diff options
author | Neels Hofmeyr <neels@hofmeyr.de> | 2018-07-16 19:41:27 +0200 |
---|---|---|
committer | Neels Hofmeyr <neels@hofmeyr.de> | 2018-07-16 19:53:29 +0200 |
commit | 7eec2dab31f7058b0d97f7bfd7d243a02f02358f (patch) | |
tree | 80a53074b8b23d49661b3c7118b060acebe7b754 | |
parent | 1bfa6c7b627bd9c1f6e44cb385a02ca3b0535f58 (diff) |
add scripts/verify_log_statements.py
This came up in
https://gerrit.osmocom.org/#/c/osmo-bsc/+/9671/6//COMMIT_MSG@36
The errors it finds in the current code base are numerous, and many are
intended LOGP .. LOGPC calls. It doesn't make sense to enforce this, but so far
this can be used manually.
Change-Id: Id79389f090a2fded7ff01dc7e3fe9774e7f22ca0
-rwxr-xr-x | scripts/verify_log_statements.py | 87 |
1 files changed, 87 insertions, 0 deletions
diff --git a/scripts/verify_log_statements.py b/scripts/verify_log_statements.py new file mode 100755 index 0000000..e7752e1 --- /dev/null +++ b/scripts/verify_log_statements.py @@ -0,0 +1,87 @@ +#!/usr/bin/env python3 +__doc__ = ''' +With regex magic, try to pinpoint all LOG* macro calls that lack a final newline. +Also find those that have non-printable characters or extra newlines. + +Usage: + + ./verify_log_statements.py [-d|--debug] [dir] [file] [...] + +Without args, default to '.' +''' + +import re +import sys +import codecs +import os.path + +# This regex matches the entire LOGxx(...) statement over multiple lines. +# It pinpoints the format string by looking for the first arg that contains quotes. +# It then matches any number of separate quoted strings, and accepts 0 or more args after that. +log_statement_re = re.compile(r'^[ \t]*LOG[_A-Z]+\(([^";,]*,)* *(("[^"]*"[^";,]*)*)(,[^;]*|)\);', + re.MULTILINE | re.DOTALL) +fmt_re = re.compile(r'("[^"]*".*)*fmt') + +errors_found = 0 +debug = ('-d' in sys.argv) or ('--debug' in sys.argv) + +args = [x for x in sys.argv[1:] if not (x == '-d' or x == '--debug')] +if not args: + args = ['.'] + + +def check_file(f): + global errors_found + if not (f.endswith('.h') or f.endswith('.c') or f.endswith('.cpp')): + return + + for log in log_statement_re.finditer(codecs.open(f, "r", "utf-8").read()): + quoted = log.group(2) + + # Skip 'LOG("bla" fmt )' strings that typically appear as #defines. + if fmt_re.match(quoted): + if debug: + print('Skipping define:', f, '\n'+log.group(0)) + continue + + # Drop PRI* parts of 'LOG("bla %"PRIu64" foo")' + for n in (16,32,64): + quoted = quoted.replace('PRIu' + str(n), '') + quoted = quoted.replace('PRId' + str(n), '') + + # Use py eval to join separate string constants: drop any tabs/newlines + # that are not in quotes, between separate string constants. + try: + quoted = eval('(' + quoted + '\n)' ) + except: + # hopefully eval broke because of some '## args' macro def + if debug: + print('Ignoring:', f, '\n'+log.group(0)) + continue + + # check for errors... + + # final newline + if not quoted.endswith('\n'): + print('Missing final newline:', f, '\n'+log.group(0)) + errors_found += 1 + + # disallowed chars and extra newlines + for c in quoted[:-1]: + if not c.isprintable() and not c == '\t': + if c == '\n': + msg = 'Extraneous newline' + else: + msg = 'Illegal char' + print('%s %r in' % (msg, c), f, '\n' + log.group(0)) + errors_found += 1 + +for f in args: + if os.path.isdir(f): + for parent_path, subdirs, files in os.walk(f, None, None): + for ff in files: + check_file(os.path.join(parent_path, ff)) + else: + check_file(f) + +sys.exit(errors_found) |