def test_create_adviser_dataframe_with_results(self) -> None:
"""Test create of adviser dataframe from adviser documents."""
adviser_files = Adviser.aggregate_adviser_results(
repo_path=self._ADVISER_FOLDER_PATH, is_local=True)
adviser_version = "0.21.1"
justifications_collected = []
justifications_collected = Adviser.create_adviser_dataframe(
adviser_version=adviser_version,
adviser_files=adviser_files,
justifications_collected=justifications_collected,
)
adviser_dataframe = pd.DataFrame(justifications_collected)
assert adviser_dataframe.shape[0] == 3
def test_create_adviser_dataframe_heatmap(self) -> None:
"""Test create of adviser dataframe for heatmap plot from adviser documents."""
adviser_files = Adviser.aggregate_adviser_results(
repo_path=self._ADVISER_FOLDER_PATH, is_local=True)
adviser_version = "0.21.1"
justifications_collected = []
justifications_collected = Adviser.create_adviser_dataframe(
adviser_version=adviser_version,
adviser_files=adviser_files,
justifications_collected=justifications_collected,
)
adviser_dataframe = pd.DataFrame(justifications_collected)
adviser_heatmap_df = Adviser.create_adviser_results_dataframe_heatmap(
adviser_type_dataframe=adviser_dataframe,
number_days=1,
)
last_date = [column for column in adviser_heatmap_df.columns][-1]
csv = adviser_heatmap_df[[last_date]].to_csv(header=False)
assert csv
async def on_thamos_workflow_finished(*, action, base_repo_url, check_run_id,
installation, payload, **kwargs):
"""Advise workflow has finished, now we need to send a check-run to the PR."""
_LOGGER.info("on_thamos_workflow_finished: %s", kwargs)
github_api: RawGitHubAPI = RUNTIME_CONTEXT.app_installation_client
_LOGGER.info("on_thamos_workflow_finished: github_api=%s", github_api)
repo = base_repo_url.split("/", 4)[-1] # i.e.: thoth-station/Qeb-Hwt
check_runs_url = f"https://api.github.com/repos/{repo}/check-runs/{check_run_id}"
_LOGGER.info("on_thamos_workflow_finished: check_runs_url=%s",
check_runs_url)
advise_url: str
conclusion: str
justification: str
report: str
text: str
report_message: str
async with aiohttp.ClientSession() as session:
_LOGGER.info("on_thamos_workflow_finished: payload=%s", payload)
if "exception" in payload:
exception = payload["exception"]
else:
exception = None
analysis_id = payload["analysis_id"]
_LOGGER.info("on_thamos_workflow_finished: analysis_id=%s",
analysis_id)
advise_url = urljoin(ADVISE_API_URL, analysis_id)
_LOGGER.info("on_thamos_workflow_finished: advise_url=%s", advise_url)
if exception:
_LOGGER.info("on_thamos_workflow_finished: exception=%s",
exception)
if "error_type" in payload:
error_type: str = payload["error_type"]
if error_type and error_type == "MissingThothYamlFile":
conclusion = "action_required"
else:
conclusion = "failure"
else:
conclusion = "failure"
justification = exception
if exception == "Internal server error occurred, please contact administrator with provided details.":
justification += (
"\nThoth Team is working to solve the issue as soon as possible. Thanks for your patience!"
)
report = "Report not produced."
text = report
report_message = ""
if analysis_id:
# TODO: Find alternative solution to this workround
attempts = 1
max_attempts = 6
while attempts < max_attempts:
try:
async with session.get(advise_url) as response:
_LOGGER.info(
"on_thamos_workflow_finished: response=%s",
response)
_LOGGER.info(
"on_thamos_workflow_finished: attempts=%s",
attempts)
if response.status == 200:
attempts = max_attempts
else:
attempts += 1
except Exception:
continue
async with session.get(advise_url) as response:
if response.status != 200:
conclusion = "failure"
justification = "Could not retrieve analysis results."
report = ""
text = "Report cannot be provided, Please open an issue on Qeb-Hwt."
report_message = ""
else:
adviser_payload: dict = await response.json()
adviser_result: dict = adviser_payload["result"]
if adviser_result["error"]:
error_msg: str = adviser_result["error_msg"]
conclusion = "failure"
justification = f"Analysis has encountered errors: {error_msg}."
if error_msg == "No direct dependencies found":
conclusion = "neutral"
if adviser_result["report"]:
report = adviser_result["report"]
text = "See the report below for more details."
report_message = "See the document below for more details."
else:
conclusion = "failure"
justification = f"Analysis has encountered errors: {error_msg}."
if adviser_result["report"]:
report = adviser_result["report"]
text = "See the report below for more details."
report_message = "See the document below for more details."
else:
text = "Analysis report is missing."
report_message = "See the document below for more details."
else:
conclusion = "success"
adviser_report: dict = adviser_result["report"]
justification = Adviser.create_pretty_report_from_json(
report=adviser_report, is_justification=True)
# Complete report
report = Adviser.create_pretty_report_from_json(
report=adviser_report)
_LOGGER.info(
"on_thamos_workflow_finished: len(report)=%s",
len(report))
# TODO: Split report results to include only relevant information
if len(report) > MAX_CHARACTERS_LENGTH:
_LOGGER.warning(
"on_thamos_workflow_finished: reduced len(report)=%s",
len(report))
text = f"Analysis report:\n{report}"
report_message = "See the document below for more details."
else:
analysis_id = "No-analysis-run"
_LOGGER.info(
"on_thamos_workflow_finished: sending check run: check_runs_url=%s",
check_runs_url)
_LOGGER.info(
"on_thamos_workflow_finished: sending check run: conclusion=%s",
conclusion)
_LOGGER.info(
"on_thamos_workflow_finished: sending check run: advise_url=%s",
advise_url)
_LOGGER.info(
"on_thamos_workflow_finished: sending check run: analysis_id=%s",
analysis_id)
_LOGGER.info("on_thamos_workflow_finished: sending check run: text=%s",
text)
_LOGGER.info("on_thamos_workflow_finished: sending check run: text=%s",
report_message)
try:
_LOGGER.info(
"on_thamos_workflow_finished: installation_id=%s, check_run_url=%s",
installation, check_runs_url)
await github_api.patch(
check_runs_url,
preview_api_version="antiope",
data={
"name": CHECK_RUN_NAME,
"status": "completed",
"conclusion": conclusion,
"completed_at": f"{datetime.utcnow().isoformat()}Z",
"details_url": advise_url,
"external_id": analysis_id,
"output": {
"title":
"Thoth's Advise",
"text":
text,
"summary":
(f"Thoth's adviser finished with conclusion: '{conclusion}'\n\n"
f"Justification:\n{justification}\n\n"
f"{report_message}"),
},
},
)
except gidgethub.BadRequest as exc:
_LOGGER.error(exc)
_LOGGER.info(
f"on_thamos_workflow_finished: finished with `thamos advise`, updated %s",
check_run_id)
def explore_adviser_files(
current_initial_date: datetime.date,
current_end_date: datetime.date,
total_justifications: List[Dict[str, Any]],
store_on_ceph: bool = False,
store_on_public_bucket: bool = False,
):
"""Explore adviser files to gather info for contributors."""
daily_processed_dataframes: Dict[str, pd.DataFrame] = {}
adviser_files = Adviser.aggregate_adviser_results(
start_date=current_initial_date, end_date=current_end_date)
if not adviser_files:
_LOGGER.info("No adviser files identifed!")
return total_justifications
dataframes = Adviser.create_adviser_dataframes(adviser_files=adviser_files)
daily_justifications = retrieve_processed_justifications_dataframe(
date_=current_initial_date, dataframes=dataframes)
daily_processed_dataframes["adviser_justifications"] = pd.DataFrame(
daily_justifications)
if not daily_processed_dataframes[
"adviser_justifications"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser justifications:"
f'\n{daily_processed_dataframes["adviser_justifications"].to_csv(header=False, sep="`", index=False)}'
)
daily_statistics = retrieve_processed_statistics_dataframe(
date_=current_initial_date, dataframes=dataframes)
daily_processed_dataframes["adviser_statistics"] = pd.DataFrame(
daily_statistics)
if not daily_processed_dataframes[
"adviser_statistics"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser statistics success rate:"
f'\n{daily_processed_dataframes["adviser_statistics"].to_csv(header=False, sep="`", index=False)}'
)
daily_inputs_info = retrieve_processed_inputs_info_dataframe(
date_=current_initial_date, dataframes=dataframes)
daily_processed_dataframes["adviser_integration_info"] = pd.DataFrame(
daily_inputs_info["integration_info"])
daily_processed_dataframes["adviser_recommendation_info"] = pd.DataFrame(
daily_inputs_info["recommendation_info"])
daily_processed_dataframes["adviser_solver_info"] = pd.DataFrame(
daily_inputs_info["solver_info"])
daily_processed_dataframes["adviser_base_image_info"] = pd.DataFrame(
daily_inputs_info["base_image_info"])
daily_processed_dataframes["adviser_hardware_info"] = pd.DataFrame(
daily_inputs_info["hardware_info"])
if not daily_processed_dataframes[
"adviser_integration_info"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser integration info stats:"
f'\n{daily_processed_dataframes["adviser_integration_info"].to_csv(header=False, sep="`", index=False)}'
)
if not daily_processed_dataframes[
"adviser_recommendation_info"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser recomendation info stats:"
f'\n{daily_processed_dataframes["adviser_recommendation_info"].to_csv(header=False, sep="`", index=False)}'
)
if not daily_processed_dataframes[
"adviser_solver_info"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser solver info stats:"
f'\n{daily_processed_dataframes["adviser_solver_info"].to_csv(header=False, sep="`", index=False)}'
)
if not daily_processed_dataframes[
"adviser_base_image_info"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser base image info stats:"
f'\n{daily_processed_dataframes["adviser_base_image_info"].to_csv(header=False, sep="`", index=False)}'
)
if not daily_processed_dataframes[
"adviser_hardware_info"].empty and not store_on_ceph:
_LOGGER.info(
"Adviser hardware info stats:"
f'\n{daily_processed_dataframes["adviser_hardware_info"].to_csv(header=False, sep="`", index=False)}'
)
if store_on_ceph:
for result_class, processed_df in daily_processed_dataframes.items():
save_results_to_ceph(
processed_df=processed_df,
result_class=result_class,
date_filter=current_initial_date,
store_to_public_ceph=store_on_public_bucket,
)
total_justifications += daily_justifications
return daily_processed_dataframes
def test_get_adviser_files(self) -> None:
"""Test retrieving adviser results from local path."""
adviser_files = Adviser.aggregate_adviser_results(
repo_path=self._ADVISER_FOLDER_PATH, is_local=True)
assert adviser_files