ClickHouse/ci/praktika/hook_html.py

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

230 lines
8.0 KiB
Python
Raw Normal View History

2024-10-24 11:17:00 +00:00
import dataclasses
import json
2024-10-01 19:19:35 +00:00
from pathlib import Path
2024-10-24 11:17:00 +00:00
from typing import List
2024-10-01 19:19:35 +00:00
from praktika._environment import _Environment
from praktika.gh import GH
from praktika.parser import WorkflowConfigParser
2024-11-08 16:48:17 +00:00
from praktika.result import Result, ResultInfo, _ResultS3
2024-10-01 19:19:35 +00:00
from praktika.runtime import RunConfig
from praktika.s3 import S3
from praktika.settings import Settings
2024-10-29 21:09:03 +00:00
from praktika.utils import Utils
2024-10-24 11:17:00 +00:00
@dataclasses.dataclass
class GitCommit:
2024-10-29 21:09:03 +00:00
# date: str
# message: str
2024-10-24 11:17:00 +00:00
sha: str
@staticmethod
2024-10-29 21:09:03 +00:00
def from_json(file) -> List["GitCommit"]:
2024-10-24 11:17:00 +00:00
commits = []
2024-10-29 21:09:03 +00:00
json_data = None
2024-10-24 11:17:00 +00:00
try:
2024-10-29 21:09:03 +00:00
with open(file, "r", encoding="utf-8") as f:
json_data = json.load(f)
2024-10-24 11:17:00 +00:00
commits = [
GitCommit(
2024-10-29 21:09:03 +00:00
# message=commit["messageHeadline"],
sha=commit["sha"],
# date=commit["committedDate"],
2024-10-24 11:17:00 +00:00
)
2024-10-29 21:09:03 +00:00
for commit in json_data
2024-10-24 11:17:00 +00:00
]
except Exception as e:
print(
2024-10-29 21:09:03 +00:00
f"ERROR: Failed to deserialize commit's data [{json_data}], ex: [{e}]"
2024-10-24 11:17:00 +00:00
)
return commits
2024-10-01 19:19:35 +00:00
2024-10-29 21:09:03 +00:00
@classmethod
def update_s3_data(cls):
env = _Environment.get()
sha = env.SHA
if not sha:
print("WARNING: Failed to retrieve commit sha")
return
commits = cls.pull_from_s3()
for commit in commits:
if sha == commit.sha:
print(
f"INFO: Sha already present in commits data [{sha}] - skip data update"
)
return
commits.append(GitCommit(sha=sha))
cls.push_to_s3(commits)
return
2024-10-01 19:19:35 +00:00
@classmethod
2024-10-29 21:09:03 +00:00
def dump(cls, commits):
commits_ = []
for commit in commits:
commits_.append(dataclasses.asdict(commit))
with open(cls.file_name(), "w", encoding="utf8") as f:
json.dump(commits_, f)
@classmethod
def pull_from_s3(cls):
local_path = Path(cls.file_name())
file_name = local_path.name
env = _Environment.get()
s3_path = f"{Settings.HTML_S3_PATH}/{cls.get_s3_prefix(pr_number=env.PR_NUMBER, branch=env.BRANCH)}/{file_name}"
if not S3.copy_file_from_s3(s3_path=s3_path, local_path=local_path):
print(f"WARNING: failed to cp file [{s3_path}] from s3")
return []
return cls.from_json(local_path)
@classmethod
def push_to_s3(cls, commits):
print(f"INFO: push commits data to s3, commits num [{len(commits)}]")
cls.dump(commits)
local_path = Path(cls.file_name())
file_name = local_path.name
env = _Environment.get()
s3_path = f"{Settings.HTML_S3_PATH}/{cls.get_s3_prefix(pr_number=env.PR_NUMBER, branch=env.BRANCH)}/{file_name}"
if not S3.copy_file_to_s3(s3_path=s3_path, local_path=local_path, text=True):
print(f"WARNING: failed to cp file [{local_path}] to s3")
@classmethod
def get_s3_prefix(cls, pr_number, branch):
prefix = ""
assert pr_number or branch
if pr_number and pr_number > 0:
prefix += f"{pr_number}"
else:
prefix += f"{branch}"
return prefix
@classmethod
def file_name(cls):
return f"{Settings.TEMP_DIR}/commits.json"
2024-10-24 11:17:00 +00:00
2024-10-29 21:09:03 +00:00
# def _get_pr_commits(pr_number):
# res = []
# if not pr_number:
# return res
# output = Shell.get_output(f"gh pr view {pr_number} --json commits")
# if output:
# res = GitCommit.from_json(output)
# return res
2024-10-24 11:17:00 +00:00
2024-10-29 21:09:03 +00:00
class HtmlRunnerHooks:
@classmethod
def configure(cls, _workflow):
2024-10-01 19:19:35 +00:00
# generate pending Results for all jobs in the workflow
if _workflow.enable_cache:
skip_jobs = RunConfig.from_fs(_workflow.name).cache_success
2024-11-08 16:48:17 +00:00
job_cache_records = RunConfig.from_fs(_workflow.name).cache_jobs
2024-10-01 19:19:35 +00:00
else:
skip_jobs = []
env = _Environment.get()
results = []
for job in _workflow.jobs:
if job.name not in skip_jobs:
result = Result.generate_pending(job.name)
else:
2024-11-08 16:48:17 +00:00
result = Result.generate_skipped(job.name, job_cache_records[job.name])
2024-10-01 19:19:35 +00:00
results.append(result)
summary_result = Result.generate_pending(_workflow.name, results=results)
2024-11-05 12:59:14 +00:00
summary_result.links.append(env.CHANGE_URL)
summary_result.links.append(env.RUN_URL)
2024-10-01 19:19:35 +00:00
summary_result.start_time = Utils.timestamp()
2024-11-08 16:48:17 +00:00
assert _ResultS3.copy_result_to_s3_with_version(summary_result, version=0)
2024-11-05 12:59:14 +00:00
page_url = env.get_report_url(settings=Settings)
2024-10-01 19:19:35 +00:00
print(f"CI Status page url [{page_url}]")
res1 = GH.post_commit_status(
name=_workflow.name,
status=Result.Status.PENDING,
description="",
2024-11-08 16:48:17 +00:00
url=env.get_report_url(settings=Settings, latest=True),
2024-10-01 19:19:35 +00:00
)
res2 = GH.post_pr_comment(
comment_body=f"Workflow [[{_workflow.name}]({page_url})], commit [{_Environment.get().SHA[:8]}]",
or_update_comment_with_substring=f"Workflow [",
)
if not (res1 or res2):
2024-10-24 11:17:00 +00:00
Utils.raise_with_error(
"Failed to set both GH commit status and PR comment with Workflow Status, cannot proceed"
2024-10-01 19:19:35 +00:00
)
2024-10-24 11:17:00 +00:00
if env.PR_NUMBER:
2024-10-29 21:09:03 +00:00
# TODO: enable for branch, add commit number limiting
GitCommit.update_s3_data()
2024-10-01 19:19:35 +00:00
@classmethod
def pre_run(cls, _workflow, _job):
result = Result.from_fs(_job.name)
2024-11-08 16:48:17 +00:00
_ResultS3.update_workflow_results(
workflow_name=_workflow.name, new_sub_results=result
2024-10-01 19:19:35 +00:00
)
@classmethod
def run(cls, _workflow, _job):
pass
@classmethod
def post_run(cls, _workflow, _job, info_errors):
result = Result.from_fs(_job.name)
2024-11-08 16:48:17 +00:00
_ResultS3.upload_result_files_to_s3(result)
_ResultS3.copy_result_to_s3(result)
2024-10-01 19:19:35 +00:00
env = _Environment.get()
2024-11-08 16:48:17 +00:00
new_sub_results = [result]
new_result_info = ""
2024-10-01 19:19:35 +00:00
env_info = env.REPORT_INFO
if env_info:
print(
f"WARNING: some info lines are set in Environment - append to report [{env_info}]"
)
info_errors += env_info
if info_errors:
info_errors = [f" | {error}" for error in info_errors]
info_str = f"{_job.name}:\n"
info_str += "\n".join(info_errors)
print("Update workflow results with new info")
2024-11-08 16:48:17 +00:00
new_result_info = info_str
2024-10-01 19:19:35 +00:00
if not result.is_ok():
print(
"Current job failed - find dependee jobs in the workflow and set their statuses to skipped"
)
workflow_config_parsed = WorkflowConfigParser(_workflow).parse()
for dependee_job in workflow_config_parsed.workflow_yaml_config.jobs:
if _job.name in dependee_job.needs:
if _workflow.get_job(dependee_job.name).run_unless_cancelled:
continue
print(
f"NOTE: Set job [{dependee_job.name}] status to [{Result.Status.SKIPPED}] due to current failure"
)
2024-11-08 16:48:17 +00:00
new_sub_results.append(
2024-10-01 19:19:35 +00:00
Result(
name=dependee_job.name,
status=Result.Status.SKIPPED,
info=ResultInfo.SKIPPED_DUE_TO_PREVIOUS_FAILURE
+ f" [{_job.name}]",
)
)
2024-11-08 16:48:17 +00:00
updated_status = _ResultS3.update_workflow_results(
new_info=new_result_info,
new_sub_results=new_sub_results,
workflow_name=_workflow.name,
2024-10-01 19:19:35 +00:00
)
2024-11-08 16:48:17 +00:00
if updated_status:
print(f"Update GH commit status [{result.name}]: [{updated_status}]")
2024-10-01 19:19:35 +00:00
GH.post_commit_status(
2024-11-08 16:48:17 +00:00
name=_workflow.name,
status=GH.convert_to_gh_status(updated_status),
2024-10-01 19:19:35 +00:00
description="",
2024-11-08 16:48:17 +00:00
url=env.get_report_url(settings=Settings, latest=True),
2024-10-01 19:19:35 +00:00
)