Source code for plainbox.impl.session.jobs

# This file is part of Checkbox.
#
# Copyright 2012-2015 Canonical Ltd.
# Written by:
#   Zygmunt Krynicki <zygmunt.krynicki@canonical.com>
#
# Checkbox is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License version 3,
# as published by the Free Software Foundation.
#
# Checkbox is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Checkbox.  If not, see <http://www.gnu.org/licenses/>.
"""
Job State.

:mod:`plainbox.impl.session.jobs` -- jobs state handling
========================================================

This module contains a helper class for associating job state within a
particular session. The :class:`JobState` class holds references to a
:class:`JobDefinition` and :class:`JobResult` as well as a list of inhibitors
that prevent the job from being runnable in a particular session.
"""

from enum import IntEnum
import logging

from plainbox.abc import IJobResult
from plainbox.i18n import gettext as _
from plainbox.impl import pod
from plainbox.impl.resource import ResourceExpression
from plainbox.impl.result import MemoryJobResult
from plainbox.impl.unit.job import JobDefinition

logger = logging.getLogger("plainbox.session.jobs")


[docs]class InhibitionCause(IntEnum): """ There are four possible not-ready causes. UNDESIRED: This job was not selected to run in this session PENDING_DEP: This job depends on another job which was not started yet FAILED_DEP: This job depends on another job which was started and failed PENDING_RESOURCE: This job has a resource requirement expression that uses a resource produced by another job which was not started yet FAILED_RESOURCE: This job has a resource requirement that evaluated to a false value """ UNDESIRED = 0 PENDING_DEP = 1 FAILED_DEP = 2 PENDING_RESOURCE = 3 FAILED_RESOURCE = 4
[docs]def cause_convert_assign_filter( instance: pod.POD, field: pod.Field, old: "Any", new: "Any") -> "Any": """ Assign filter for for JobReadinessInhibitor.cause. Custom assign filter for the JobReadinessInhibitor.cause field that produces a very specific error message. """ try: return pod.type_convert_assign_filter(instance, field, old, new) except ValueError: raise ValueError(_("unsupported value for cause"))
[docs]class JobReadinessInhibitor(pod.POD): """ Class representing the cause of a job not being ready to execute. It is intended to be consumed by UI layers and to provide them with enough information to render informative error messages or other visual feedback that will aid the user in understanding why a job cannot be started. There are four possible not ready causes: UNDESIRED: This job was not selected to run in this session PENDING_DEP: This job depends on another job which was not started yet FAILED_DEP: This job depends on another job which was started and failed PENDING_RESOURCE: This job has a resource requirement expression that uses a resource produced by another job which was not started yet FAILED_RESOURCE: This job has a resource requirement that evaluated to a false value All causes apart from UNDESIRED use the related_job property to encode a job that is related to the problem. The PENDING_RESOURCE and FAILED_RESOURCE causes also store related_expression that describes the relevant requirement expression. There are three attributes that can be accessed: cause: Encodes the reason why a job is not ready, see :class:`InhibitionCause`. related_job: Provides additional context for the problem. This is not the job that is affected, rather, the job that is causing the problem. related_expression: Provides additional context for the problem caused by a failing resource expression. """ # XXX: PENDING_RESOURCE is not strict, there are multiple states that are # clumped here which is something I don't like. A resource may be still # "pending" as in PENDING_DEP (it has not ran yet) or it could have ran but # failed to produce any data, it could also be prevented from running # because it has unmet dependencies. In essence it tells us nothing about # if related_job.can_start() is true or not. # # XXX: FAILED_RESOURCE is "correct" but somehow misleading, FAILED_RESOURCE # is used to represent a resource expression that evaluated to a non-True # value cause = pod.Field( doc="cause (constant) of the inhibitor", type=InhibitionCause, initial=pod.MANDATORY, assign_filter_list=[cause_convert_assign_filter, pod.read_only_assign_filter]) related_job = pod.Field( doc="an (optional) job reference", type=JobDefinition, assign_filter_list=[pod.read_only_assign_filter]) related_expression = pod.Field( doc="an (optional) resource expression reference", type=ResourceExpression, assign_filter_list=[pod.read_only_assign_filter]) def __init__(self, cause, related_job=None, related_expression=None): """ Initialize a new inhibitor with the specified cause. If cause is other than UNDESIRED a related_job is necessary. If cause is either PENDING_RESOURCE or FAILED_RESOURCE related_expression is necessary as well. A ValueError is raised when this is violated. """ super().__init__(cause, related_job, related_expression) if (self.cause != InhibitionCause.UNDESIRED and self.related_job is None): raise ValueError( # TRANSLATORS: please don't translate related_job, None and # cause _("related_job must not be None when cause is {}").format( self.cause.name)) if (self.cause in (InhibitionCause.PENDING_RESOURCE, InhibitionCause.FAILED_RESOURCE) and self.related_expression is None): raise ValueError(_( # TRANSLATORS: please don't translate related_expression, None # and cause. "related_expression must not be None when cause is {}" ).format(self.cause.name)) def __repr__(self): """Get a custom debugging representation of an inhibitor.""" return "<{} cause:{} related_job:{!r} related_expression:{!r}>".format( self.__class__.__name__, self.cause.name, self.related_job, self.related_expression) def __str__(self): """Get a human-readable text representation of an inhibitor.""" if self.cause == InhibitionCause.UNDESIRED: # TRANSLATORS: as in undesired job return _("undesired") elif self.cause == InhibitionCause.PENDING_DEP: return _("required dependency {!r} did not run yet").format( self.related_job.id) elif self.cause == InhibitionCause.FAILED_DEP: return _("required dependency {!r} has failed").format( self.related_job.id) elif self.cause == InhibitionCause.PENDING_RESOURCE: return _( "resource expression {!r} could not be evaluated because" " the resource it depends on did not run yet" ).format(self.related_expression.text) else: assert self.cause == InhibitionCause.FAILED_RESOURCE return _("resource expression {!r} evaluates to false").format( self.related_expression.text)
# A global instance of :class:`JobReadinessInhibitor` with the UNDESIRED cause. # This is used a lot and it makes no sense to instantiate all the time. UndesiredJobReadinessInhibitor = JobReadinessInhibitor( InhibitionCause.UNDESIRED) JOB_VALUE = object()
[docs]class OverridableJobField(pod.Field): """ A custom Field for modeling job values that can be overridden. A readable-writable field that has a special initial value ``JOB_VALUE`` which is interpreted as "load this value from the corresponding job definition". This field class facilitates implementation of fields that have some per-job value but can be also overridden in a session state context. """ def __init__(self, job_field, doc=None, type=None, notify=False, assign_filter_list=None): """Initialize a new overridable job field.""" super().__init__( doc, type, JOB_VALUE, None, notify, assign_filter_list) self.job_field = job_field def __get__(self, instance, owner): """Get an overriden (if any) value of an overridable job field.""" value = super().__get__(instance, owner) if value is JOB_VALUE: return getattr(instance.job, self.job_field) else: return value
[docs]def job_assign_filter(instance, field, old_value, new_value): """ A custom setter for the JobState.job. .. warning:: This setter should not exist. job attribute should be read-only. This is a temporary kludge to get session restoring over DBus working. Once a solution that doesn't involve setting a JobState's job attribute is implemented, please remove this awful method. """ return new_value
[docs]def job_via_assign_filter(instance, field, old_value, new_value): """A custom setter for JobState.via_job.""" if (old_value is not pod.UNSET and not isinstance(new_value, JobDefinition) and new_value is not None): raise TypeError("via_job must be the actual job, not the checksum") return new_value
[docs]class JobState(pod.POD): """ Class representing the state of a job in a session. Contains the following basic properties of each job: * the readiness_inhibitor_list that prevent the job form starting * the result (outcome) of the run (IJobResult) * the effective category identifier * the effective certification status * the job that was used to create it (via_job) For convenience (to SessionState implementation) it also has a reference to the job itself. This class is a pure state holder an will typically collaborate with the SessionState class and the UI layer. """ job = pod.Field( doc="the job associated with this state", type=JobDefinition, initial=pod.MANDATORY, assign_filter_list=[job_assign_filter]) readiness_inhibitor_list = pod.Field( doc="the list of readiness inhibitors of the associated job", type="List[JobReadinessInhibitor]", initial_fn=lambda: [UndesiredJobReadinessInhibitor]) result = pod.Field( doc="the result of running the associated job", type=IJobResult, initial_fn=lambda: MemoryJobResult({}), notify=True) result_history = pod.Field( doc="a tuple of result_history of the associated job", type=tuple, initial=(), notify=True, assign_filter_list=[pod.typed, pod.typed.sequence(IJobResult)]) via_job = pod.Field( doc="the parent job definition", type=JobDefinition, assign_filter_list=[job_via_assign_filter]) effective_category_id = OverridableJobField( job_field="category_id", doc="the effective categorization of this test in a session", type=str) effective_certification_status = OverridableJobField( job_field="certification_status", doc="the effective certification status of this job", type=str) effective_auto_retry = OverridableJobField( job_field="auto_retry", doc="the ability to automatically retry this job if it fails", type=str) attempts = pod.Field( doc="number of attempts remaining (in auto-retry mode)", type=int, initial_fn=lambda: 3) # NOTE: the `result` property just exposes the last result from the # `result_history` tuple above. The API is used everywhere so it should not # be broken in any way but the way forward is the sequence stored in # `result_history`. # # The one particularly annoying part of this implementation is that each # job state always has at least one result. Even if there was no testing # done yet. This OUTCOME_NONE result needs to be filtered out at various # times. I think it would be better if we could not have it in the # sequence-based API anymore. Otherwise each test will have two # result_history (more if you count things like resuming a session). @result.change_notifier def _result_changed(self, old, new): # Don't track the initial assignment over UNSET if old is pod.UNSET: return assert new != old assert isinstance(new, IJobResult) if new.is_hollow: return logger.debug( "Appending result %r to history: %r", new, self.result_history) self.result_history += (new,)
[docs] def can_start(self): """Quickly check if the associated job can run right now.""" return len(self.readiness_inhibitor_list) == 0 and self.attempts > 0
[docs] def get_readiness_description(self): """Get a human readable description of the current readiness state.""" if self.readiness_inhibitor_list: return _("job cannot be started: {}").format( ", ".join((str(inhibitor) for inhibitor in self.readiness_inhibitor_list))) else: return _("job can be started")
[docs] def apply_overrides(self, override_list: "List[Tuple[str, Any]]"): """ Apply overrides to effective jop values. This method is automatically called by :class:`SessionDeviceContext` to implement effective overrides originating from test plan data. :param override_list: A list, as exposed by values of :attr:`TestPlanUnitSupport.override_list`, composed of a sequence of pairs ``(field, value)``, where ``field`` is the name of the field to override (without the prefix ``effective_``) and value is any valid value of that field. :raises AttributeError: If any of the ``field``s refer to an unknown field. :raises ValueError: If any of the ``field``s refer to fields that are not designated as overridable. :raises ValueError: If the ``value`` supplied is incorrect for the given field. :raises TypeError: If the type of the ``value`` supplied is incorrect for the given field. .. note:: Consult field specification for details on what types and values are valid for that field. Example: >>> from plainbox.vendor.mock import Mock >>> job = Mock(spec=JobDefinition) >>> job_state = JobState(job) >>> job_state.apply_overrides([ ... ('category_id', 'new-category-id'), ... ('certification_status', 'blocker')]) >>> job_state.effective_category_id 'new-category-id' >>> job_state.effective_certification_status 'blocker' """ for field, value in override_list: effective_field = 'effective_{}'.format(field) effective_field_obj = getattr(self.__class__, effective_field) if not isinstance(effective_field_obj, OverridableJobField): raise ValueError(_('{!r} is not overridable').format(field)) setattr(self, effective_field, value) logger.debug("Applied overrides %r to job %r", override_list, self.job)
comments powered by Disqus