class Pygemony(object):
"""
The main driver of pygemony, pulls the seperate pieces together.
"""
def __init__(self, user=None, token=None, owner=None, repo=None):
# todo_found contains a list of the following layout:
# ['file_path', 'line_number', 'todo_message', 'md5 of todo']
self.blacklist = ['build', '.git']
self.todo_found = []
self.github = GithubAPIManager(user, token, owner, repo)
# TODO(ian): Add support for parsing more than one file type
self.language = self.lookup_language()
def _sanitize_todo_line(self, lines):
"""
Strips tab, newline, and comment characters form the TODO line.
:param str lines: The found line containing a TODO
:rtype: str
:return: The sanitized TODO line.
"""
# We're mainly aiming to remove newlines and tab characters here.
lines = lines.replace('\n', '')
while ' ' in lines or '\t' in lines:
lines = lines.replace(' ', '')
for lang in self.language:
lines = lines.replace(lang.single_comment, '')
return lines
@staticmethod
def hash_todo(todo_content, file_name):
"""
Hashes the TODO line with the file name
:param str todo_content: The line in the file containing TODO
:param str file_name: The file name containing the TODO line.
:rtype: str
:return: The MD5 hash of the `todo_content` and `file_name`
"""
m = hashlib.md5()
m.update('{0}-{1}'.format(todo_content, file_name))
return str(m.hexdigest())
def parse_for_todo(self, f, file_):
"""
Searches (line-by-line) through a file's content and and looks for
lines containing TODO.
:param file_handle f: The handle to the file that is currently being
searched
:param str file_: The name of the file currently being searched
"""
for i, line in enumerate(f.readlines()):
if "TODO" in line and self._starts_with_comment(line):
line = self._sanitize_todo_line(line)
self.todo_found.append([file_, i, line, self.hash_todo(line, file_)])
def parse_by_extension(self, files):
"""
Parses the list of the directory for files with an acceptable
extension. The extension is determined by data returned from github on
the languages used in the project.
:param list files: The list of all files in the current repository
:rtype: generator(str)
:return: Generates a list of acceptable-to-parse files.
"""
for lang in self.language:
for ext in lang.file_exts:
for file_ in fn_filter(files, ext):
yield file_
def find_all_files(self, root):
"""
Walks the current repository directory and determines viable files
:param str root: The root directory
:rtype: list
:return: The list of files found and determined to be viable.
"""
files_found = []
for roots, _, files in walk(root):
base_dir = roots.split('/')[1]
if base_dir not in self.blacklist:
for file_ in self.parse_by_extension(files):
files_found.append(path.join(roots, file_))
return files_found
def file_handler(self):
"""
Handles IO with the file
:rtype: list
:return: The list of files found
"""
# First we need to remove any non-text files
files_found = self.find_all_files('./')
# TODO(ian): filter() over files to parse out by mimetype
for file_ in files_found:
file_type = detect_mimetype(file_)
# We're looking for startswith('text/'). Mimetype returns
# None if it can't determine file type. Remove if either is True
try:
if file_type[0].startswith("application") or file_type[0] is None:
files_found.remove(file_)
except (AttributeError, IndexError) as e:
print "Failed to open file {} with error of {}".format(file_, e)
for file_ in files_found:
try:
with open(file_, 'r') as f:
self.parse_for_todo(f, file_)
except IOError as e:
print "Failed to open file {} with error of {}".format(file_, e)
return files_found
def run(self):
"""
Starts the process of finding TODOs
"""
self.file_handler()
self.github.commit(self.todo_found)
def lookup_language(self):
"""
Constructs langauge classes based on what is found in github data.
:rtype: list
:return: A list of language classes that will be found in a github
repo.
"""
lang_map = {'cpp': LanguageCPP,
'python': LanguagePython,
'javascript': LanguageJavascript,
'c': LanguageC,
'go': LanguageGo,
'erlang': LanguageErlang}
langs = [i for i in self.github.get_languages()]
for i in langs:
self.blacklist.append(lang_map[str(langs[0][0]).lower()]().ignore_dir)
return [lang_map[str(langs[0][0]).lower()]()]
def _starts_with_comment(self, line):
"""
Verifies a line (containing the word TODO) starts with a comment, if it
does, we deem it to be commit-viable.
:param str line: The line that contains "TODO"
:rtype: bool
:return: True if line starts with a comment (is a valid TODO statement)
"""
comments = self._create_comment_start_list()
for comment in comments:
if line.startswith(comment):
return True
def _create_comment_start_list(self):
"""
Create a list of comments from each language class associated with the
current repo.
:rtype: list
:return: A list of strings containing all line-start comments.
"""
comments = []
for lang in self.language:
comments.append(lang.single_comment)
comments.append(lang.multi_comment[0])
return comments
class Pygemony(object):
def __init__(self, user=None, token=None, owner=None, repo=None):
# todo_found contains a list of the following layout:
# ['file_path', 'line_number', 'todo_message', 'md5 of todo']
self.blacklist = ['build', '.git']
self.todo_found = []
self.github = GithubAPIManager(user, token, owner, repo)
# TODO(ian): Add support for parsing more than one file type
self.language = self.lookup_language()
def find_end_comment(self, f):
# TODO(ian): Remove this function as we no longer support multiline TODO
todo_content = []
x = f
count = 0
for line in x.readlines():
todo_content.append(line)
if self.language.multi_comment[1] in line:
return todo_content
if count > 20:
return None
todo_content.append(line)
count += 1
def _sanitize_todo_line(self, lines):
# We're mainly aiming to remove newlines and tab characters here.
lines = lines.replace('\n', '')
while ' ' in lines or '\t' in lines:
lines = lines.replace(' ', '')
for lang in self.language:
lines = lines.replace(lang.single_comment, '')
return lines
@staticmethod
def hash_todo(todo_content, file_name):
m = hashlib.md5()
m.update('{0}-{1}'.format(todo_content, file_name))
return str(m.hexdigest())
def parse_for_todo(self, f, file_):
for i, line in enumerate(f.readlines()):
if "TODO" in line and self._starts_with_comment(line):
line = self._sanitize_todo_line(line)
self.todo_found.append([file_, i, line, self.hash_todo(line, file_)])
def parse_by_extension(self, files):
for lang in self.language:
for ext in lang.file_exts:
for file_ in fn_filter(files, ext):
yield file_
def find_all_files(self, root):
files_found = []
for roots, _, files in walk(root):
base_dir = roots.split('/')[1]
if base_dir not in self.blacklist:
for file_ in self.parse_by_extension(files):
files_found.append(path.join(roots, file_))
return files_found
def file_handler(self):
# First we need to remove any non-text files
files_found = self.find_all_files('./')
# TODO(ian): filter() over files to parse out by mimetype
for file_ in files_found:
file_type = detect_mimetype(file_)
# We're looking for startswith('text/'). Mimetype returns
# None if it can't determine file type. Remove if either is True
try:
if file_type[0].startswith("application") or file_type[0] is None:
files_found.remove(file_)
except (AttributeError, IndexError) as e:
print "Failed to open file {} with error of {}".format(file_, e)
for file_ in files_found:
try:
with open(file_, 'r') as f:
self.parse_for_todo(f, file_)
except IOError as e:
print "Failed to open file {} with error of {}".format(file_, e)
return files_found
def run(self):
self.file_handler()
self.github.commit(self.todo_found)
def lookup_language(self):
lang_map = {'cpp': LanguageCPP,
'python': LanguagePython,
'javascript': LanguageJavascript,
'c': LanguageC,
'go': LanguageGo}
langs = [i for i in self.github.get_languages()]
for i in langs:
self.blacklist.append(lang_map[str(langs[0][0]).lower()]().ignore_dir)
return [lang_map[str(langs[0][0]).lower()]()]
def _starts_with_comment(self, line):
comments = self._create_comment_start_list()
for comment in comments:
if line.startswith(comment):
return True
def _create_comment_start_list(self):
comments = []
for lang in self.language:
comments.append(lang.single_comment)
comments.append(lang.multi_comment[0])
return comments