ref: move grouping module into issues

pyproject.toml

@@ -396,15 +396,6 @@ module = [
     "sentry.eventtypes.error",
     "sentry.feedback.migrations.*",
     "sentry.flags.migrations.*",
-    "sentry.grouping.api",
-    "sentry.grouping.component",
-    "sentry.grouping.enhancer.parser",
-    "sentry.grouping.fingerprinting.*",
-    "sentry.grouping.grouping_info",
-    "sentry.grouping.ingest.*",
-    "sentry.grouping.parameterization",
-    "sentry.grouping.utils",
-    "sentry.grouping.variants",
     "sentry.hybridcloud.*",
     "sentry.identity.discord.*",
     "sentry.identity.github_enterprise.*",
@@ -482,6 +473,15 @@ module = [
     "sentry.issues.escalating.escalating_group_forecast",
     "sentry.issues.escalating.escalating_issues_alg",
     "sentry.issues.escalating.forecasts",
+    "sentry.issues.grouping.api",
+    "sentry.issues.grouping.component",
+    "sentry.issues.grouping.enhancer.parser",
+    "sentry.issues.grouping.fingerprinting.*",
+    "sentry.issues.grouping.grouping_info",
+    "sentry.issues.grouping.ingest.*",
+    "sentry.issues.grouping.parameterization",
+    "sentry.issues.grouping.utils",
+    "sentry.issues.grouping.variants",
     "sentry.issues.ignored",
     "sentry.issues.ingest",
     "sentry.issues.issue_occurrence",
@@ -682,7 +682,6 @@ module = [
     "tests.sentry.eventstream.kafka.*",
     "tests.sentry.eventtypes.*",
     "tests.sentry.flags.providers.*",
-    "tests.sentry.grouping.*",
     "tests.sentry.hybridcloud.*",
     "tests.sentry.incidents.handlers.*",
     "tests.sentry.incidents.serializers.*",
@@ -702,6 +701,7 @@ module = [
     "tests.sentry.issues.auto_source_code_config.*",
     "tests.sentry.issues.endpoints.*",
     "tests.sentry.issues.escalating.*",
+    "tests.sentry.issues.grouping.*",
     "tests.sentry.issues.ownership.*",
     "tests.sentry.issues.test_attributes",
     "tests.sentry.issues.test_group_attributes_dataset",

src/sentry/api/endpoints/project_rule_actions.py

@@ -10,7 +10,7 @@
 from sentry.api.base import region_silo_endpoint
 from sentry.api.bases import ProjectAlertRulePermission, ProjectEndpoint
 from sentry.api.serializers.rest_framework import DummyRuleSerializer
-from sentry.grouping.grouptype import ErrorGroupType
+from sentry.issues.grouping.grouptype import ErrorGroupType
 from sentry.models.rule import Rule
 from sentry.notifications.notification_action.utils import should_fire_workflow_actions
 from sentry.notifications.types import TEST_NOTIFICATION_ID

src/sentry/core/endpoints/project_details.py

@@ -37,10 +37,10 @@
 from sentry.dynamic_sampling import get_supported_biases_ids, get_user_biases
 from sentry.dynamic_sampling.types import DynamicSamplingMode
 from sentry.dynamic_sampling.utils import has_custom_dynamic_sampling, has_dynamic_sampling
-from sentry.grouping.enhancer import Enhancements
-from sentry.grouping.enhancer.exceptions import InvalidEnhancerConfig
-from sentry.grouping.fingerprinting import FingerprintingRules, InvalidFingerprintingConfig
 from sentry.ingest.inbound_filters import FilterTypes
+from sentry.issues.grouping.enhancer import Enhancements
+from sentry.issues.grouping.enhancer.exceptions import InvalidEnhancerConfig
+from sentry.issues.grouping.fingerprinting import FingerprintingRules, InvalidFingerprintingConfig
 from sentry.issues.highlights import HighlightContextField
 from sentry.lang.native.sources import (
     InvalidSourcesError,

src/sentry/event_manager.py

@@ -49,35 +49,41 @@
 from sentry.eventtypes import EventType
 from sentry.eventtypes.transaction import TransactionEvent
 from sentry.exceptions import HashDiscarded
-from sentry.grouping.api import (
+from sentry.ingest.inbound_filters import FilterStatKeys
+from sentry.ingest.transaction_clusterer.datasource.redis import (
+    record_transaction_name as record_transaction_name_for_clustering,
+)
+from sentry.integrations.tasks.kick_off_status_syncs import kick_off_status_syncs
+from sentry.issues.grouping.api import (
     NULL_GROUPHASH_INFO,
     GroupHashInfo,
     GroupingConfig,
     get_grouping_config_dict_for_project,
 )
-from sentry.grouping.enhancer import get_enhancements_version
-from sentry.grouping.grouptype import ErrorGroupType
-from sentry.grouping.ingest.config import is_in_transition, update_or_set_grouping_config_if_needed
-from sentry.grouping.ingest.hashing import (
+from sentry.issues.grouping.enhancer import get_enhancements_version
+from sentry.issues.grouping.grouptype import ErrorGroupType
+from sentry.issues.grouping.ingest.config import (
+    is_in_transition,
+    update_or_set_grouping_config_if_needed,
+)
+from sentry.issues.grouping.ingest.hashing import (
     find_grouphash_with_group,
     get_or_create_grouphashes,
     maybe_run_background_grouping,
     maybe_run_secondary_grouping,
     run_primary_grouping,
 )
-from sentry.grouping.ingest.metrics import record_hash_calculation_metrics, record_new_group_metrics
-from sentry.grouping.ingest.seer import maybe_check_seer_for_matching_grouphash
-from sentry.grouping.ingest.utils import (
+from sentry.issues.grouping.ingest.metrics import (
+    record_hash_calculation_metrics,
+    record_new_group_metrics,
+)
+from sentry.issues.grouping.ingest.seer import maybe_check_seer_for_matching_grouphash
+from sentry.issues.grouping.ingest.utils import (
     add_group_id_to_grouphashes,
     check_for_group_creation_load_shed,
     is_non_error_type_group,
 )
-from sentry.grouping.variants import BaseVariant
-from sentry.ingest.inbound_filters import FilterStatKeys
-from sentry.ingest.transaction_clusterer.datasource.redis import (
-    record_transaction_name as record_transaction_name_for_clustering,
-)
-from sentry.integrations.tasks.kick_off_status_syncs import kick_off_status_syncs
+from sentry.issues.grouping.variants import BaseVariant
 from sentry.issues.issue_occurrence import IssueOccurrence
 from sentry.issues.producer import PayloadType, produce_occurrence_to_kafka
 from sentry.killswitches import killswitch_matches_context

src/sentry/feedback/endpoints/organization_feedback_categories.py

@@ -20,7 +20,7 @@
     query_recent_feedbacks_with_ai_labels,
     query_top_ai_labels_by_feedback_count,
 )
-from sentry.grouping.utils import hash_from_values
+from sentry.issues.grouping.utils import hash_from_values
 from sentry.models.organization import Organization
 from sentry.seer.seer_setup import has_seer_access
 from sentry.seer.signed_seer_api import sign_with_seer_secret

src/sentry/feedback/endpoints/organization_feedback_summary.py

@@ -15,7 +15,7 @@
 from sentry.api.bases.organization import OrganizationEndpoint, OrganizationUserReportsPermission
 from sentry.api.utils import get_date_range_from_stats_period
 from sentry.exceptions import InvalidParams
-from sentry.grouping.utils import hash_from_values
+from sentry.issues.grouping.utils import hash_from_values
 from sentry.issues.grouptype import FeedbackGroup
 from sentry.models.group import Group, GroupStatus
 from sentry.models.organization import Organization

src/sentry/grouping/__init__.py

@@ -1,152 +1,3 @@
-"""
-sentry.grouping
-~~~~~~~~~~~~~~~
-
-This package implements various grouping-related functionality in Sentry.
-It's an evolution of an earlier grouping system and the backwards compatibility
-with that grouping system has created a certain amount of complexity.
-
-General Overview
-----------------
-
-Events are grouped together into issues by calculating hashes.  When a hash
-is already associated with a group that group is reused.  The grouping code
-can however generate more than one hash, the sum of which will be added to a
-group.  In the database this is represented in the `GroupHash` model.
-
-Grouping hashes can be created from different methods.  These are called
-variants and are implemented in `sentry.grouping.variants`.  The
-`get_grouping_variants_for_event` function of the grouping API evaluates all
-variants that would be available for the event and returns them.
-
-Afterward each variant can produce hashes which are then used for grouping.
-Some variants (like the component variants) will use the component system
-to annotate the inputs into the hashing function so that they can be shown
-to the user.  These values are not persisted, but the configuration inputs are
-stored in the event.
-
-For reproducibility an event config is persisted in the event dictionary
-consisting of two main sources of configuration: event enhancers and the
-version of the grouping config.  Enhancers are custom rules applied to the
-default grouping algorithm (specifically the stacktrace) and the base version
-picks one of the many versions of the algorithm.
-
-The default version of the algorithm is selected on a per project basis.
-
-Variants
---------
-
-`ChecksumVariant`:
-    This is the legacy variant of the grouping algorithm.  Clients at one point
-    in the past were able to provide a grouping hash (the checksum) which was
-    used for grouping exclusively.
-
-`HashedChecksumVariant`:
-    A ChecksumVariant with a hash value which has been normalized by running the
-    provided checksum through a hashing function.
-
-`FallbackVariant`:
-    This variant produces always the same hash.  It's used if nothing else works.
-
-`ComponentVariant`:
-    This is the main variant which uses the underlying component-based grouping
-    strategy system.  It produces hashes but can also expose the component
-    tree with annotations so users can debug it.
-
-`CustomFingerprintVariant`:
-    This variant is similar to the checksum variant but instead produces
-    fingerprint values which are then hashed.  This can be used both for
-    events produced by the server as well as events which are fingerprinted
-    by the SDK.  If the special `{{ default }}` value is used then this
-    variant is not used.
-
-`BuiltInFingerprintVariant`:
-    Same as the custom fingerprint variant but produced by pre-defined
-    built-in rules.  This is used for server-side fingerprinting.
-
-`SaltedComponentVariant`:
-    This variant is used when the server or client produces a fingerprint
-    that refers (with the special `{{ default }}` value) to the default
-    grouping algorithm and the default grouping algorithm uses the
-    component based grouping system.
-
-Component-Based Grouping
-------------------------
-
-The component-based grouping system produces a tree of nodes (referred to
-as a component) to produce the hash.  These are defined in
-`sentry.grouping.component`.  A component has an `id` which is a string
-identifying the component in the tree.  This `id` is not unique in the tree
-but unique on one level of the hierarchy.  It has an optional `hint` which is
-a string that carries information for the user about why a value was used, not
-used, or why it might have been modified.  A component can additionally  carry `values`,
-which are either components or primitives (strings or integers).  Lastly a
-component has a boolean `contributes` flag.  When set to `False`, this flag
-removes a component (and its children) entirely from the grouping output.
-
-Here an example of how components can be used::
-
-    function_name = 'lambda$1234'
-    threads = BaseGroupingComponent(
-        id="function",
-        values=[function_name],
-        contributes=False,
-        hint='Unused because generated code'
-    )
-
-Strategies and Strategy Configurations
---------------------------------------
-
-Component-based grouping is triggered by strategies.  Strategies define the
-logic for how to generate a component for an interface in the event.  Each
-strategy can only produce a component for one interface.  For instance, a
-stacktrace strategy can produce a component tree for a stacktrace.  Because
-events can have different forms and different strategies for the same interface,
-strategy configurations define which ones are picked.
-
-For instance, there was a `frame:legacy` strategy, which was the legacy
-version of the `frame` strategy.  Then there are the new ones (`frame:v1`,
-`frame:v2`, etc.).  The strategy configuration defines which one is used.
-These are in `sentry.grouping.strategies.configurations`.  A strategy can
-inherit from another one, in which case a lot of behavior is inherited unless
-overridden.
-
-This for instance is how one of the configurations is defined::
-
-    register_strategy_config(
-        id="newstyle:SomeDate",
-        base="newstyle:AnotherDate",
-        delegates=["frame:v4"],
-    )
-
-The configuration ID (`newstyle:YYYY-MM-DD`) is defined in the project
-options and then becomes the strategy configuration of choice for all new
-events.  Because in this case it inherits from another strategy, the default
-configurations from that strategy are reused.  Here the `frame` is changed
-to version `v4`.
-
-Note that here the frame is defined as a delegate.  A delegate is a strategy
-that is used for an interface which by itself is not used for grouping.  This
-means that just because an event has a frame, the frame strategy does not necessarily
-activate.  Only if another interface recurses down into a frame will this strategy
-will be used.
-
-Fingerprinting and Enhancements
--------------------------------
-
-Server-side fingerprinting and enhancements are also folded into this grouping
-system.  The former is in `sentry.grouping.fingerprinting` and can fundamentally
-override the default grouping system.  Enhancements are used by the stacktrace
-strategies to improve how stacktraces are used for grouping.  Enhancements
-activate in two places: as part of stacktrace normalization to update the
-`in-app` flag and later on during grouping by the stacktrace strategy, to add or remove
-frames from the grouping algorithm.
-
-Testing
--------
-
-Tests for grouping are in `tests/sentry/grouping`.  They are snapshot-based,
-and snapshots exist for all tests and all versions of the grouping algorithm.
-If you add a new one, copy the folder in the snapshot directory over to match
-the new name, which makes it easier to evaluate differences.
-"""
+# Shim for backward compatibility with getsentry
+# The grouping module has been moved to sentry.issues.grouping
+from sentry.issues.grouping import *  # noqa: F401, F403

src/sentry/grouping/grouptype.py

@@ -1,40 +1,3 @@
-from dataclasses import dataclass
-from typing import TypeVar
-
-from sentry.issues.grouptype import GroupCategory, GroupType
-from sentry.models.group import DEFAULT_TYPE_ID
-from sentry.types.group import PriorityLevel
-from sentry.workflow_engine.endpoints.validators.error_detector import ErrorDetectorValidator
-from sentry.workflow_engine.handlers.detector.base import DetectorHandler
-from sentry.workflow_engine.models.data_source import DataPacket
-from sentry.workflow_engine.types import (
-    DetectorEvaluationResult,
-    DetectorGroupKey,
-    DetectorSettings,
-)
-
-T = TypeVar("T")
-
-
-class ErrorDetectorHandler(DetectorHandler):
-    def evaluate(
-        self, data_packet: DataPacket[T]
-    ) -> dict[DetectorGroupKey, DetectorEvaluationResult]:
-        # placeholder
-        return {}
-
-
-@dataclass(frozen=True)
-class ErrorGroupType(GroupType):
-    type_id = DEFAULT_TYPE_ID
-    slug = "error"
-    description = "Error"
-    category = GroupCategory.ERROR.value
-    category_v2 = GroupCategory.ERROR.value
-    default_priority = PriorityLevel.MEDIUM
-    released = True
-    detector_settings = DetectorSettings(
-        handler=ErrorDetectorHandler,
-        validator=ErrorDetectorValidator,
-        config_schema={"type": "object", "additionalProperties": False},
-    )
+# Shim for backward compatibility with getsentry
+# The grouping module has been moved to sentry.issues.grouping
+from sentry.issues.grouping.grouptype import *  # noqa: F401, F403

src/sentry/grouping/ingest/__init__.py

@@ -0,0 +1,3 @@
+# Shim for backward compatibility with getsentry
+# The grouping module has been moved to sentry.issues.grouping
+from sentry.issues.grouping.ingest import *  # noqa: F401, F403