Quellcodebibliothek Statistik Leitseite products/Sources/formale Sprachen/C/Firefox/taskcluster/gecko_taskgraph/actions/   (Browser von der Mozilla Stiftung Version 136.0.1©)  Datei vom 10.2.2025 mit Größe 9 kB image not shown  

Quelle  confirm_failure.py   Sprache: Python

 
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.


import json
import logging
from functools import partial

from taskgraph.util.taskcluster import get_artifact, get_task_definition, list_artifacts

from .registry import register_callback_action
from .retrigger import retrigger_action
from .util import add_args_to_command, create_tasks, fetch_graph_and_labels

logger = logging.getLogger(__name__)


def get_failures(task_id, task_definition):
    """Returns a dict containing properties containing a list of
    directories containing test failures and a separate list of
    individual test failures from the errorsummary.log artifact for
    the task.

    Find test path to pass to the task in
    MOZHARNESS_TEST_PATHS.  If no appropriate test path can be
    determined, nothing is returned.
    """

    def fix_wpt_name(test):
        # TODO: find other cases to handle
        if ".any." in test:
            test = "%s.any.js" % test.split(".any.")[0]
        if ".window.html" in test:
            test = test.replace(".window.html"".window.js")

        if test.startswith("/_mozilla"):
            test = "testing/web-platform/mozilla/tests" + test[len("_mozilla") :]
        else:
            test = "testing/web-platform/tests/" + test.strip("/")
        # some wpt tests have params, those are not supported
        test = test.split("?")[0]

        return test

    # collect dirs that don't have a specific manifest
    dirs = []
    tests = []

    artifacts = list_artifacts(task_id)
    for artifact in artifacts:
        if "name" not in artifact or not artifact["name"].endswith("errorsummary.log"):
            continue

        stream = get_artifact(task_id, artifact["name"])
        if not stream:
            continue

        # We handle the stream as raw bytes because it may contain invalid
        # UTF-8 characters in portions other than those containing the error
        # messages we're looking for.
        for line in stream.read().split(b"\n"):
            if not line.strip():
                continue

            l = json.loads(line)
            if "group_results" in l.keys() and l["status"] != "OK":
                dirs.append(l["group_results"].group())

            elif "test" in l.keys():
                if not l["test"]:
                    print("Warning: no testname in errorsummary line: %s" % l)
                    continue

                test_path = l["test"].split(" ")[0]
                found_path = False

                # tests with url params (wpt), will get confused here
                if "?" not in test_path:
                    test_path = test_path.split(":")[-1]

                # edge case where a crash on shutdown has a "test" name == group name
                if (
                    test_path.endswith(".toml")
                    or test_path.endswith(".ini")
                    or test_path.endswith(".list")
                ):
                    # TODO: consider running just the manifest
                    continue

                # edge cases with missing test names
                if (
                    test_path is None
                    or test_path == "None"
                    or "SimpleTest" in test_path
                ):
                    continue

                if "signature" in l.keys():
                    # dealing with a crash
                    found_path = True
                    if "web-platform" in task_definition["extra"]["suite"]:
                        test_path = fix_wpt_name(test_path)
                else:
                    if "status" not in l and "expected" not in l:
                        continue

                    if l["status"] != l["expected"]:
                        if l["status"not in l.get("known_intermittent", []):
                            found_path = True
                            if "web-platform" in task_definition["extra"]["suite"]:
                                test_path = fix_wpt_name(test_path)

                if found_path and test_path:
                    fpath = test_path.replace("\\""/")
                    tval = {"path": fpath, "group": l["group"]}
                    # only store one failure per test
                    if not [t for t in tests if t["path"] == fpath]:
                        tests.append(tval)

                # only run the failing test not both test + dir
                if l["group"in dirs:
                    dirs.remove(l["group"])

            # TODO: 10 is too much; how to get only NEW failures?
            if len(tests) > 10:
                break

    dirs = [{"path""""group": d} for d in list(set(dirs))]
    return {"dirs": dirs, "tests": tests}


def get_repeat_args(task_definition, failure_group):
    task_name = task_definition["metadata"]["name"]
    repeatable_task = False
    if (
        "crashtest" in task_name
        or "mochitest" in task_name
        or "reftest" in task_name
        or "xpcshell" in task_name
        or "web-platform" in task_name
        and "jsreftest" not in task_name
    ):
        repeatable_task = True

    repeat_args = ""
    if not repeatable_task:
        return repeat_args

    if failure_group == "dirs":
        # execute 3 total loops
        repeat_args = ["--repeat=2"if repeatable_task else []
    elif failure_group == "tests":
        # execute 5 total loops
        repeat_args = ["--repeat=4"if repeatable_task else []

    return repeat_args


def confirm_modifier(task, input):
    if task.label != input["label"]:
        return task

    logger.debug(f"Modifying paths for {task.label}")

    # If the original task has defined test paths
    suite = input.get("suite")
    test_path = input.get("test_path")
    test_group = input.get("test_group")
    if test_path or test_group:
        repeat_args = input.get("repeat_args")

        if repeat_args:
            task.task["payload"]["command"] = add_args_to_command(
                task.task["payload"]["command"], extra_args=repeat_args
            )

        # TODO: do we need this attribute?
        task.attributes["test_path"] = test_path

        task.task["payload"]["env"]["MOZHARNESS_TEST_PATHS"] = json.dumps(
            {suite: [test_group]}, sort_keys=True
        )
        task.task["payload"]["env"]["MOZHARNESS_CONFIRM_PATHS"] = json.dumps(
            {suite: [test_path]}, sort_keys=True
        )
        task.task["payload"]["env"]["MOZLOG_DUMP_ALL_TESTS"] = "1"

        task.task["metadata"]["name"] = task.label
        task.task["tags"]["action"] = "confirm-failure"
    return task


@register_callback_action(
    name="confirm-failures",
    title="Confirm failures in job",
    symbol="cf",
    description="Re-run Tests for original manifest, directories or tests for failing tests.",
    order=150,
    context=[{"kind""test"}],
    schema={
        "type""object",
        "properties": {
            "label": {"type""string""description""A task label"},
            "suite": {"type""string""description""Test suite"},
            "test_path": {"type""string""description""A full path to test"},
            "test_group": {
                "type""string",
                "description""A full path to group name",
            },
            "repeat_args": {
                "type""string",
                "description""args to pass to test harness",
            },
        },
        "additionalProperties"False,
    },
)
def confirm_failures(parameters, graph_config, input, task_group_id, task_id):
    task_definition = get_task_definition(task_id)
    decision_task_id, full_task_graph, label_to_taskid, _ = fetch_graph_and_labels(
        parameters, graph_config
    )

    # create -cf label; ideally make this a common function
    task_definition["metadata"]["name"].split("-")
    cfname = "%s-cf" % task_definition["metadata"]["name"]

    if cfname not in full_task_graph.tasks:
        raise Exception(f"{cfname} was not found in the task-graph")

    to_run = [cfname]

    suite = task_definition["extra"]["suite"]
    if "-coverage" in suite:
        suite = suite[: suite.index("-coverage")]
    if "-qr" in suite:
        suite = suite[: suite.index("-qr")]
    failures = get_failures(task_id, task_definition)

    if failures["dirs"] == [] and failures["tests"] == []:
        logger.info("need to retrigger task as no specific test failures found")
        retrigger_action(parameters, graph_config, input, decision_task_id, task_id)
        return

    # for each unique failure, create a new confirm failure job
    for failure_group in failures:
        for failure_path in failures[failure_group]:
            repeat_args = get_repeat_args(task_definition, failure_group)

            input = {
                "label": cfname,
                "suite": suite,
                "test_path": failure_path["path"],
                "test_group": failure_path["group"],
                "repeat_args": repeat_args,
            }

            logger.info("confirm_failures: %s" % failures)
            create_tasks(
                graph_config,
                to_run,
                full_task_graph,
                label_to_taskid,
                parameters,
                decision_task_id,
                modifier=partial(confirm_modifier, input=input),
            )

Messung V0.5
C=97 H=100 G=98

¤ Dauer der Verarbeitung: 0.16 Sekunden  (vorverarbeitet)  ¤

*© Formatika GbR, Deutschland






Wurzel

Suchen

Beweissystem der NASA

Beweissystem Isabelle

NIST Cobol Testsuite

Cephes Mathematical Library

Wiener Entwicklungsmethode

Haftungshinweis

Die Informationen auf dieser Webseite wurden nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit, noch Qualität der bereit gestellten Informationen zugesichert.

Bemerkung:

Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.