esp-idf/tools/ci/python_packages/gitlab_api.py

213 lines
8.0 KiB
Python
Raw Normal View History

import argparse
2019-11-26 22:22:14 -05:00
import os
import re
import tarfile
import tempfile
2019-11-26 22:22:14 -05:00
import zipfile
from functools import wraps
2019-11-26 22:22:14 -05:00
import gitlab
class Gitlab(object):
JOB_NAME_PATTERN = re.compile(r'(\w+)(\s+(\d+)/(\d+))?')
2019-11-26 22:22:14 -05:00
DOWNLOAD_ERROR_MAX_RETRIES = 3
2019-11-26 22:22:14 -05:00
def __init__(self, project_id=None):
config_data_from_env = os.getenv('PYTHON_GITLAB_CONFIG')
2019-11-26 22:22:14 -05:00
if config_data_from_env:
# prefer to load config from env variable
with tempfile.NamedTemporaryFile('w', delete=False) as temp_file:
2019-11-26 22:22:14 -05:00
temp_file.write(config_data_from_env)
config_files = [temp_file.name]
else:
# otherwise try to use config file at local filesystem
config_files = None
gitlab_id = os.getenv('LOCAL_GITLAB_HTTPS_HOST') # if None, will use the default gitlab server
self.gitlab_inst = gitlab.Gitlab.from_config(gitlab_id=gitlab_id, config_files=config_files)
2019-11-26 22:22:14 -05:00
self.gitlab_inst.auth()
if project_id:
self.project = self.gitlab_inst.projects.get(project_id)
else:
self.project = None
def get_project_id(self, name, namespace=None):
"""
search project ID by name
:param name: project name
:param namespace: namespace to match when we have multiple project with same name
:return: project ID
"""
projects = self.gitlab_inst.projects.list(search=name)
res = []
2019-11-26 22:22:14 -05:00
for project in projects:
if namespace is None:
if len(projects) == 1:
res.append(project.id)
2019-11-26 22:22:14 -05:00
break
if project.namespace['path'] == namespace:
if project.name == name:
res.insert(0, project.id)
else:
res.append(project.id)
if not res:
2019-11-26 22:22:14 -05:00
raise ValueError("Can't find project")
return res[0]
2019-11-26 22:22:14 -05:00
def download_artifacts(self, job_id, destination):
"""
download full job artifacts and extract to destination.
:param job_id: Gitlab CI job ID
:param destination: extract artifacts to path.
"""
job = self.project.jobs.get(job_id)
with tempfile.NamedTemporaryFile(delete=False) as temp_file:
job.artifacts(streamed=True, action=temp_file.write)
with zipfile.ZipFile(temp_file.name, 'r') as archive_file:
2019-11-26 22:22:14 -05:00
archive_file.extractall(destination)
def retry_download(func):
"""
This wrapper will only catch IOError and retry the whole function.
So only use it with download functions, read() inside and atomic
functions
"""
@wraps(func)
def wrapper(self, *args, **kwargs):
retried = 0
while True:
try:
res = func(self, *args, **kwargs)
except (IOError, EOFError) as e:
retried += 1
if retried > self.DOWNLOAD_ERROR_MAX_RETRIES:
raise e # get out of the loop
else:
print('Retried for the {} time'.format(retried))
continue
else:
break
return res
return wrapper
2021-06-08 03:04:06 -04:00
@retry_download # type: ignore
2019-11-26 22:22:14 -05:00
def download_artifact(self, job_id, artifact_path, destination=None):
"""
download specific path of job artifacts and extract to destination.
:param job_id: Gitlab CI job ID
:param artifact_path: list of path in artifacts (relative path to artifact root path)
:param destination: destination of artifact. Do not save to file if destination is None
:return: A list of artifact file raw data.
"""
job = self.project.jobs.get(job_id)
raw_data_list = []
for a_path in artifact_path:
try:
data = job.artifact(a_path)
except gitlab.GitlabGetError as e:
print("Failed to download '{}' from job {}".format(a_path, job_id))
2019-11-26 22:22:14 -05:00
raise e
raw_data_list.append(data)
if destination:
file_path = os.path.join(destination, a_path)
try:
os.makedirs(os.path.dirname(file_path))
except OSError:
# already exists
pass
with open(file_path, 'wb') as f:
2019-11-26 22:22:14 -05:00
f.write(data)
return raw_data_list
def find_job_id(self, job_name, pipeline_id=None, job_status='success'):
2019-11-26 22:22:14 -05:00
"""
Get Job ID from job name of specific pipeline
:param job_name: job name
:param pipeline_id: If None, will get pipeline id from CI pre-defined variable.
:param job_status: status of job. One pipeline could have multiple jobs with same name after retry.
job_status is used to filter these jobs.
2019-11-26 22:22:14 -05:00
:return: a list of job IDs (parallel job will generate multiple jobs)
"""
job_id_list = []
if pipeline_id is None:
pipeline_id = os.getenv('CI_PIPELINE_ID')
2019-11-26 22:22:14 -05:00
pipeline = self.project.pipelines.get(pipeline_id)
jobs = pipeline.jobs.list(all=True)
for job in jobs:
match = self.JOB_NAME_PATTERN.match(job.name)
if match:
if match.group(1) == job_name and job.status == job_status:
job_id_list.append({'id': job.id, 'parallel_num': match.group(3)})
2019-11-26 22:22:14 -05:00
return job_id_list
2021-06-08 03:04:06 -04:00
@retry_download # type: ignore
2019-11-26 22:22:14 -05:00
def download_archive(self, ref, destination, project_id=None):
"""
Download archive of certain commit of a repository and extract to destination path
:param ref: commit or branch name
:param destination: destination path of extracted archive file
:param project_id: download project of current instance if project_id is None
:return: root path name of archive file
"""
if project_id is None:
project = self.project
else:
project = self.gitlab_inst.projects.get(project_id)
with tempfile.NamedTemporaryFile(delete=False) as temp_file:
try:
project.repository_archive(sha=ref, streamed=True, action=temp_file.write)
except gitlab.GitlabGetError as e:
print('Failed to archive from project {}'.format(project_id))
2019-11-26 22:22:14 -05:00
raise e
print('archive size: {:.03f}MB'.format(float(os.path.getsize(temp_file.name)) / (1024 * 1024)))
2019-11-26 22:22:14 -05:00
with tarfile.open(temp_file.name, 'r') as archive_file:
2019-11-26 22:22:14 -05:00
root_name = archive_file.getnames()[0]
archive_file.extractall(destination)
return os.path.join(os.path.realpath(destination), root_name)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('action')
parser.add_argument('project_id', type=int)
parser.add_argument('--pipeline_id', '-i', type=int, default=None)
parser.add_argument('--ref', '-r', default='master')
parser.add_argument('--job_id', '-j', type=int, default=None)
parser.add_argument('--job_name', '-n', default=None)
parser.add_argument('--project_name', '-m', default=None)
parser.add_argument('--destination', '-d', default=None)
parser.add_argument('--artifact_path', '-a', nargs='*', default=None)
2019-11-26 22:22:14 -05:00
args = parser.parse_args()
gitlab_inst = Gitlab(args.project_id)
if args.action == 'download_artifacts':
2019-11-26 22:22:14 -05:00
gitlab_inst.download_artifacts(args.job_id, args.destination)
if args.action == 'download_artifact':
2019-11-26 22:22:14 -05:00
gitlab_inst.download_artifact(args.job_id, args.artifact_path, args.destination)
elif args.action == 'find_job_id':
2019-11-26 22:22:14 -05:00
job_ids = gitlab_inst.find_job_id(args.job_name, args.pipeline_id)
print(';'.join([','.join([str(j['id']), j['parallel_num']]) for j in job_ids]))
elif args.action == 'download_archive':
2019-11-26 22:22:14 -05:00
gitlab_inst.download_archive(args.ref, args.destination)
elif args.action == 'get_project_id':
2019-11-26 22:22:14 -05:00
ret = gitlab_inst.get_project_id(args.project_name)
print('project id: {}'.format(ret))