") + exit_code = 1 + + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-git-commit-check.sh b/test/lint/lint-git-commit-check.sh deleted file mode 100755 index f77373ed00..0000000000 --- a/test/lint/lint-git-commit-check.sh +++ /dev/null @@ -1,48 +0,0 @@ -#!/usr/bin/env bash -# Copyright (c) 2020-2021 The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -# -# Linter to check that commit messages have a new line before the body -# or no body at all - -export LC_ALL=C - -EXIT_CODE=0 - -while getopts "?" opt; do - case $opt in - ?) - echo "Usage: $0 [N]" - echo " COMMIT_RANGE='<commit range>' $0" - echo " $0 -?" - echo "Checks unmerged commits, the previous N commits, or a commit range." - echo "COMMIT_RANGE='47ba2c3...ee50c9e' $0" - exit ${EXIT_CODE} - ;; - esac -done - -if [ -z "${COMMIT_RANGE}" ]; then - if [ -n "$1" ]; then - COMMIT_RANGE="HEAD~$1...HEAD" - else - # This assumes that the target branch of the pull request will be master. - MERGE_BASE=$(git merge-base HEAD master) - COMMIT_RANGE="$MERGE_BASE..HEAD" - fi -fi - -while IFS= read -r commit_hash || [[ -n "$commit_hash" ]]; do - n_line=0 - while IFS= read -r line || [[ -n "$line" ]]; do - n_line=$((n_line+1)) - length=${#line} - if [ $n_line -eq 2 ] && [ "$length" -ne 0 ]; then - echo "The subject line of commit hash ${commit_hash} is followed by a non-empty line. Subject lines should always be followed by a blank line." - EXIT_CODE=1 - fi - done < <(git log --format=%B -n 1 "$commit_hash") -done < <(git log "${COMMIT_RANGE}" --format=%H) - -exit ${EXIT_CODE} 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} diff --git a/test/lint/lint-includes.py b/test/lint/lint-includes.py new file mode 100755 index 0000000000..b29c7f8b4d --- /dev/null +++ b/test/lint/lint-includes.py @@ -0,0 +1,179 @@ +#!/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 for duplicate includes. +# Guard against accidental introduction of new Boost dependencies. +# Check includes: Check for duplicate includes. Enforce bracket syntax includes. + +import os +import re +import sys + +from subprocess import check_output, CalledProcessError + + +EXCLUDED_DIRS = ["src/leveldb/", + "src/crc32c/", + "src/secp256k1/", + "src/minisketch/", + "src/univalue/"] + +EXPECTED_BOOST_INCLUDES = ["boost/algorithm/string.hpp", + "boost/algorithm/string/classification.hpp", + "boost/algorithm/string/replace.hpp", + "boost/algorithm/string/split.hpp", + "boost/date_time/posix_time/posix_time.hpp", + "boost/multi_index/hashed_index.hpp", + "boost/multi_index/ordered_index.hpp", + "boost/multi_index/sequenced_index.hpp", + "boost/multi_index_container.hpp", + "boost/process.hpp", + "boost/signals2/connection.hpp", + "boost/signals2/optional_last_value.hpp", + "boost/signals2/signal.hpp", + "boost/test/included/unit_test.hpp", + "boost/test/unit_test.hpp"] + + +def get_toplevel(): + return check_output(["git", "rev-parse", "--show-toplevel"], universal_newlines=True, encoding="utf8").rstrip("\n") + + +def list_files_by_suffix(suffixes): + exclude_args = [":(exclude)" + dir for dir in EXCLUDED_DIRS] + + files_list = check_output(["git", "ls-files", "src"] + exclude_args, universal_newlines=True, encoding="utf8").splitlines() + + return [file for file in files_list if file.endswith(suffixes)] + + +def find_duplicate_includes(include_list): + tempset = set() + duplicates = set() + + for inclusion in include_list: + if inclusion in tempset: + duplicates.add(inclusion) + else: + tempset.add(inclusion) + + return duplicates + + +def find_included_cpps(): + included_cpps = list() + + try: + included_cpps = check_output(["git", "grep", "-E", r"^#include [<\"][^>\"]+\.cpp[>\"]", "--", "*.cpp", "*.h"], universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + return included_cpps + + +def find_extra_boosts(): + included_boosts = list() + filtered_included_boost_set = set() + exclusion_set = set() + + try: + included_boosts = check_output(["git", "grep", "-E", r"^#include <boost/", "--", "*.cpp", "*.h"], universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + for boost in included_boosts: + filtered_included_boost_set.add(re.findall(r'(?<=\<).+?(?=\>)', boost)[0]) + + for expected_boost in EXPECTED_BOOST_INCLUDES: + for boost in filtered_included_boost_set: + if expected_boost in boost: + exclusion_set.add(boost) + + extra_boosts = set(filtered_included_boost_set.difference(exclusion_set)) + + return extra_boosts + + +def find_quote_syntax_inclusions(): + exclude_args = [":(exclude)" + dir for dir in EXCLUDED_DIRS] + quote_syntax_inclusions = list() + + try: + quote_syntax_inclusions = check_output(["git", "grep", r"^#include \"", "--", "*.cpp", "*.h"] + exclude_args, universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + return quote_syntax_inclusions + + +def main(): + exit_code = 0 + + os.chdir(get_toplevel()) + + # Check for duplicate includes + for filename in list_files_by_suffix((".cpp", ".h")): + with open(filename, "r", encoding="utf8") as file: + include_list = [line.rstrip("\n") for line in file if re.match(r"^#include", line)] + + duplicates = find_duplicate_includes(include_list) + + if duplicates: + print(f"Duplicate include(s) in {filename}:") + for duplicate in duplicates: + print(duplicate) + print("") + exit_code = 1 + + # Check if code includes .cpp-files + included_cpps = find_included_cpps() + + if included_cpps: + print("The following files #include .cpp files:") + for included_cpp in included_cpps: + print(included_cpp) + print("") + exit_code = 1 + + # Guard against accidental introduction of new Boost dependencies + extra_boosts = find_extra_boosts() + + if extra_boosts: + for boost in extra_boosts: + print(f"A new Boost dependency in the form of \"{boost}\" appears to have been introduced:") + print(check_output(["git", "grep", boost, "--", "*.cpp", "*.h"], universal_newlines=True, encoding="utf8")) + exit_code = 1 + + # Check if Boost dependencies are no longer used + for expected_boost in EXPECTED_BOOST_INCLUDES: + try: + check_output(["git", "grep", "-q", r"^#include <%s>" % expected_boost, "--", "*.cpp", "*.h"], universal_newlines=True, encoding="utf8") + except CalledProcessError as e: + if e.returncode > 1: + raise e + else: + print(f"Good job! The Boost dependency \"{expected_boost}\" is no longer used. " + "Please remove it from EXPECTED_BOOST_INCLUDES in test/lint/lint-includes.py " + "to make sure this dependency is not accidentally reintroduced.\n") + exit_code = 1 + + # Enforce bracket syntax includes + quote_syntax_inclusions = find_quote_syntax_inclusions() + + if quote_syntax_inclusions: + print("Please use bracket syntax includes (\"#include <foo.h>\") instead of quote syntax includes:") + for quote_syntax_inclusion in quote_syntax_inclusions: + print(quote_syntax_inclusion) + exit_code = 1 + + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-includes.sh b/test/lint/lint-includes.sh deleted file mode 100755 index 9e72831ee9..0000000000 --- a/test/lint/lint-includes.sh +++ /dev/null @@ -1,103 +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 for duplicate includes. -# Guard against accidental introduction of new Boost dependencies. -# Check includes: Check for duplicate includes. Enforce bracket syntax includes. - -export LC_ALL=C -IGNORE_REGEXP="/(leveldb|secp256k1|minisketch|univalue|crc32c)/" - -# cd to root folder of git repo for git ls-files to work properly -cd "$(dirname "$0")/../.." || exit 1 - -filter_suffix() { - git ls-files | grep -E "^src/.*\.${1}"'$' | grep -Ev "${IGNORE_REGEXP}" -} - -EXIT_CODE=0 - -for HEADER_FILE in $(filter_suffix h); do - DUPLICATE_INCLUDES_IN_HEADER_FILE=$(grep -E "^#include " < "${HEADER_FILE}" | sort | uniq -d) - if [[ ${DUPLICATE_INCLUDES_IN_HEADER_FILE} != "" ]]; then - echo "Duplicate include(s) in ${HEADER_FILE}:" - echo "${DUPLICATE_INCLUDES_IN_HEADER_FILE}" - echo - EXIT_CODE=1 - fi -done - -for CPP_FILE in $(filter_suffix cpp); do - DUPLICATE_INCLUDES_IN_CPP_FILE=$(grep -E "^#include " < "${CPP_FILE}" | sort | uniq -d) - if [[ ${DUPLICATE_INCLUDES_IN_CPP_FILE} != "" ]]; then - echo "Duplicate include(s) in ${CPP_FILE}:" - echo "${DUPLICATE_INCLUDES_IN_CPP_FILE}" - echo - EXIT_CODE=1 - fi -done - -INCLUDED_CPP_FILES=$(git grep -E "^#include [<\"][^>\"]+\.cpp[>\"]" -- "*.cpp" "*.h") -if [[ ${INCLUDED_CPP_FILES} != "" ]]; then - echo "The following files #include .cpp files:" - echo "${INCLUDED_CPP_FILES}" - echo - EXIT_CODE=1 -fi - -EXPECTED_BOOST_INCLUDES=( - boost/algorithm/string.hpp - boost/algorithm/string/classification.hpp - boost/algorithm/string/replace.hpp - boost/algorithm/string/split.hpp - boost/date_time/posix_time/posix_time.hpp - boost/multi_index/hashed_index.hpp - boost/multi_index/ordered_index.hpp - boost/multi_index/sequenced_index.hpp - boost/multi_index_container.hpp - boost/process.hpp - boost/signals2/connection.hpp - boost/signals2/optional_last_value.hpp - boost/signals2/signal.hpp - boost/test/included/unit_test.hpp - boost/test/unit_test.hpp -) - -for BOOST_INCLUDE in $(git grep '^#include <boost/' -- "*.cpp" "*.h" | cut -f2 -d: | cut -f2 -d'<' | cut -f1 -d'>' | sort -u); do - IS_EXPECTED_INCLUDE=0 - for EXPECTED_BOOST_INCLUDE in "${EXPECTED_BOOST_INCLUDES[@]}"; do - if [[ "${BOOST_INCLUDE}" == "${EXPECTED_BOOST_INCLUDE}" ]]; then - IS_EXPECTED_INCLUDE=1 - break - fi - done - if [[ ${IS_EXPECTED_INCLUDE} == 0 ]]; then - EXIT_CODE=1 - echo "A new Boost dependency in the form of \"${BOOST_INCLUDE}\" appears to have been introduced:" - git grep "${BOOST_INCLUDE}" -- "*.cpp" "*.h" - echo - fi -done - -for EXPECTED_BOOST_INCLUDE in "${EXPECTED_BOOST_INCLUDES[@]}"; do - if ! git grep -q "^#include <${EXPECTED_BOOST_INCLUDE}>" -- "*.cpp" "*.h"; then - echo "Good job! The Boost dependency \"${EXPECTED_BOOST_INCLUDE}\" is no longer used." - echo "Please remove it from EXPECTED_BOOST_INCLUDES in $0" - echo "to make sure this dependency is not accidentally reintroduced." - echo - EXIT_CODE=1 - fi -done - -QUOTE_SYNTAX_INCLUDES=$(git grep '^#include "' -- "*.cpp" "*.h" | grep -Ev "${IGNORE_REGEXP}") -if [[ ${QUOTE_SYNTAX_INCLUDES} != "" ]]; then - echo "Please use bracket syntax includes (\"#include <foo.h>\") instead of quote syntax includes:" - echo "${QUOTE_SYNTAX_INCLUDES}" - echo - EXIT_CODE=1 -fi - -exit ${EXIT_CODE} diff --git a/test/lint/lint-locale-dependence.py b/test/lint/lint-locale-dependence.py new file mode 100755 index 0000000000..2abf1be6b3 --- /dev/null +++ b/test/lint/lint-locale-dependence.py @@ -0,0 +1,259 @@ +#!/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. +# +# Be aware that bitcoind and bitcoin-qt differ in terms of localization: Qt +# opts in to POSIX localization by running setlocale(LC_ALL, "") on startup, +# whereas no such call is made in bitcoind. +# +# Qt runs setlocale(LC_ALL, "") on initialization. This installs the locale +# specified by the user's LC_ALL (or LC_*) environment variable as the new +# C locale. +# +# In contrast, bitcoind does not opt in to localization -- no call to +# setlocale(LC_ALL, "") is made and the environment variables LC_* are +# thus ignored. +# +# This results in situations where bitcoind is guaranteed to be running +# with the classic locale ("C") whereas the locale of bitcoin-qt will vary +# depending on the user's environment variables. +# +# An example: Assuming the environment variable LC_ALL=de_DE then the +# call std::to_string(1.23) will return "1.230000" in bitcoind but +# "1,230000" in bitcoin-qt. +# +# From the Qt documentation: +# "On Unix/Linux Qt is configured to use the system locale settings by default. +# This can cause a conflict when using POSIX functions, for instance, when +# converting between data types such as floats and strings, since the notation +# may differ between locales. To get around this problem, call the POSIX function +# setlocale(LC_NUMERIC,"C") right after initializing QApplication, QGuiApplication +# or QCoreApplication to reset the locale that is used for number formatting to +# "C"-locale." +# +# See https://doc.qt.io/qt-5/qcoreapplication.html#locale-settings and +# https://stackoverflow.com/a/34878283 for more details. +# +# TODO: Reduce KNOWN_VIOLATIONS by replacing uses of locale dependent snprintf with strprintf. + +import re +import sys + +from subprocess import check_output, CalledProcessError + + +KNOWN_VIOLATIONS = [ + "src/dbwrapper.cpp:.*vsnprintf", + "src/test/dbwrapper_tests.cpp:.*snprintf", + "src/test/fuzz/locale.cpp:.*setlocale", + "src/test/fuzz/string.cpp:.*strtol", + "src/test/fuzz/string.cpp:.*strtoul", + "src/test/util_tests.cpp:.*strtoll" +] + +REGEXP_EXTERNAL_DEPENDENCIES_EXCLUSIONS = [ + "src/crypto/ctaes/", + "src/leveldb/", + "src/secp256k1/", + "src/minisketch/", + "src/tinyformat.h", + "src/univalue/" +] + +LOCALE_DEPENDENT_FUNCTIONS = [ + "alphasort", # LC_COLLATE (via strcoll) + "asctime", # LC_TIME (directly) + "asprintf", # (via vasprintf) + "atof", # LC_NUMERIC (via strtod) + "atoi", # LC_NUMERIC (via strtol) + "atol", # LC_NUMERIC (via strtol) + "atoll", # (via strtoll) + "atoq", + "btowc", # LC_CTYPE (directly) + "ctime", # (via asctime or localtime) + "dprintf", # (via vdprintf) + "fgetwc", + "fgetws", + "fold_case", # boost::locale::fold_case + "fprintf", # (via vfprintf) + "fputwc", + "fputws", + "fscanf", # (via __vfscanf) + "fwprintf", # (via __vfwprintf) + "getdate", # via __getdate_r => isspace // __localtime_r + "getwc", + "getwchar", + "is_digit", # boost::algorithm::is_digit + "is_space", # boost::algorithm::is_space + "isalnum", # LC_CTYPE + "isalpha", # LC_CTYPE + "isblank", # LC_CTYPE + "iscntrl", # LC_CTYPE + "isctype", # LC_CTYPE + "isdigit", # LC_CTYPE + "isgraph", # LC_CTYPE + "islower", # LC_CTYPE + "isprint", # LC_CTYPE + "ispunct", # LC_CTYPE + "isspace", # LC_CTYPE + "isupper", # LC_CTYPE + "iswalnum", # LC_CTYPE + "iswalpha", # LC_CTYPE + "iswblank", # LC_CTYPE + "iswcntrl", # LC_CTYPE + "iswctype", # LC_CTYPE + "iswdigit", # LC_CTYPE + "iswgraph", # LC_CTYPE + "iswlower", # LC_CTYPE + "iswprint", # LC_CTYPE + "iswpunct", # LC_CTYPE + "iswspace", # LC_CTYPE + "iswupper", # LC_CTYPE + "iswxdigit", # LC_CTYPE + "isxdigit", # LC_CTYPE + "localeconv", # LC_NUMERIC + LC_MONETARY + "mblen", # LC_CTYPE + "mbrlen", + "mbrtowc", + "mbsinit", + "mbsnrtowcs", + "mbsrtowcs", + "mbstowcs", # LC_CTYPE + "mbtowc", # LC_CTYPE + "mktime", + "normalize", # boost::locale::normalize + "printf", # LC_NUMERIC + "putwc", + "putwchar", + "scanf", # LC_NUMERIC + "setlocale", + "snprintf", + "sprintf", + "sscanf", + "std::locale::global", + "std::to_string", + "stod", + "stof", + "stoi", + "stol", + "stold", + "stoll", + "stoul", + "stoull", + "strcasecmp", + "strcasestr", + "strcoll", # LC_COLLATE + #"strerror", + "strfmon", + "strftime", # LC_TIME + "strncasecmp", + "strptime", + "strtod", # LC_NUMERIC + "strtof", + "strtoimax", + "strtol", # LC_NUMERIC + "strtold", + "strtoll", + "strtoq", + "strtoul", # LC_NUMERIC + "strtoull", + "strtoumax", + "strtouq", + "strxfrm", # LC_COLLATE + "swprintf", + "to_lower", # boost::locale::to_lower + "to_title", # boost::locale::to_title + "to_upper", # boost::locale::to_upper + "tolower", # LC_CTYPE + "toupper", # LC_CTYPE + "towctrans", + "towlower", # LC_CTYPE + "towupper", # LC_CTYPE + "trim", # boost::algorithm::trim + "trim_left", # boost::algorithm::trim_left + "trim_right", # boost::algorithm::trim_right + "ungetwc", + "vasprintf", + "vdprintf", + "versionsort", + "vfprintf", + "vfscanf", + "vfwprintf", + "vprintf", + "vscanf", + "vsnprintf", + "vsprintf", + "vsscanf", + "vswprintf", + "vwprintf", + "wcrtomb", + "wcscasecmp", + "wcscoll", # LC_COLLATE + "wcsftime", # LC_TIME + "wcsncasecmp", + "wcsnrtombs", + "wcsrtombs", + "wcstod", # LC_NUMERIC + "wcstof", + "wcstoimax", + "wcstol", # LC_NUMERIC + "wcstold", + "wcstoll", + "wcstombs", # LC_CTYPE + "wcstoul", # LC_NUMERIC + "wcstoull", + "wcstoumax", + "wcswidth", + "wcsxfrm", # LC_COLLATE + "wctob", + "wctomb", # LC_CTYPE + "wctrans", + "wctype", + "wcwidth", + "wprintf" +] + + +def find_locale_dependent_function_uses(): + regexp_locale_dependent_functions = "|".join(LOCALE_DEPENDENT_FUNCTIONS) + exclude_args = [":(exclude)" + excl for excl in REGEXP_EXTERNAL_DEPENDENCIES_EXCLUSIONS] + git_grep_command = ["git", "grep", "-E", "[^a-zA-Z0-9_\\`'\"<>](" + regexp_locale_dependent_functions + "(_r|_s)?)[^a-zA-Z0-9_\\`'\"<>]", "--", "*.cpp", "*.h"] + exclude_args + git_grep_output = list() + + try: + git_grep_output = check_output(git_grep_command, universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + return git_grep_output + + +def main(): + exit_code = 0 + + regexp_ignore_known_violations = "|".join(KNOWN_VIOLATIONS) + git_grep_output = find_locale_dependent_function_uses() + + for locale_dependent_function in LOCALE_DEPENDENT_FUNCTIONS: + matches = [line for line in git_grep_output + if re.search("[^a-zA-Z0-9_\\`'\"<>]" + locale_dependent_function + "(_r|_s)?[^a-zA-Z0-9_\\`'\"<>]", line) + and not re.search("\\.(c|cpp|h):\\s*(//|\\*|/\\*|\").*" + locale_dependent_function, line) + and not re.search(regexp_ignore_known_violations, line)] + if matches: + print(f"The locale dependent function {locale_dependent_function}(...) appears to be used:") + for match in matches: + print(match) + print("") + exit_code = 1 + + if exit_code == 1: + print("Unnecessary locale depedence can cause bugs that are very tricky to isolate and fix. Please avoid using locale dependent functions if possible.\n") + print(f"Advice not applicable in this specific case? Add an exception by updating the ignore list in {sys.argv[0]}") + + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-locale-dependence.sh b/test/lint/lint-locale-dependence.sh deleted file mode 100755 index 7d608eed6a..0000000000 --- a/test/lint/lint-locale-dependence.sh +++ /dev/null @@ -1,241 +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. - -export LC_ALL=C - -# Be aware that bitcoind and bitcoin-qt differ in terms of localization: Qt -# opts in to POSIX localization by running setlocale(LC_ALL, "") on startup, -# whereas no such call is made in bitcoind. -# -# Qt runs setlocale(LC_ALL, "") on initialization. This installs the locale -# specified by the user's LC_ALL (or LC_*) environment variable as the new -# C locale. -# -# In contrast, bitcoind does not opt in to localization -- no call to -# setlocale(LC_ALL, "") is made and the environment variables LC_* are -# thus ignored. -# -# This results in situations where bitcoind is guaranteed to be running -# with the classic locale ("C") whereas the locale of bitcoin-qt will vary -# depending on the user's environment variables. -# -# An example: Assuming the environment variable LC_ALL=de_DE then the -# call std::to_string(1.23) will return "1.230000" in bitcoind but -# "1,230000" in bitcoin-qt. -# -# From the Qt documentation: -# "On Unix/Linux Qt is configured to use the system locale settings by default. -# This can cause a conflict when using POSIX functions, for instance, when -# converting between data types such as floats and strings, since the notation -# may differ between locales. To get around this problem, call the POSIX function -# setlocale(LC_NUMERIC,"C") right after initializing QApplication, QGuiApplication -# or QCoreApplication to reset the locale that is used for number formatting to -# "C"-locale." -# -# See https://doc.qt.io/qt-5/qcoreapplication.html#locale-settings and -# https://stackoverflow.com/a/34878283 for more details. - -# TODO: Reduce KNOWN_VIOLATIONS by replacing uses of locale dependent snprintf with strprintf. -KNOWN_VIOLATIONS=( - "src/dbwrapper.cpp:.*vsnprintf" - "src/test/dbwrapper_tests.cpp:.*snprintf" - "src/test/fuzz/locale.cpp" - "src/test/fuzz/string.cpp" - "src/test/util_tests.cpp" -) - -REGEXP_IGNORE_EXTERNAL_DEPENDENCIES="^src/(crypto/ctaes/|leveldb/|secp256k1/|minisketch/|tinyformat.h|univalue/)" - -LOCALE_DEPENDENT_FUNCTIONS=( - alphasort # LC_COLLATE (via strcoll) - asctime # LC_TIME (directly) - asprintf # (via vasprintf) - atof # LC_NUMERIC (via strtod) - atoi # LC_NUMERIC (via strtol) - atol # LC_NUMERIC (via strtol) - atoll # (via strtoll) - atoq - btowc # LC_CTYPE (directly) - ctime # (via asctime or localtime) - dprintf # (via vdprintf) - fgetwc - fgetws - fold_case # boost::locale::fold_case - fprintf # (via vfprintf) - fputwc - fputws - fscanf # (via __vfscanf) - fwprintf # (via __vfwprintf) - getdate # via __getdate_r => isspace // __localtime_r - getwc - getwchar - is_digit # boost::algorithm::is_digit - is_space # boost::algorithm::is_space - isalnum # LC_CTYPE - isalpha # LC_CTYPE - isblank # LC_CTYPE - iscntrl # LC_CTYPE - isctype # LC_CTYPE - isdigit # LC_CTYPE - isgraph # LC_CTYPE - islower # LC_CTYPE - isprint # LC_CTYPE - ispunct # LC_CTYPE - isspace # LC_CTYPE - isupper # LC_CTYPE - iswalnum # LC_CTYPE - iswalpha # LC_CTYPE - iswblank # LC_CTYPE - iswcntrl # LC_CTYPE - iswctype # LC_CTYPE - iswdigit # LC_CTYPE - iswgraph # LC_CTYPE - iswlower # LC_CTYPE - iswprint # LC_CTYPE - iswpunct # LC_CTYPE - iswspace # LC_CTYPE - iswupper # LC_CTYPE - iswxdigit # LC_CTYPE - isxdigit # LC_CTYPE - localeconv # LC_NUMERIC + LC_MONETARY - mblen # LC_CTYPE - mbrlen - mbrtowc - mbsinit - mbsnrtowcs - mbsrtowcs - mbstowcs # LC_CTYPE - mbtowc # LC_CTYPE - mktime - normalize # boost::locale::normalize - printf # LC_NUMERIC - putwc - putwchar - scanf # LC_NUMERIC - setlocale - snprintf - sprintf - sscanf - std::locale::global - std::to_string - stod - stof - stoi - stol - stold - stoll - stoul - stoull - strcasecmp - strcasestr - strcoll # LC_COLLATE -# strerror - strfmon - strftime # LC_TIME - strncasecmp - strptime - strtod # LC_NUMERIC - strtof - strtoimax - strtol # LC_NUMERIC - strtold - strtoll - strtoq - strtoul # LC_NUMERIC - strtoull - strtoumax - strtouq - strxfrm # LC_COLLATE - swprintf - to_lower # boost::locale::to_lower - to_title # boost::locale::to_title - to_upper # boost::locale::to_upper - tolower # LC_CTYPE - toupper # LC_CTYPE - towctrans - towlower # LC_CTYPE - towupper # LC_CTYPE - trim # boost::algorithm::trim - trim_left # boost::algorithm::trim_left - trim_right # boost::algorithm::trim_right - ungetwc - vasprintf - vdprintf - versionsort - vfprintf - vfscanf - vfwprintf - vprintf - vscanf - vsnprintf - vsprintf - vsscanf - vswprintf - vwprintf - wcrtomb - wcscasecmp - wcscoll # LC_COLLATE - wcsftime # LC_TIME - wcsncasecmp - wcsnrtombs - wcsrtombs - wcstod # LC_NUMERIC - wcstof - wcstoimax - wcstol # LC_NUMERIC - wcstold - wcstoll - wcstombs # LC_CTYPE - wcstoul # LC_NUMERIC - wcstoull - wcstoumax - wcswidth - wcsxfrm # LC_COLLATE - wctob - wctomb # LC_CTYPE - wctrans - wctype - wcwidth - wprintf -) - -function join_array { - local IFS="$1" - shift - echo "$*" -} - -REGEXP_IGNORE_KNOWN_VIOLATIONS=$(join_array "|" "${KNOWN_VIOLATIONS[@]}") - -# Invoke "git grep" only once in order to minimize run-time -REGEXP_LOCALE_DEPENDENT_FUNCTIONS=$(join_array "|" "${LOCALE_DEPENDENT_FUNCTIONS[@]}") -GIT_GREP_OUTPUT=$(git grep -E "[^a-zA-Z0-9_\`'\"<>](${REGEXP_LOCALE_DEPENDENT_FUNCTIONS}(_r|_s)?)[^a-zA-Z0-9_\`'\"<>]" -- "*.cpp" "*.h") - -EXIT_CODE=0 -for LOCALE_DEPENDENT_FUNCTION in "${LOCALE_DEPENDENT_FUNCTIONS[@]}"; do - MATCHES=$(grep -E "[^a-zA-Z0-9_\`'\"<>]${LOCALE_DEPENDENT_FUNCTION}(_r|_s)?[^a-zA-Z0-9_\`'\"<>]" <<< "${GIT_GREP_OUTPUT}" | \ - grep -vE "\.(c|cpp|h):\s*(//|\*|/\*|\").*${LOCALE_DEPENDENT_FUNCTION}") - if [[ ${REGEXP_IGNORE_EXTERNAL_DEPENDENCIES} != "" ]]; then - MATCHES=$(grep -vE "${REGEXP_IGNORE_EXTERNAL_DEPENDENCIES}" <<< "${MATCHES}") - fi - if [[ ${REGEXP_IGNORE_KNOWN_VIOLATIONS} != "" ]]; then - MATCHES=$(grep -vE "${REGEXP_IGNORE_KNOWN_VIOLATIONS}" <<< "${MATCHES}") - fi - if [[ ${MATCHES} != "" ]]; then - echo "The locale dependent function ${LOCALE_DEPENDENT_FUNCTION}(...) appears to be used:" - echo "${MATCHES}" - echo - EXIT_CODE=1 - fi -done -if [[ ${EXIT_CODE} != 0 ]]; then - echo "Unnecessary locale dependence can cause bugs that are very" - echo "tricky to isolate and fix. Please avoid using locale dependent" - echo "functions if possible." - echo - echo "Advice not applicable in this specific case? Add an exception" - echo "by updating the ignore list in $0" -fi -exit ${EXIT_CODE} diff --git a/test/lint/lint-python-utf8-encoding.py b/test/lint/lint-python-utf8-encoding.py new file mode 100755 index 0000000000..62fdc34d50 --- /dev/null +++ b/test/lint/lint-python-utf8-encoding.py @@ -0,0 +1,73 @@ +#!/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. +# +# Make sure we explicitly open all text files using UTF-8 (or ASCII) encoding to +# avoid potential issues on the BSDs where the locale is not always set. + +import sys +import re + +from subprocess import check_output, CalledProcessError + +EXCLUDED_DIRS = ["src/crc32c/"] + + +def get_exclude_args(): + return [":(exclude)" + dir for dir in EXCLUDED_DIRS] + + +def check_fileopens(): + fileopens = list() + + try: + fileopens = check_output(["git", "grep", r" open(", "--", "*.py"] + get_exclude_args(), universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + filtered_fileopens = [fileopen for fileopen in fileopens if not re.search(r"encoding=.(ascii|utf8|utf-8).|open\([^,]*, ['\"][^'\"]*b[^'\"]*['\"]", fileopen)] + + return filtered_fileopens + + +def check_checked_outputs(): + checked_outputs = list() + + try: + checked_outputs = check_output(["git", "grep", "check_output(", "--", "*.py"] + get_exclude_args(), universal_newlines=True, encoding="utf8").splitlines() + except CalledProcessError as e: + if e.returncode > 1: + raise e + + filtered_checked_outputs = [checked_output for checked_output in checked_outputs if re.search(r"universal_newlines=True", checked_output) and not re.search(r"encoding=.(ascii|utf8|utf-8).", checked_output)] + + return filtered_checked_outputs + + +def main(): + exit_code = 0 + + nonexplicit_utf8_fileopens = check_fileopens() + if nonexplicit_utf8_fileopens: + print("Python's open(...) seems to be used to open text files without explicitly specifying encoding='utf8':\n") + for fileopen in nonexplicit_utf8_fileopens: + print(fileopen) + exit_code = 1 + + nonexplicit_utf8_checked_outputs = check_checked_outputs() + if nonexplicit_utf8_checked_outputs: + if nonexplicit_utf8_fileopens: + print("\n") + print("Python's check_output(...) seems to be used to get program outputs without explicitly specifying encoding='utf8':\n") + for checked_output in nonexplicit_utf8_checked_outputs: + print(checked_output) + exit_code = 1 + + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-python-utf8-encoding.sh b/test/lint/lint-python-utf8-encoding.sh deleted file mode 100755 index 6e5b18fc23..0000000000 --- a/test/lint/lint-python-utf8-encoding.sh +++ /dev/null @@ -1,28 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2018-2020 The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -# -# Make sure we explicitly open all text files using UTF-8 (or ASCII) encoding to -# avoid potential issues on the BSDs where the locale is not always set. - -export LC_ALL=C -EXIT_CODE=0 -OUTPUT=$(git grep " open(" -- "*.py" ":(exclude)src/crc32c/" | grep -vE "encoding=.(ascii|utf8|utf-8)." | grep -vE "open\([^,]*, ['\"][^'\"]*b[^'\"]*['\"]") -if [[ ${OUTPUT} != "" ]]; then - echo "Python's open(...) seems to be used to open text files without explicitly" - echo "specifying encoding=\"utf8\":" - echo - echo "${OUTPUT}" - EXIT_CODE=1 -fi -OUTPUT=$(git grep "check_output(" -- "*.py" ":(exclude)src/crc32c/"| grep "universal_newlines=True" | grep -vE "encoding=.(ascii|utf8|utf-8).") -if [[ ${OUTPUT} != "" ]]; then - echo "Python's check_output(...) seems to be used to get program outputs without explicitly" - echo "specifying encoding=\"utf8\":" - echo - echo "${OUTPUT}" - EXIT_CODE=1 -fi -exit ${EXIT_CODE} diff --git a/test/lint/lint-python.py b/test/lint/lint-python.py new file mode 100755 index 0000000000..4d16facfea --- /dev/null +++ b/test/lint/lint-python.py @@ -0,0 +1,131 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 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 for specified flake8 and mypy warnings in python files. +""" + +import os +import pkg_resources +import subprocess +import sys + +DEPS = ['flake8', 'mypy', 'pyzmq'] +MYPY_CACHE_DIR = f"{os.getenv('BASE_ROOT_DIR', '')}/test/.mypy_cache" +FILES_ARGS = ['git', 'ls-files', 'test/functional/*.py', 'contrib/devtools/*.py'] + +ENABLED = ( + 'E101,' # indentation contains mixed spaces and tabs + 'E112,' # expected an indented block + 'E113,' # unexpected indentation + 'E115,' # expected an indented block (comment) + 'E116,' # unexpected indentation (comment) + 'E125,' # continuation line with same indent as next logical line + 'E129,' # visually indented line with same indent as next logical line + 'E131,' # continuation line unaligned for hanging indent + 'E133,' # closing bracket is missing indentation + 'E223,' # tab before operator + 'E224,' # tab after operator + 'E242,' # tab after ',' + 'E266,' # too many leading '#' for block comment + 'E271,' # multiple spaces after keyword + 'E272,' # multiple spaces before keyword + 'E273,' # tab after keyword + 'E274,' # tab before keyword + 'E275,' # missing whitespace after keyword + 'E304,' # blank lines found after function decorator + 'E306,' # expected 1 blank line before a nested definition + 'E401,' # multiple imports on one line + 'E402,' # module level import not at top of file + 'E502,' # the backslash is redundant between brackets + 'E701,' # multiple statements on one line (colon) + 'E702,' # multiple statements on one line (semicolon) + 'E703,' # statement ends with a semicolon + 'E711,' # comparison to None should be 'if cond is None:' + 'E714,' # test for object identity should be "is not" + 'E721,' # do not compare types, use "isinstance()" + 'E742,' # do not define classes named "l", "O", or "I" + 'E743,' # do not define functions named "l", "O", or "I" + 'E901,' # SyntaxError: invalid syntax + 'E902,' # TokenError: EOF in multi-line string + 'F401,' # module imported but unused + 'F402,' # import module from line N shadowed by loop variable + 'F403,' # 'from foo_module import *' used; unable to detect undefined names + 'F404,' # future import(s) name after other statements + 'F405,' # foo_function may be undefined, or defined from star imports: bar_module + 'F406,' # "from module import *" only allowed at module level + 'F407,' # an undefined __future__ feature name was imported + 'F601,' # dictionary key name repeated with different values + 'F602,' # dictionary key variable name repeated with different values + 'F621,' # too many expressions in an assignment with star-unpacking + 'F622,' # two or more starred expressions in an assignment (a, *b, *c = d) + 'F631,' # assertion test is a tuple, which are always True + 'F632,' # use ==/!= to compare str, bytes, and int literals + 'F701,' # a break statement outside of a while or for loop + 'F702,' # a continue statement outside of a while or for loop + 'F703,' # a continue statement in a finally block in a loop + 'F704,' # a yield or yield from statement outside of a function + 'F705,' # a return statement with arguments inside a generator + 'F706,' # a return statement outside of a function/method + 'F707,' # an except: block as not the last exception handler + 'F811,' # redefinition of unused name from line N + 'F812,' # list comprehension redefines 'foo' from line N + 'F821,' # undefined name 'Foo' + 'F822,' # undefined name name in __all__ + 'F823,' # local variable name … referenced before assignment + 'F831,' # duplicate argument name in function definition + 'F841,' # local variable 'foo' is assigned to but never used + 'W191,' # indentation contains tabs + 'W291,' # trailing whitespace + 'W292,' # no newline at end of file + 'W293,' # blank line contains whitespace + 'W601,' # .has_key() is deprecated, use "in" + 'W602,' # deprecated form of raising exception + 'W603,' # "<>" is deprecated, use "!=" + 'W604,' # backticks are deprecated, use "repr()" + 'W605,' # invalid escape sequence "x" + 'W606,' # 'async' and 'await' are reserved keywords starting with Python 3.7 +) + + +def check_dependencies(): + working_set = {pkg.key for pkg in pkg_resources.working_set} + + for dep in DEPS: + if dep not in working_set: + print(f"Skipping Python linting since {dep} is not installed.") + exit(0) + + +def main(): + check_dependencies() + + if len(sys.argv) > 1: + flake8_files = sys.argv[1:] + else: + files_args = ['git', 'ls-files', '*.py'] + flake8_files = subprocess.check_output(files_args).decode("utf-8").splitlines() + + flake8_args = ['flake8', '--ignore=B,C,E,F,I,N,W', f'--select={ENABLED}'] + flake8_files + flake8_env = os.environ.copy() + flake8_env["PYTHONWARNINGS"] = "ignore" + + try: + subprocess.check_call(flake8_args, env=flake8_env) + except subprocess.CalledProcessError: + exit(1) + + mypy_files = subprocess.check_output(FILES_ARGS).decode("utf-8").splitlines() + mypy_args = ['mypy', '--show-error-codes'] + mypy_files + + try: + subprocess.check_call(mypy_args) + except subprocess.CalledProcessError: + exit(1) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-python.sh b/test/lint/lint-python.sh deleted file mode 100755 index 7d7857d325..0000000000 --- a/test/lint/lint-python.sh +++ /dev/null @@ -1,111 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2017-2021 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 for specified flake8 warnings in python files. - -export LC_ALL=C -export MYPY_CACHE_DIR="${BASE_ROOT_DIR}/test/.mypy_cache" - -enabled=( - E101 # indentation contains mixed spaces and tabs - E112 # expected an indented block - E113 # unexpected indentation - E115 # expected an indented block (comment) - E116 # unexpected indentation (comment) - E125 # continuation line with same indent as next logical line - E129 # visually indented line with same indent as next logical line - E131 # continuation line unaligned for hanging indent - E133 # closing bracket is missing indentation - E223 # tab before operator - E224 # tab after operator - E242 # tab after ',' - E266 # too many leading '#' for block comment - E271 # multiple spaces after keyword - E272 # multiple spaces before keyword - E273 # tab after keyword - E274 # tab before keyword - E275 # missing whitespace after keyword - E304 # blank lines found after function decorator - E306 # expected 1 blank line before a nested definition - E401 # multiple imports on one line - E402 # module level import not at top of file - E502 # the backslash is redundant between brackets - E701 # multiple statements on one line (colon) - E702 # multiple statements on one line (semicolon) - E703 # statement ends with a semicolon - E711 # comparison to None should be 'if cond is None:' - E714 # test for object identity should be "is not" - E721 # do not compare types, use "isinstance()" - E742 # do not define classes named "l", "O", or "I" - E743 # do not define functions named "l", "O", or "I" - E901 # SyntaxError: invalid syntax - E902 # TokenError: EOF in multi-line string - F401 # module imported but unused - F402 # import module from line N shadowed by loop variable - F403 # 'from foo_module import *' used; unable to detect undefined names - F404 # future import(s) name after other statements - F405 # foo_function may be undefined, or defined from star imports: bar_module - F406 # "from module import *" only allowed at module level - F407 # an undefined __future__ feature name was imported - F601 # dictionary key name repeated with different values - F602 # dictionary key variable name repeated with different values - F621 # too many expressions in an assignment with star-unpacking - F622 # two or more starred expressions in an assignment (a, *b, *c = d) - F631 # assertion test is a tuple, which are always True - F632 # use ==/!= to compare str, bytes, and int literals - F701 # a break statement outside of a while or for loop - F702 # a continue statement outside of a while or for loop - F703 # a continue statement in a finally block in a loop - F704 # a yield or yield from statement outside of a function - F705 # a return statement with arguments inside a generator - F706 # a return statement outside of a function/method - F707 # an except: block as not the last exception handler - F811 # redefinition of unused name from line N - F812 # list comprehension redefines 'foo' from line N - F821 # undefined name 'Foo' - F822 # undefined name name in __all__ - F823 # local variable name … referenced before assignment - F831 # duplicate argument name in function definition - F841 # local variable 'foo' is assigned to but never used - W191 # indentation contains tabs - W291 # trailing whitespace - W292 # no newline at end of file - W293 # blank line contains whitespace - W601 # .has_key() is deprecated, use "in" - W602 # deprecated form of raising exception - W603 # "<>" is deprecated, use "!=" - W604 # backticks are deprecated, use "repr()" - W605 # invalid escape sequence "x" - W606 # 'async' and 'await' are reserved keywords starting with Python 3.7 -) - -if ! command -v flake8 > /dev/null; then - echo "Skipping Python linting since flake8 is not installed." - exit 0 -elif PYTHONWARNINGS="ignore" flake8 --version | grep -q "Python 2"; then - echo "Skipping Python linting since flake8 is running under Python 2. Install the Python 3 version of flake8." - exit 0 -fi - -EXIT_CODE=0 - -# shellcheck disable=SC2046 -if ! PYTHONWARNINGS="ignore" flake8 --ignore=B,C,E,F,I,N,W --select=$(IFS=","; echo "${enabled[*]}") $( - if [[ $# == 0 ]]; then - git ls-files "*.py" - else - echo "$@" - fi -); then - EXIT_CODE=1 -fi - -mapfile -t FILES < <(git ls-files "test/functional/*.py" "contrib/devtools/*.py") -if ! mypy --show-error-codes "${FILES[@]}"; then - EXIT_CODE=1 -fi - -exit $EXIT_CODE diff --git a/test/lint/lint-shell-locale.py b/test/lint/lint-shell-locale.py new file mode 100755 index 0000000000..f3dfe18a95 --- /dev/null +++ b/test/lint/lint-shell-locale.py @@ -0,0 +1,67 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 2018-2022 The Bitcoin Core developers +# Distributed under the MIT software license, see the accompanying +# file COPYING or http://www.opensource.org/licenses/mit-license.php. + +""" +Make sure all shell scripts are: +a.) explicitly opt out of locale dependence using + "export LC_ALL=C" or "export LC_ALL=C.UTF-8", or +b.) explicitly opt in to locale dependence using the annotation below. +""" + +import subprocess +import sys +import re + +OPT_IN_LINE = '# This script is intentionally locale dependent by not setting \"export LC_ALL=C\"' + +OPT_OUT_LINES = [ + 'export LC_ALL=C', + 'export LC_ALL=C.UTF-8', +] + +def get_shell_files_list(): + command = [ + 'git', + 'ls-files', + '--', + '*.sh', + ] + try: + return subprocess.check_output(command, stderr = subprocess.STDOUT).decode('utf-8').splitlines() + except subprocess.CalledProcessError as e: + if e.returncode > 1: # return code is 1 when match is empty + print(e.output.decode('utf-8'), end='') + sys.exit(1) + return [] + +def main(): + exit_code = 0 + shell_files = get_shell_files_list() + for file_path in shell_files: + if re.search('src/(secp256k1|minisketch|univalue)/', file_path): + continue + + with open(file_path, 'r', encoding='utf-8') as file_obj: + contents = file_obj.read() + + if OPT_IN_LINE in contents: + continue + + non_comment_pattern = re.compile(r'^\s*((?!#).+)$', re.MULTILINE) + non_comment_lines = re.findall(non_comment_pattern, contents) + if not non_comment_lines: + continue + + first_non_comment_line = non_comment_lines[0] + if first_non_comment_line not in OPT_OUT_LINES: + print(f'Missing "export LC_ALL=C" (to avoid locale dependence) as first non-comment non-empty line in {file_path}') + exit_code = 1 + + return sys.exit(exit_code) + +if __name__ == '__main__': + main() + diff --git a/test/lint/lint-shell-locale.sh b/test/lint/lint-shell-locale.sh deleted file mode 100755 index 4c6b8a57e6..0000000000 --- a/test/lint/lint-shell-locale.sh +++ /dev/null @@ -1,25 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2018-2020 The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -# -# Make sure all shell scripts: -# a.) explicitly opt out of locale dependence using -# "export LC_ALL=C" or "export LC_ALL=C.UTF-8", or -# b.) explicitly opt in to locale dependence using the annotation below. - -export LC_ALL=C - -EXIT_CODE=0 -for SHELL_SCRIPT in $(git ls-files -- "*.sh" | grep -vE "src/(secp256k1|minisketch|univalue)/"); do - if grep -q "# This script is intentionally locale dependent by not setting \"export LC_ALL=C\"" "${SHELL_SCRIPT}"; then - continue - fi - FIRST_NON_COMMENT_LINE=$(grep -vE '^(#.*)?$' "${SHELL_SCRIPT}" | head -1) - if [[ ${FIRST_NON_COMMENT_LINE} != "export LC_ALL=C" && ${FIRST_NON_COMMENT_LINE} != "export LC_ALL=C.UTF-8" ]]; then - echo "Missing \"export LC_ALL=C\" (to avoid locale dependence) as first non-comment non-empty line in ${SHELL_SCRIPT}" - EXIT_CODE=1 - fi -done -exit ${EXIT_CODE} diff --git a/test/lint/lint-submodule.py b/test/lint/lint-submodule.py new file mode 100755 index 0000000000..89d4c80f55 --- /dev/null +++ b/test/lint/lint-submodule.py @@ -0,0 +1,23 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 2022 The Bitcoin Core developers +# Distributed under the MIT software license, see the accompanying +# file COPYING or http://www.opensource.org/licenses/mit-license.php. + +""" +This script checks for git modules +""" + +import subprocess +import sys + +def main(): + submodules_list = subprocess.check_output(['git', 'submodule', 'status', '--recursive'], + universal_newlines = True, encoding = 'utf8').rstrip('\n') + if submodules_list: + print("These submodules were found, delete them:\n", submodules_list) + sys.exit(1) + sys.exit(0) + +if __name__ == '__main__': + main() diff --git a/test/lint/lint-submodule.sh b/test/lint/lint-submodule.sh deleted file mode 100755 index d9aa021df7..0000000000 --- a/test/lint/lint-submodule.sh +++ /dev/null @@ -1,20 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2020 The Bitcoin Core developers -# Distributed under the MIT software license, see the accompanying -# file COPYING or http://www.opensource.org/licenses/mit-license.php. -# -# This script checks for git modules -export LC_ALL=C -EXIT_CODE=0 - -CMD=$(git submodule status --recursive) -if test -n "$CMD"; -then - echo These submodules were found, delete them: - echo "$CMD" - EXIT_CODE=1 -fi - -exit $EXIT_CODE - diff --git a/test/lint/lint-tests.py b/test/lint/lint-tests.py new file mode 100755 index 0000000000..849ddcb961 --- /dev/null +++ b/test/lint/lint-tests.py @@ -0,0 +1,87 @@ +#!/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 the test suite naming conventions +""" + +import re +import subprocess +import sys + + +def grep_boost_fixture_test_suite(): + command = [ + "git", + "grep", + "-E", + r"^BOOST_FIXTURE_TEST_SUITE\(", + "--", + "src/test/**.cpp", + "src/wallet/test/**.cpp", + ] + return subprocess.check_output(command, universal_newlines=True, encoding="utf8") + + +def check_matching_test_names(test_suite_list): + not_matching = [ + x + for x in test_suite_list + if re.search(r"/(.*?)\.cpp:BOOST_FIXTURE_TEST_SUITE\(\1, .*\)", x) is None + ] + if len(not_matching) > 0: + not_matching = "\n".join(not_matching) + error_msg = ( + "The test suite in file src/test/foo_tests.cpp should be named\n" + '"foo_tests". Please make sure the following test suites follow\n' + "that convention:\n\n" + f"{not_matching}\n" + ) + print(error_msg) + return 1 + return 0 + + +def get_duplicates(input_list): + """ + From https://stackoverflow.com/a/9835819 + """ + seen = set() + dupes = set() + for x in input_list: + if x in seen: + dupes.add(x) + else: + seen.add(x) + return dupes + + +def check_unique_test_names(test_suite_list): + output = [re.search(r"\((.*?),", x) for x in test_suite_list] + output = [x.group(1) for x in output if x is not None] + output = get_duplicates(output) + output = sorted(list(output)) + + if len(output) > 0: + output = "\n".join(output) + error_msg = ( + "Test suite names must be unique. The following test suite names\n" + f"appear to be used more than once:\n\n{output}" + ) + print(error_msg) + return 1 + return 0 + + +def main(): + test_suite_list = grep_boost_fixture_test_suite().splitlines() + exit_code = check_matching_test_names(test_suite_list) + exit_code |= check_unique_test_names(test_suite_list) + sys.exit(exit_code) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-tests.sh b/test/lint/lint-tests.sh deleted file mode 100755 index 35d11023eb..0000000000 --- a/test/lint/lint-tests.sh +++ /dev/null @@ -1,35 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2018 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 the test suite naming conventions - -export LC_ALL=C -EXIT_CODE=0 - -NAMING_INCONSISTENCIES=$(git grep -E '^BOOST_FIXTURE_TEST_SUITE\(' -- \ - "src/test/**.cpp" "src/wallet/test/**.cpp" | \ - grep -vE '/(.*?)\.cpp:BOOST_FIXTURE_TEST_SUITE\(\1, .*\)$') -if [[ ${NAMING_INCONSISTENCIES} != "" ]]; then - echo "The test suite in file src/test/foo_tests.cpp should be named" - echo "\"foo_tests\". Please make sure the following test suites follow" - echo "that convention:" - echo - echo "${NAMING_INCONSISTENCIES}" - EXIT_CODE=1 -fi - -TEST_SUITE_NAME_COLLISSIONS=$(git grep -E '^BOOST_FIXTURE_TEST_SUITE\(' -- \ - "src/test/**.cpp" "src/wallet/test/**.cpp" | cut -f2 -d'(' | cut -f1 -d, | \ - sort | uniq -d) -if [[ ${TEST_SUITE_NAME_COLLISSIONS} != "" ]]; then - echo "Test suite names must be unique. The following test suite names" - echo "appear to be used more than once:" - echo - echo "${TEST_SUITE_NAME_COLLISSIONS}" - EXIT_CODE=1 -fi - -exit ${EXIT_CODE} diff --git a/test/lint/lint-whitespace.py b/test/lint/lint-whitespace.py new file mode 100755 index 0000000000..d98fc8d9a2 --- /dev/null +++ b/test/lint/lint-whitespace.py @@ -0,0 +1,135 @@ +#!/usr/bin/env python3 +# +# Copyright (c) 2017-2022 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 for new lines in diff that introduce trailing whitespace or +# tab characters instead of spaces. + +# We can't run this check unless we know the commit range for the PR. + +import argparse +import os +import re +import sys + +from subprocess import check_output + +EXCLUDED_DIRS = ["depends/patches/", + "contrib/guix/patches/", + "src/leveldb/", + "src/crc32c/", + "src/secp256k1/", + "src/minisketch/", + "src/univalue/", + "doc/release-notes/", + "src/qt/locale"] + +def parse_args(): + """Parse command line arguments.""" + parser = argparse.ArgumentParser( + description=""" + Check for new lines in diff that introduce trailing whitespace + or tab characters instead of spaces in unstaged changes, the + previous n commits, or a commit-range. + """, + epilog=f""" + You can manually set the commit-range with the COMMIT_RANGE + environment variable (e.g. "COMMIT_RANGE='47ba2c3...ee50c9e' + {sys.argv[0]}"). Defaults to current merge base when neither + prev-commits nor the environment variable is set. + """) + + parser.add_argument("--prev-commits", "-p", required=False, help="The previous n commits to check") + + return parser.parse_args() + + +def report_diff(selection): + filename = "" + seen = False + seenln = False + + print("The following changes were suspected:") + + for line in selection: + if re.match(r"^diff", line): + filename = line + seen = False + elif re.match(r"^@@", line): + linenumber = line + seenln = False + else: + if not seen: + # The first time a file is seen with trailing whitespace or a tab character, we print the + # filename (preceded by a newline). + print("") + print(filename) + seen = True + if not seenln: + print(linenumber) + seenln = True + print(line) + + +def get_diff(commit_range, check_only_code): + exclude_args = [":(exclude)" + dir for dir in EXCLUDED_DIRS] + + if check_only_code: + what_files = ["*.cpp", "*.h", "*.md", "*.py", "*.sh"] + else: + what_files = ["."] + + diff = check_output(["git", "diff", "-U0", commit_range, "--"] + what_files + exclude_args, universal_newlines=True, encoding="utf8") + + return diff + + +def main(): + args = parse_args() + + if not os.getenv("COMMIT_RANGE"): + if args.prev_commits: + commit_range = "HEAD~" + args.prev_commits + "...HEAD" + else: + # This assumes that the target branch of the pull request will be master. + merge_base = check_output(["git", "merge-base", "HEAD", "master"], universal_newlines=True, encoding="utf8").rstrip("\n") + commit_range = merge_base + "..HEAD" + else: + commit_range = os.getenv("COMMIT_RANGE") + + whitespace_selection = [] + tab_selection = [] + + # Check if trailing whitespace was found in the diff. + for line in get_diff(commit_range, check_only_code=False).splitlines(): + if re.match(r"^(diff --git|\@@|^\+.*\s+$)", line): + whitespace_selection.append(line) + + whitespace_additions = [i for i in whitespace_selection if i.startswith("+")] + + # Check if tab characters were found in the diff. + for line in get_diff(commit_range, check_only_code=True).splitlines(): + if re.match(r"^(diff --git|\@@|^\+.*\t)", line): + tab_selection.append(line) + + tab_additions = [i for i in tab_selection if i.startswith("+")] + + ret = 0 + + if len(whitespace_additions) > 0: + print("This diff appears to have added new lines with trailing whitespace.") + report_diff(whitespace_selection) + ret = 1 + + if len(tab_additions) > 0: + print("This diff appears to have added new lines with tab characters instead of spaces.") + report_diff(tab_selection) + ret = 1 + + sys.exit(ret) + + +if __name__ == "__main__": + main() diff --git a/test/lint/lint-whitespace.sh b/test/lint/lint-whitespace.sh deleted file mode 100755 index 9d55c71eb5..0000000000 --- a/test/lint/lint-whitespace.sh +++ /dev/null @@ -1,115 +0,0 @@ -#!/usr/bin/env bash -# -# Copyright (c) 2017-2021 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 for new lines in diff that introduce trailing whitespace. - -# We can't run this check unless we know the commit range for the PR. - -export LC_ALL=C -while getopts "?" opt; do - case $opt in - ?) - echo "Usage: $0 [N]" - echo " COMMIT_RANGE='<commit range>' $0" - echo " $0 -?" - echo "Checks unstaged changes, the previous N commits, or a commit range." - echo "COMMIT_RANGE='47ba2c3...ee50c9e' $0" - exit 0 - ;; - esac -done - -if [ -z "${COMMIT_RANGE}" ]; then - if [ -n "$1" ]; then - COMMIT_RANGE="HEAD~$1...HEAD" - else - # This assumes that the target branch of the pull request will be master. - MERGE_BASE=$(git merge-base HEAD master) - COMMIT_RANGE="$MERGE_BASE..HEAD" - fi -fi - -showdiff() { - if ! git diff -U0 "${COMMIT_RANGE}" -- "." ":(exclude)depends/patches/" ":(exclude)contrib/guix/patches/" ":(exclude)src/leveldb/" ":(exclude)src/crc32c/" ":(exclude)src/secp256k1/" ":(exclude)src/minisketch/" ":(exclude)src/univalue/" ":(exclude)doc/release-notes/" ":(exclude)src/qt/locale/"; then - echo "Failed to get a diff" - exit 1 - fi -} - -showcodediff() { - if ! git diff -U0 "${COMMIT_RANGE}" -- *.cpp *.h *.md *.py *.sh ":(exclude)src/leveldb/" ":(exclude)src/crc32c/" ":(exclude)src/secp256k1/" ":(exclude)src/minisketch/" ":(exclude)src/univalue/" ":(exclude)doc/release-notes/" ":(exclude)src/qt/locale/"; then - echo "Failed to get a diff" - exit 1 - fi -} - -RET=0 - -# Check if trailing whitespace was found in the diff. -if showdiff | grep -E -q '^\+.*\s+$'; then - echo "This diff appears to have added new lines with trailing whitespace." - echo "The following changes were suspected:" - FILENAME="" - SEEN=0 - SEENLN=0 - while read -r line; do - if [[ "$line" =~ ^diff ]]; then - FILENAME="$line" - SEEN=0 - elif [[ "$line" =~ ^@@ ]]; then - LINENUMBER="$line" - SEENLN=0 - else - if [ "$SEEN" -eq 0 ]; then - # The first time a file is seen with trailing whitespace, we print the - # filename (preceded by a newline). - echo - echo "$FILENAME" - SEEN=1 - fi - if [ "$SEENLN" -eq 0 ]; then - echo "$LINENUMBER" - SEENLN=1 - fi - echo "$line" - fi - done < <(showdiff | grep -E '^(diff --git |@@|\+.*\s+$)') - RET=1 -fi - -# Check if tab characters were found in the diff. -if showcodediff | perl -nle '$MATCH++ if m{^\+.*\t}; END{exit 1 unless $MATCH>0}' > /dev/null; then - echo "This diff appears to have added new lines with tab characters instead of spaces." - echo "The following changes were suspected:" - FILENAME="" - SEEN=0 - SEENLN=0 - while read -r line; do - if [[ "$line" =~ ^diff ]]; then - FILENAME="$line" - SEEN=0 - elif [[ "$line" =~ ^@@ ]]; then - LINENUMBER="$line" - SEENLN=0 - else - if [ "$SEEN" -eq 0 ]; then - # The first time a file is seen with a tab character, we print the - # filename (preceded by a newline). - echo - echo "$FILENAME" - SEEN=1 - fi - if [ "$SEENLN" -eq 0 ]; then - echo "$LINENUMBER" - SEENLN=1 - fi - echo "$line" - fi - done < <(showcodediff | perl -nle 'print if m{^(diff --git |@@|\+.*\t)}') - RET=1 -fi - -exit $RET |