def _identify_gh_repo(homepage):
"""Returns code repository dict filled with homepage, if homepage is GH repo
(None otherwise)
"""
if parse_gh_repo(homepage):
return {'url': homepage, 'type': 'git'}
return None
def get_files_github_url(self, github_url):
"""Clone the repository from GitHub and retrieve manifest files from it."""
manifest_data = []
repo_suffix = parse_gh_repo(github_url)
try:
self.del_temp_files()
repo_url = urljoin(self.PREFIX_URL, repo_suffix)
check_valid_repo = get(repo_url)
if check_valid_repo.status_code == 200:
repo_clone_url = urljoin(self.PREFIX_GIT_URL, repo_suffix,
'.git')
Git.clone(repo_clone_url, self.CLONED_DIR)
for file_obj in self.get_manifest_files():
file_content = None
filename = file_obj.get('filename')
filepath = file_obj.get('filepath')
with open(filepath, 'rb') as m_file:
file_content = m_file.read().decode('utf-8')
manifest_data.append({
"filename":
filename,
"content":
file_content,
"filepath":
filepath.replace(self.CLONED_DIR, '')
})
except Exception:
raise HTTPError(500, "Error in reading repo from github.")
finally:
self.del_temp_files()
return manifest_data
def _get_repo_name(self, url):
"""Retrieve GitHub repo from a preceding Mercator scan."""
parsed = parse_gh_repo(url)
if not parsed:
self.log.debug('Could not parse Github repo URL %s', url)
else:
self._repo_url = 'https://github.com/' + parsed
return parsed
def _get_repo_name(self, url):
"""Get GitHub repo URL."""
parsed = parse_gh_repo(url)
if not parsed:
logger.debug('Could not parse Github repo URL %s', url)
else:
self._repo_url = 'https://github.com/' + parsed
return parsed
def isGhRepo(node_args, key):
"""Predicate if the repository is on GitHub."""
try:
val = reduce(lambda m, k: m[k], key if isinstance(key, list) else [key], node_args)
if parse_gh_repo(val):
return True
else:
return False
except Exception:
return False
def isGhRepo(node_args, key):
try:
val = reduce(lambda m, k: m[k],
key if isinstance(key, list) else [key], node_args)
if parse_gh_repo(val):
return True
else:
return False
except Exception:
return False
def _handle_java(self, data):
"""Handle data from pom.xml."""
# we expect pom.xml to be there, since it's always downloaded to top level by InitTask
pom = data.get('pom.xml')
if pom is None:
return None
key_map = (('name', ), ('version', ), ('description', ),
('url', 'homepage'), ('licenses', 'declared_licenses'))
# handle licenses
transformed = self.transform_keys(pom, key_map)
if transformed['name'] is None:
transformed['name'] = "{}:{}".format(pom.get('groupId'),
pom.get('artifactId'))
# dependencies with scope 'compile' and 'runtime' are needed at runtime;
# dependencies with scope 'provided' are not necessarily runtime dependencies,
# but they are commonly used for example in web applications
dependencies_dict = pom.get('dependencies', {}).get('compile', {})
dependencies_dict.update(
pom.get('dependencies', {}).get('runtime', {}))
dependencies_dict.update(
pom.get('dependencies', {}).get('provided', {}))
# dependencies with scope 'test' are only needed for testing;
dev_dependencies_dict = pom.get('dependencies', {}).get('test', {})
transformed['dependencies'] = [
k.rstrip(':') + ' ' + v for k, v in dependencies_dict.items()
]
transformed['devel_dependencies'] = [
k.rstrip(':') + ' ' + v for k, v in dev_dependencies_dict.items()
]
# handle code_repository
if 'scm_url' in pom:
# TODO: there's no way we can tell 100 % what the type is, but we could
# try to handle at least some cases, e.g. github will always be git etc
repo_type = 'git' if parse_gh_repo(pom['scm_url']) else 'unknown'
transformed['code_repository'] = {
'url': pom['scm_url'],
'type': repo_type
}
return transformed
def normalize(self):
"""Normalize output from Mercator for pom.xml (Maven)."""
if not self._raw_data:
return {}
if self._data['name'] is None:
self._data['name'] = "{}:{}".format(
self._raw_data.get('groupId'),
self._raw_data.get('artifactId'))
# dependencies with scope 'compile' and 'runtime' are needed at runtime;
# dependencies with scope 'provided' are not necessarily runtime dependencies,
# but they are commonly used for example in web applications
dependencies_dict = self._raw_data.get('dependencies',
{}).get('compile', {})
dependencies_dict.update(
self._raw_data.get('dependencies', {}).get('runtime', {}))
dependencies_dict.update(
self._raw_data.get('dependencies', {}).get('provided', {}))
# dependencies with scope 'test' are only needed for testing;
dev_dependencies_dict = self._raw_data.get('dependencies',
{}).get('test', {})
self._data['dependencies'] = [
k.rstrip(':') + ' ' + v for k, v in dependencies_dict.items()
]
self._data['devel_dependencies'] = [
k.rstrip(':') + ' ' + v for k, v in dev_dependencies_dict.items()
]
# handle code_repository
if 'scm_url' in self._raw_data:
# TODO: there's no way we can tell 100 % what the type is, but we could
# try to handle at least some cases, e.g. github will always be git etc
repo_type = 'git' if parse_gh_repo(
self._raw_data['scm_url']) else 'unknown'
self._data['code_repository'] = {
'url': self._raw_data['scm_url'],
'type': repo_type
}
return self._data
def _get_github_readme(self, url):
repo_tuple = parse_gh_repo(url)
if repo_tuple:
project, repo = repo_tuple.split('/')
else:
return None
for readme_type, extensions in self.README_TYPES.items():
for extension in extensions:
if extension:
extension = '.' + extension
url = self._GITHUB_README_PATH.format(project=project, repo=repo,
extension=extension)
response = requests.get(url)
if response.status_code != 200:
self.log.debug('No README%s found for type "%s" at "%s"', extension,
readme_type, url)
continue
self.log.debug('README%s found for type "%s" at "%s"', extension, readme_type, url)
return {'type': readme_type, 'content': response.text}
def _identify_gh_repo(homepage):
"""Return code repository dict filled with homepage."""
if parse_gh_repo(homepage):
return {'url': homepage, 'type': 'git'}
return None
def get_manifest_details(self, github_url):
"""Retrieve manifest files from cloned repository."""
manifest_data = []
supported_manifests = {
'requirements.txt': True,
'pom.xml': True,
'package.json': True
}
repo_tuple = parse_gh_repo(github_url)
if repo_tuple:
project, repo = repo_tuple.split('/')
else:
return None
last_commit_url = 'https://api.github.com/repos/{project}/{repo}/git/refs/heads/' \
'master'.format(project=project, repo=repo)
trees_url = 'https://api.github.com/repos/{project}/{repo}/git/trees/{sha}?recursive=1'
raw_content_path = 'https://raw.githubusercontent.com/{project}/{repo}/master/{filename}'
# Fetch the latest commit of the repo
try:
resp = requests.get(last_commit_url)
except exceptions.RequestException as e:
print(e)
return None
last_commit = ''
if resp.status_code == 200:
try:
last_commit = resp.json()['object']['sha']
except KeyError as e:
print(e)
return None
# Fetch the contents tree using the last commit sha
try:
resp = requests.get(trees_url.format(project=project, repo=repo, sha=last_commit))
except exceptions.RequestException as e:
print(e)
return None
if resp.status_code == 200:
try:
tree = resp.json()['tree']
except KeyError as e:
print(e)
return None
for t in tree:
try:
if supported_manifests[os.path.basename(t['path'])]:
manifest_data.append({
'filename': os.path.basename(t['path']),
'download_url': raw_content_path.format(
project=project, repo=repo, filename=t['path']),
'filepath': os.path.dirname(t['path'])
})
except KeyError as e:
print(e)
continue
print(manifest_data)
return manifest_data
def test_parse_gh_repo_nok(self, url):
"""Test parse_gh_repo()."""
assert parse_gh_repo(url) is None
def test_parse_gh_repo_ok(self, url):
"""Test parse_gh_repo()."""
assert parse_gh_repo(url) == 'foo/bar'
def test_parse_gh_repo_nok(self, url):
assert parse_gh_repo(url) is None
def test_parse_gh_repo_ok(self, url):
assert parse_gh_repo(url) == 'foo/bar'
