\n"""
# escape() and unescape() takes care of &, < and >.
html_escape_table = {
'"': """,
"'": "'"
}
html_unescape_table = {v: k for k, v in html_escape_table.items()}
def html_escape(text):
return escape(text, html_escape_table)
def git_blame(line, path, file, blame_options):
git_blame_dict = {}
head, tail = os.path.split(file)
if head != "":
path = head
try:
os.chdir(path)
except:
return {}
try:
result = subprocess.check_output('git blame -L %d %s %s --porcelain -- %s' % (
line, " -w" if "-w" in blame_options else "", " -M" if "-M" in blame_options else "", file))
result = result.decode(locale.getpreferredencoding())
except:
return {}
if result.startswith('fatal'):
return {}
disallowed_characters = '<>'
for line in result.split('\n')[1:]:
space_pos = line.find(' ')
if space_pos > 30:
break
key = line[:space_pos]
val = line[space_pos + 1:]
for character in disallowed_characters:
val = val.replace(character, "")
git_blame_dict[key] = val
datetime_object = datetime.date.fromtimestamp(float(git_blame_dict['author-time']))
year = datetime_object.strftime("%Y")
month = datetime_object.strftime("%m")
day = datetime_object.strftime("%d")
git_blame_dict['author-time'] = '%s/%s/%s' % (day, month, year)
return git_blame_dict
def tr_str(td_th, line, id, cwe, severity, message, author, author_mail, date, add_author, tr_class=None, htmlfile=None, message_class=None):
ret = ''
if htmlfile:
ret += '<%s>%d' % (td_th, htmlfile, line, line, td_th)
for item in (id, cwe, severity):
ret += '<%s>%s' % (td_th, item, td_th)
else:
for item in (line, id, cwe, severity):
ret += '<%s>%s' % (td_th, item, td_th)
if message_class:
message_attribute = ' class="%s"' % message_class
else:
message_attribute = ''
ret += '<%s%s>%s' % (td_th, message_attribute, html_escape(message), td_th)
if add_author:
for item in (author, author_mail, date):
ret += '<%s>%s' % (td_th, item, td_th)
if tr_class:
tr_attributes = ' class="%s"' % tr_class
else:
tr_attributes = ''
return '%s' % (tr_attributes, ret)
class AnnotateCodeFormatter(HtmlFormatter):
errors = []
def wrap(self, source, outfile):
line_no = 1
for i, t in HtmlFormatter.wrap(self, source, outfile):
# If this is a source code line we want to add a span tag at the
# end.
if i == 1:
for error in self.errors:
if error['line'] == line_no:
try:
if error['inconclusive'] == 'true':
# only print verbose msg if it really differs
# from actual message
if error.get('verbose') and (error['verbose'] != error['msg']):
index = t.rfind('\n')
t = t[:index] + HTML_EXPANDABLE_INCONCLUSIVE % (error['msg'], html_escape(error['verbose'].replace("\\012", '\n'))) + t[index + 1:]
else:
t = t.replace('\n', HTML_INCONCLUSIVE % error['msg'])
except KeyError:
if error.get('verbose') and (error['verbose'] != error['msg']):
index = t.rfind('\n')
t = t[:index] + HTML_EXPANDABLE_ERROR % (error['msg'], html_escape(error['verbose'].replace("\\012", '\n'))) + t[index + 1:]
else:
t = t.replace('\n', HTML_ERROR % error['msg'])
line_no = line_no + 1
yield i, t
class CppCheckHandler(XmlContentHandler):
"""Parses the cppcheck xml file and produces a list of all its errors."""
def __init__(self):
XmlContentHandler.__init__(self)
self.errors = []
self.version = '1'
self.versionCppcheck = ''
def startElement(self, name, attributes):
if name == 'results':
self.version = attributes.get('version', self.version)
if self.version == '1':
self.handleVersion1(name, attributes)
else:
self.handleVersion2(name, attributes)
def handleVersion1(self, name, attributes):
if name != 'error':
return
self.errors.append({
'file': attributes.get('file', ''),
'line': int(attributes.get('line', 0)),
'locations': [{
'file': attributes.get('file', ''),
'line': int(attributes.get('line', 0)),
}],
'id': attributes['id'],
'severity': attributes['severity'],
'msg': attributes['msg']
})
def handleVersion2(self, name, attributes):
if name == 'cppcheck':
self.versionCppcheck = attributes['version']
if name == 'error':
error = {
'locations': [],
'file': '',
'line': 0,
'id': attributes['id'],
'severity': attributes['severity'],
'msg': attributes['msg'],
'verbose': attributes.get('verbose')
}
if 'inconclusive' in attributes:
error['inconclusive'] = attributes['inconclusive']
if 'cwe' in attributes:
error['cwe'] = attributes['cwe']
self.errors.append(error)
elif name == 'location':
assert self.errors
error = self.errors[-1]
locations = error['locations']
file = attributes['file']
line = int(attributes['line'])
if not locations:
error['file'] = file
error['line'] = line
locations.append({
'file': file,
'line': line,
'info': attributes.get('info')
})
if __name__ == '__main__':
# Configure all the options this little utility is using.
parser = optparse.OptionParser()
parser.add_option('--title', dest='title',
help='The title of the project.',
default='[project name]')
parser.add_option('--file', dest='file', action="append",
help='The cppcheck xml output file to read defects '
'from. You can combine results from several '
'xml reports i.e. "--file file1.xml --file file2.xml ..". '
'Default is reading from stdin.')
parser.add_option('--report-dir', dest='report_dir',
help='The directory where the HTML report content is '
'written.')
parser.add_option('--source-dir', dest='source_dir',
help='Base directory where source code files can be '
'found.')
parser.add_option('--add-author-information', dest='add_author_information',
help='Initially set to false'
'Adds author, author-mail and time to htmlreport')
parser.add_option('--source-encoding', dest='source_encoding',
help='Encoding of source code.', default='utf-8')
parser.add_option('--blame-options', dest='blame_options',
help='[-w, -M] blame options which you can use to get author and author mail '
'-w --> not including white spaces and returns original author of the line '
'-M --> not including moving of lines and returns original author of the line')
# Parse options and make sure that we have an output directory set.
options, args = parser.parse_args()
try:
sys.argv[1]
except IndexError: # no arguments give, print --help
parser.print_help()
quit()
if not options.report_dir:
parser.error('No report directory set.')
# Get the directory where source code files are located.
cwd = os.getcwd()
source_dir = os.getcwd()
if options.source_dir:
source_dir = options.source_dir
add_author_information = False
if options.add_author_information:
add_author_information = True
blame_options = ''
if options.blame_options:
blame_options = options.blame_options
add_author_information = True
# Parse the xml from all files defined in file argument
# or from stdin. If no input is provided, stdin is used
# Produce a simple list of errors.
print('Parsing xml report.')
try:
contentHandler = CppCheckHandler()
for fname in options.file or [sys.stdin]:
xml_parse(fname, contentHandler)
except (XmlParseException, ValueError) as msg:
print('Failed to parse cppcheck xml file: %s' % msg)
sys.exit(1)
# We have a list of errors. But now we want to group them on
# each source code file. Lets create a files dictionary that
# will contain a list of all the errors in that file. For each
# file we will also generate a HTML filename to use.
files = {}
file_no = 0
for error in contentHandler.errors:
filename = error['file']
if filename not in files.keys():
files[filename] = {
'errors': [], 'htmlfile': str(file_no) + '.html'}
file_no = file_no + 1
files[filename]['errors'].append(error)
# Make sure that the report directory is created if it doesn't exist.
print('Creating %s directory' % options.report_dir)
if not os.path.exists(options.report_dir):
os.makedirs(options.report_dir)
# Generate a HTML file with syntax highlighted source code for each
# file that contains one or more errors.
print('Processing errors')
decode_errors = []
for filename, data in sorted(files.items()):
htmlfile = data['htmlfile']
errors = []
for error in data['errors']:
for location in error['locations']:
if filename == location['file']:
newError = dict(error)
del newError['locations']
newError['line'] = location['line']
if location.get('info'):
newError['msg'] = location['info']
newError['severity'] = 'information'
del newError['verbose']
errors.append(newError)
lines = []
for error in errors:
lines.append(error['line'])
if filename == '':
continue
source_filename = os.path.join(source_dir, filename)
try:
with io.open(source_filename, 'r', encoding=options.source_encoding) as input_file:
content = input_file.read()
except IOError:
if error['id'] == 'unmatchedSuppression':
continue # file not found, bail out
else:
sys.stderr.write("ERROR: Source file '%s' not found.\n" %
source_filename)
continue
except UnicodeDecodeError:
sys.stderr.write("WARNING: Unicode decode error in '%s'.\n" %
source_filename)
decode_errors.append(source_filename[2:]) # "[2:]" gets rid of "./" at beginning
continue
htmlFormatter = AnnotateCodeFormatter(linenos=True,
style='colorful',
hl_lines=lines,
lineanchors='line',
encoding=options.source_encoding)
htmlFormatter.errors = errors
with io.open(os.path.join(options.report_dir, htmlfile), 'w', encoding='utf-8') as output_file:
output_file.write(HTML_HEAD %
(options.title,
htmlFormatter.get_style_defs('.highlight'),
options.title,
filename,
filename.split('/')[-1]))
for error in sorted(errors, key=lambda k: k['line']):
output_file.write(" %s %s" % (data['htmlfile'], error['line'], error['id'], error['line']))
output_file.write(HTML_HEAD_END)
try:
lexer = guess_lexer_for_filename(source_filename, '', stripnl=False)
except ClassNotFound:
try:
lexer = guess_lexer(content, stripnl=False)
except ClassNotFound:
sys.stderr.write("ERROR: Couldn't determine lexer for the file' " + source_filename + " '. Won't be able to syntax highlight this file.")
output_file.write("\n
Could not generate content because pygments failed to determine the code type.
")
output_file.write("\n
Sorry about this.
")
continue
if options.source_encoding:
lexer.encoding = options.source_encoding
output_file.write(
highlight(content, lexer, htmlFormatter).decode(
options.source_encoding))
output_file.write(HTML_FOOTER % contentHandler.versionCppcheck)
print(' ' + filename)
# Generate a master index.html file that will contain a list of
# all the errors created.
print('Creating index.html')
with io.open(os.path.join(options.report_dir, 'index.html'),
'w') as output_file:
stats_count = 0
stats = []
for filename, data in sorted(files.items()):
for error in data['errors']:
stats.append(error['id']) # get the stats
stats_count += 1
counter = Counter(stats)
stat_html = []
# the following lines sort the stat primary by value (occurrences),
# but if two IDs occur equally often, then we sort them alphabetically by warning ID
try:
cnt_max = counter.most_common()[0][1]
except IndexError:
cnt_max = 0
try:
cnt_min = counter.most_common()[-1][1]
except IndexError:
cnt_min = 0
stat_fmt = "\n
{}
{}
"
for occurrences in reversed(range(cnt_min, cnt_max + 1)):
for _id in [k for k, v in sorted(counter.items()) if v == occurrences]:
stat_html.append(stat_fmt.format(_id, _id, dict(counter.most_common())[_id], _id))
output_file.write(HTML_HEAD.replace('id="menu"', 'id="menu_index"', 1).replace("Defects:", "Defect summary;", 1) % (options.title, '', options.title, '', ''))
output_file.write('\n ')
output_file.write('\n
')
output_file.write(
'\n %s' %
tr_str('th', 'Line', 'Id', 'CWE', 'Severity', 'Message', 'Author', 'Author mail', 'Date (DD/MM/YYYY)', add_author=add_author_information))
for filename, data in sorted(files.items()):
if filename in decode_errors: # don't print a link but a note
output_file.write("\n
%s
" % filename)
output_file.write("\n
Could not generated due to UnicodeDecodeError
")
else:
if filename.endswith('*'): # assume unmatched suppression
output_file.write(
"\n
" %
(data['htmlfile'], filename))
for error in sorted(data['errors'], key=lambda k: k['line']):
if add_author_information:
git_blame_dict = git_blame(error['line'], source_dir, error['file'], blame_options)
else:
git_blame_dict = {}
message_class = None
try:
if error['inconclusive'] == 'true':
message_class = 'inconclusive'
error['severity'] += ", inconcl."
except KeyError:
pass
try:
if error['cwe']:
cwe_url = "" + error['cwe'] + ""
except KeyError:
cwe_url = ""
if error['severity'] == 'error':
message_class = 'error'
is_file = filename != '' and not filename.endswith('*')
line = error["line"] if is_file else ""
htmlfile = data.get('htmlfile') if is_file else None
output_file.write(
'\n %s' %
tr_str('td', line, error["id"], cwe_url, error["severity"], error["msg"],
git_blame_dict.get('author', 'Unknown'), git_blame_dict.get('author-mail', '---'),
git_blame_dict.get('author-time', '---'),
tr_class=error["id"],
message_class=message_class,
add_author=add_author_information,
htmlfile=htmlfile))
output_file.write('\n
')
output_file.write(HTML_FOOTER % contentHandler.versionCppcheck)
if decode_errors:
sys.stderr.write("\nGenerating html failed for the following files: " + ' '.join(decode_errors))
sys.stderr.write("\nConsider changing source-encoding (for example: \"htmlreport ... --source-encoding=\"iso8859-1\"\"\n")
print('Creating style.css file')
os.chdir(cwd) # going back to the cwd to find style.css
with io.open(os.path.join(options.report_dir, 'style.css'), 'w') as css_file:
css_file.write(STYLE_FILE)
print("Creating stats.html (statistics)\n")
stats_countlist = {}
for filename, data in sorted(files.items()):
if filename == '':
continue
stats_tmplist = []
for error in sorted(data['errors'], key=lambda k: k['line']):
stats_tmplist.append(error['severity'])
stats_countlist[filename] = dict(Counter(stats_tmplist))
# get top ten for each severity
SEVERITIES = "error", "warning", "portability", "performance", "style", "unusedFunction", "information", "missingInclude", "internal"
with io.open(os.path.join(options.report_dir, 'stats.html'), 'w') as stats_file:
stats_file.write(HTML_HEAD.replace('id="menu"', 'id="menu_index"', 1).replace("Defects:", "Back to summary", 1) % (options.title, '', options.title, 'Statistics', ''))
stats_file.write(HTML_HEAD_END.replace("content", "content_index", 1))
for sev in SEVERITIES:
_sum = 0
stats_templist = {}
# if the we have an style warning but we are checking for
# portability, we have to skip it to prevent KeyError
try:
for filename in stats_countlist:
try: # also bail out if we have a file with no sev-results
_sum += stats_countlist[filename][sev]
stats_templist[filename] = int(stats_countlist[filename][sev]) # file : amount,
except KeyError:
continue
# don't print "0 style" etc, if no style warnings were found
if _sum == 0:
continue
except KeyError:
continue
stats_file.write("
Top 10 files for " + sev + " severity, total findings: " + str(_sum) + " \n")
# sort, so that the file with the most severities per type is first
stats_list_sorted = sorted(stats_templist.items(), key=operator.itemgetter(1, 0), reverse=True)
it = 0
LENGTH = 0
for i in stats_list_sorted: # printing loop
# for aesthetics: if it's the first iteration of the loop, get
# the max length of the number string
if it == 0:
LENGTH = len(str(i[1])) # <- length of longest number, now get the difference and try to make other numbers align to it
stats_file.write(" " * 3 + str(i[1]) + " " * (1 + LENGTH - len(str(i[1]))) + " " + i[0] + " \n")
it += 1
if it == 10: # print only the top 10
break
stats_file.write("
\n")
stats_file.write(HTML_FOOTER % contentHandler.versionCppcheck)
print("\nOpen '" + options.report_dir + "/index.html' to see the results.")