390 lines
13 KiB
Python
Executable File
390 lines
13 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
# Wireshark - Network traffic analyzer
|
|
# By Gerald Combs <gerald@wireshark.org>
|
|
# Copyright 1998 Gerald Combs
|
|
#
|
|
# SPDX-License-Identifier: GPL-2.0-or-later
|
|
|
|
import os
|
|
import re
|
|
import subprocess
|
|
import argparse
|
|
import signal
|
|
|
|
# This utility scans for tfs items, and works out if standard ones
|
|
# could have been used intead (from epan/tfs.c)
|
|
|
|
# TODO:
|
|
# - check how many of the definitions in epan/tfs.c are used in other dissectors
|
|
# - although even if unused, might be in external dissectors?
|
|
|
|
|
|
# Try to exit soon after Ctrl-C is pressed.
|
|
should_exit = False
|
|
|
|
def signal_handler(sig, frame):
|
|
global should_exit
|
|
should_exit = True
|
|
print('You pressed Ctrl+C - exiting')
|
|
|
|
signal.signal(signal.SIGINT, signal_handler)
|
|
|
|
|
|
# Test for whether the given file was automatically generated.
|
|
def isGeneratedFile(filename):
|
|
# Open file
|
|
f_read = open(os.path.join(filename), 'r')
|
|
lines_tested = 0
|
|
for line in f_read:
|
|
# The comment to say that its generated is near the top, so give up once
|
|
# get a few lines down.
|
|
if lines_tested > 10:
|
|
f_read.close()
|
|
return False
|
|
if (line.find('Generated automatically') != -1 or
|
|
line.find('Generated Automatically') != -1 or
|
|
line.find('Autogenerated from') != -1 or
|
|
line.find('is autogenerated') != -1 or
|
|
line.find('automatically generated by Pidl') != -1 or
|
|
line.find('Created by: The Qt Meta Object Compiler') != -1 or
|
|
line.find('This file was generated') != -1 or
|
|
line.find('This filter was automatically generated') != -1 or
|
|
line.find('This file is auto generated, do not edit!') != -1 or
|
|
line.find('This file is auto generated') != -1):
|
|
|
|
f_read.close()
|
|
return True
|
|
lines_tested = lines_tested + 1
|
|
|
|
# OK, looks like a hand-written file!
|
|
f_read.close()
|
|
return False
|
|
|
|
|
|
# Keep track of custom entries that might appear in multiple dissectors,
|
|
# so we can consider adding them to tfs.c
|
|
custom_tfs_entries = {}
|
|
def AddCustomEntry(val1, val2, file):
|
|
global custom_tfs_entries
|
|
if (val1, val2) in custom_tfs_entries:
|
|
custom_tfs_entries[(val1, val2)].append(file)
|
|
else:
|
|
custom_tfs_entries[(val1, val2)] = [file]
|
|
|
|
|
|
|
|
class TFS:
|
|
def __init__(self, file, name, val1, val2):
|
|
self.file = file
|
|
self.name = name
|
|
self.val1 = val1
|
|
self.val2 = val2
|
|
|
|
# Do some extra checks on values.
|
|
if val1.startswith(' ') or val1.endswith(' '):
|
|
print('N.B.: file=' + self.file + ' ' + self.name + ' - false val begins or ends with space \"' + self.val1 + '\"')
|
|
if val2.startswith(' ') or val2.endswith(' '):
|
|
print('N.B.: file=' + self.file + ' ' + self.name + ' - true val begins or ends with space \"' + self.val2 + '\"')
|
|
|
|
def __str__(self):
|
|
return '{' + '"' + self.val1 + '", "' + self.val2 + '"}'
|
|
|
|
|
|
class ValueString:
|
|
def __init__(self, file, name, vals):
|
|
self.file = file
|
|
self.name = name
|
|
self.raw_vals = vals
|
|
self.parsed_vals = {}
|
|
self.looks_like_tfs = True
|
|
|
|
no_lines = self.raw_vals.count('{')
|
|
if no_lines != 3:
|
|
self.looks_like_tfs = False
|
|
return
|
|
|
|
# Now parse out each entry in the value_string
|
|
matches = re.finditer(r'\{([\"a-zA-Z\s\d\,]*)\}', self.raw_vals)
|
|
for m in matches:
|
|
entry = m[1]
|
|
# Check each entry looks like part of a TFS entry.
|
|
match = re.match(r'\s*([01])\,\s*\"([a-zA-Z\d\s]*\s*)\"', entry)
|
|
if match:
|
|
if match[1] == '1':
|
|
self.parsed_vals[True] = match[2]
|
|
else:
|
|
self.parsed_vals[False] = match[2]
|
|
|
|
# Now have both entries
|
|
if len(self.parsed_vals) == 2:
|
|
break
|
|
else:
|
|
self.looks_like_tfs = False
|
|
break
|
|
|
|
def __str__(self):
|
|
return '{' + '"' + self.raw_vals + '"}'
|
|
|
|
|
|
|
|
def removeComments(code_string):
|
|
code_string = re.sub(re.compile(r"/\*.*?\*/",re.DOTALL ) ,"" ,code_string) # C-style comment
|
|
code_string = re.sub(re.compile(r"//.*?\n" ) ,"" ,code_string) # C++-style comment
|
|
return code_string
|
|
|
|
|
|
# Look for true_false_string items in a dissector file.
|
|
def findItems(filename):
|
|
items = {}
|
|
|
|
with open(filename, 'r') as f:
|
|
contents = f.read()
|
|
# Example: const true_false_string tfs_true_false = { "True", "False" };
|
|
|
|
# Remove comments so as not to trip up RE.
|
|
contents = removeComments(contents)
|
|
|
|
matches = re.finditer(r'.*const\s*true_false_string\s*([a-z_]*)\s*=\s*{\s*\"([a-zA-Z_ ]*)\"\s*,\s*\"([a-zA-Z_ ]*)\"', contents)
|
|
for m in matches:
|
|
name = m.group(1)
|
|
val1 = m.group(2)
|
|
val2 = m.group(3)
|
|
# Store this entry.
|
|
items[name] = TFS(filename, name, val1, val2)
|
|
|
|
return items
|
|
|
|
# Look for value_string entries in a dissector file.
|
|
def findValueStrings(filename):
|
|
items = {}
|
|
|
|
#static const value_string radio_type_vals[] =
|
|
#{
|
|
# { 0, "FDD"},
|
|
# { 1, "TDD"},
|
|
# { 0, NULL }
|
|
#};
|
|
|
|
with open(filename, 'r') as f:
|
|
contents = f.read()
|
|
|
|
# Remove comments so as not to trip up RE.
|
|
contents = removeComments(contents)
|
|
|
|
matches = re.finditer(r'.*const value_string\s*([a-zA-Z0-9_]*)\s*\[\s*\]\s*\=\s*\{([\{\}\d\,a-zA-Z\s\"]*)\};', contents)
|
|
for m in matches:
|
|
name = m.group(1)
|
|
vals = m.group(2)
|
|
items[name] = ValueString(filename, name, vals)
|
|
|
|
return items
|
|
|
|
|
|
|
|
def is_dissector_file(filename):
|
|
p = re.compile(r'.*packet-.*\.c')
|
|
return p.match(filename)
|
|
|
|
def findDissectorFilesInFolder(folder):
|
|
# Look at files in sorted order, to give some idea of how far through is.
|
|
files = []
|
|
|
|
for f in sorted(os.listdir(folder)):
|
|
if should_exit:
|
|
return
|
|
if is_dissector_file(f):
|
|
filename = os.path.join(folder, f)
|
|
files.append(filename)
|
|
return files
|
|
|
|
|
|
|
|
warnings_found = 0
|
|
errors_found = 0
|
|
|
|
# Check the given dissector file.
|
|
def checkFile(filename, tfs_items, look_for_common=False, check_value_strings=False):
|
|
global warnings_found
|
|
global errors_found
|
|
|
|
# Check file exists - e.g. may have been deleted in a recent commit.
|
|
if not os.path.exists(filename):
|
|
print(filename, 'does not exist!')
|
|
return
|
|
|
|
# Find items.
|
|
items = findItems(filename)
|
|
|
|
# See if any of these items already existed in tfs.c
|
|
for i in items:
|
|
for t in tfs_items:
|
|
found = False
|
|
|
|
#
|
|
# Do not do this check for plugins; plugins cannot import
|
|
# data values from libwireshark (functions, yes; data
|
|
# values, no).
|
|
#
|
|
# Test whether there's a common prefix for the file name
|
|
# and "plugin/epan/"; if so, this is a plugin, and there
|
|
# is no common path and os.path.commonprefix returns an
|
|
# empty string, otherwise it returns the common path, so
|
|
# we check whether the common path is an empty string.
|
|
#
|
|
if os.path.commonprefix([filename, 'plugin/epan/']) == '':
|
|
exact_case = False
|
|
if tfs_items[t].val1 == items[i].val1 and tfs_items[t].val2 == items[i].val2:
|
|
found = True
|
|
exact_case = True
|
|
elif tfs_items[t].val1.upper() == items[i].val1.upper() and tfs_items[t].val2.upper() == items[i].val2.upper():
|
|
found = True
|
|
|
|
if found:
|
|
print("Error:" if exact_case else "Warn: ", filename, i, "- could have used", t, 'from tfs.c instead: ', tfs_items[t],
|
|
'' if exact_case else ' (capitalisation differs)')
|
|
if exact_case:
|
|
errors_found += 1
|
|
else:
|
|
warnings_found += 1
|
|
break
|
|
if not found:
|
|
if look_for_common:
|
|
AddCustomEntry(items[i].val1, items[i].val2, filename)
|
|
|
|
if check_value_strings:
|
|
vs = findValueStrings(filename)
|
|
for v in vs:
|
|
if vs[v].looks_like_tfs:
|
|
found = False
|
|
exact_case = False
|
|
|
|
#print('Candidate', v, vs[v])
|
|
for t in tfs_items:
|
|
found = False
|
|
|
|
#
|
|
# Do not do this check for plugins; plugins cannot import
|
|
# data values from libwireshark (functions, yes; data
|
|
# values, no).
|
|
#
|
|
# Test whether there's a common prefix for the file name
|
|
# and "plugin/epan/"; if so, this is a plugin, and there
|
|
# is no common path and os.path.commonprefix returns an
|
|
# empty string, otherwise it returns the common path, so
|
|
# we check whether the common path is an empty string.
|
|
#
|
|
if os.path.commonprefix([filename, 'plugin/epan/']) == '':
|
|
exact_case = False
|
|
if tfs_items[t].val1 == vs[v].parsed_vals[True] and tfs_items[t].val2 == vs[v].parsed_vals[False]:
|
|
found = True
|
|
exact_case = True
|
|
elif tfs_items[t].val1.upper() == vs[v].parsed_vals[True].upper() and tfs_items[t].val2.upper() == vs[v].parsed_vals[False].upper():
|
|
found = True
|
|
|
|
if found:
|
|
print("Warn:" if exact_case else "Note:", filename, 'value_string', v, "- could have used", t, 'from tfs.c instead: ', tfs_items[t],
|
|
'' if exact_case else ' (capitalisation differs)')
|
|
if exact_case:
|
|
warnings_found += 1
|
|
break
|
|
|
|
|
|
|
|
#################################################################
|
|
# Main logic.
|
|
|
|
# command-line args. Controls which dissector files should be checked.
|
|
# If no args given, will just scan epan/dissectors folder.
|
|
parser = argparse.ArgumentParser(description='Check calls in dissectors')
|
|
parser.add_argument('--file', action='append',
|
|
help='specify individual dissector file to test')
|
|
parser.add_argument('--commits', action='store',
|
|
help='last N commits to check')
|
|
parser.add_argument('--open', action='store_true',
|
|
help='check open files')
|
|
parser.add_argument('--check-value-strings', action='store_true',
|
|
help='check whether value_strings could have been tfs?')
|
|
|
|
parser.add_argument('--common', action='store_true',
|
|
help='check for potential new entries for tfs.c')
|
|
|
|
|
|
args = parser.parse_args()
|
|
|
|
|
|
# Get files from wherever command-line args indicate.
|
|
files = []
|
|
if args.file:
|
|
# Add specified file(s)
|
|
for f in args.file:
|
|
if not f.startswith('epan'):
|
|
f = os.path.join('epan', 'dissectors', f)
|
|
if not os.path.isfile(f):
|
|
print('Chosen file', f, 'does not exist.')
|
|
exit(1)
|
|
else:
|
|
files.append(f)
|
|
elif args.commits:
|
|
# Get files affected by specified number of commits.
|
|
command = ['git', 'diff', '--name-only', 'HEAD~' + args.commits]
|
|
files = [f.decode('utf-8')
|
|
for f in subprocess.check_output(command).splitlines()]
|
|
# Will examine dissector files only
|
|
files = list(filter(lambda f : is_dissector_file(f), files))
|
|
elif args.open:
|
|
# Unstaged changes.
|
|
command = ['git', 'diff', '--name-only']
|
|
files = [f.decode('utf-8')
|
|
for f in subprocess.check_output(command).splitlines()]
|
|
# Only interested in dissector files.
|
|
files = list(filter(lambda f : is_dissector_file(f), files))
|
|
# Staged changes.
|
|
command = ['git', 'diff', '--staged', '--name-only']
|
|
files_staged = [f.decode('utf-8')
|
|
for f in subprocess.check_output(command).splitlines()]
|
|
# Only interested in dissector files.
|
|
files_staged = list(filter(lambda f : is_dissector_file(f), files_staged))
|
|
for f in files_staged:
|
|
if not f in files:
|
|
files.append(f)
|
|
else:
|
|
# Find all dissector files from folder.
|
|
files = findDissectorFilesInFolder(os.path.join('epan', 'dissectors'))
|
|
|
|
|
|
# If scanning a subset of files, list them here.
|
|
print('Examining:')
|
|
if args.file or args.commits or args.open:
|
|
if files:
|
|
print(' '.join(files), '\n')
|
|
else:
|
|
print('No files to check.\n')
|
|
else:
|
|
print('All dissector modules\n')
|
|
|
|
|
|
# Get standard/ shared ones.
|
|
tfs_entries = findItems(os.path.join('epan', 'tfs.c'))
|
|
|
|
# Now check the files to see if they could have used shared ones instead.
|
|
for f in files:
|
|
if should_exit:
|
|
exit(1)
|
|
if not isGeneratedFile(f):
|
|
checkFile(f, tfs_entries, look_for_common=args.common, check_value_strings=args.check_value_strings)
|
|
|
|
|
|
# Show summary.
|
|
print(warnings_found, 'warnings found')
|
|
if errors_found:
|
|
print(errors_found, 'errors found')
|
|
exit(1)
|
|
|
|
if args.common:
|
|
# Looking for items that could potentially be moved to tfs.c
|
|
for c in custom_tfs_entries:
|
|
# Only want to see items that have 3 or more occurrences.
|
|
# Even then, probably only want to consider ones that sound generic.
|
|
if len(custom_tfs_entries[c]) > 2:
|
|
print(c, 'appears', len(custom_tfs_entries[c]), 'times, in: ', custom_tfs_entries[c])
|