Skip to content

Commit

Permalink
[Backport 1.6.latest] Dc/8546 semantic models in graph selection (#8724)
Browse files Browse the repository at this point in the history
* Dc/8546 semantic models in graph selection (#8589)

(cherry picked from commit bb4214b)

* Update test_list.py to remove config "access" and remove group
comparison from SemanticModel since there is no group

---------

Co-authored-by: dave-connors-3 <[email protected]>
Co-authored-by: Gerda Shank <[email protected]>
  • Loading branch information
3 people authored Sep 27, 2023
1 parent c51271e commit c399577
Show file tree
Hide file tree
Showing 10 changed files with 281 additions and 14 deletions.
6 changes: 6 additions & 0 deletions .changes/unreleased/Fixes-20230925-233306.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,6 @@
kind: Fixes
body: semantic models in graph selection
time: 2023-09-25T23:33:06.754344+01:00
custom:
Author: dave-connors-3 michelleark
Issue: "8589"
1 change: 1 addition & 0 deletions core/dbt/cli/params.py
Original file line number Diff line number Diff line change
Expand Up @@ -368,6 +368,7 @@
type=ChoiceTuple(
[
"metric",
"semantic_model",
"source",
"analysis",
"model",
Expand Down
4 changes: 2 additions & 2 deletions core/dbt/compilation.py
Original file line number Diff line number Diff line change
Expand Up @@ -183,10 +183,10 @@ def link_node(self, node: GraphMemberNode, manifest: Manifest):
def link_graph(self, manifest: Manifest):
for source in manifest.sources.values():
self.add_node(source.unique_id)
for semantic_model in manifest.semantic_models.values():
self.add_node(semantic_model.unique_id)
for node in manifest.nodes.values():
self.link_node(node, manifest)
for semantic_model in manifest.semantic_models.values():
self.link_node(semantic_model, manifest)
for exposure in manifest.exposures.values():
self.link_node(exposure, manifest)
for metric in manifest.metrics.values():
Expand Down
46 changes: 46 additions & 0 deletions core/dbt/contracts/graph/nodes.py
Original file line number Diff line number Diff line change
Expand Up @@ -1603,6 +1603,52 @@ def primary_entity_reference(self) -> Optional[EntityReference]:
else None
)

def same_model(self, old: "SemanticModel") -> bool:
return self.model == old.same_model

def same_node_relation(self, old: "SemanticModel") -> bool:
return self.node_relation == old.node_relation

def same_description(self, old: "SemanticModel") -> bool:
return self.description == old.description

def same_defaults(self, old: "SemanticModel") -> bool:
return self.defaults == old.defaults

def same_entities(self, old: "SemanticModel") -> bool:
return self.entities == old.entities

def same_dimensions(self, old: "SemanticModel") -> bool:
return self.dimensions == old.dimensions

def same_measures(self, old: "SemanticModel") -> bool:
return self.measures == old.measures

def same_config(self, old: "SemanticModel") -> bool:
return self.config == old.config

def same_primary_entity(self, old: "SemanticModel") -> bool:
return self.primary_entity == old.primary_entity

def same_contents(self, old: Optional["SemanticModel"]) -> bool:
# existing when it didn't before is a change!
# metadata/tags changes are not "changes"
if old is None:
return True

return (
self.same_model(old)
and self.same_node_relation(old)
and self.same_description(old)
and self.same_defaults(old)
and self.same_entities(old)
and self.same_dimensions(old)
and self.same_measures(old)
and self.same_config(old)
and self.same_primary_entity(old)
and True
)


# ====================================
# Patches
Expand Down
2 changes: 1 addition & 1 deletion core/dbt/graph/cli.py
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@

INTERSECTION_DELIMITER = ","

DEFAULT_INCLUDES: List[str] = ["fqn:*", "source:*", "exposure:*", "metric:*"]
DEFAULT_INCLUDES: List[str] = ["fqn:*", "source:*", "exposure:*", "metric:*", "semantic_model:*"]
DEFAULT_EXCLUDES: List[str] = []


Expand Down
48 changes: 44 additions & 4 deletions core/dbt/graph/selector_methods.py
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@
ResultNode,
ManifestNode,
ModelNode,
SemanticModel,
)
from dbt.contracts.graph.unparsed import UnparsedVersion
from dbt.contracts.state import PreviousState
Expand Down Expand Up @@ -53,6 +54,7 @@ class MethodName(StrEnum):
SourceStatus = "source_status"
Wildcard = "wildcard"
Version = "version"
SemanticModel = "semantic_model"


def is_selected_node(fqn: List[str], node_selector: str, is_versioned: bool) -> bool:
Expand Down Expand Up @@ -144,6 +146,16 @@ def metric_nodes(self, included_nodes: Set[UniqueId]) -> Iterator[Tuple[UniqueId
continue
yield unique_id, metric

def semantic_model_nodes(
self, included_nodes: Set[UniqueId]
) -> Iterator[Tuple[UniqueId, SemanticModel]]:

for key, semantic_model in self.manifest.semantic_models.items():
unique_id = UniqueId(key)
if unique_id not in included_nodes:
continue
yield unique_id, semantic_model

def all_nodes(
self, included_nodes: Set[UniqueId]
) -> Iterator[Tuple[UniqueId, SelectorTarget]]:
Expand All @@ -152,6 +164,7 @@ def all_nodes(
self.source_nodes(included_nodes),
self.exposure_nodes(included_nodes),
self.metric_nodes(included_nodes),
self.semantic_model_nodes(included_nodes),
)

def configurable_nodes(
Expand All @@ -167,6 +180,7 @@ def non_source_nodes(
self.parsed_nodes(included_nodes),
self.exposure_nodes(included_nodes),
self.metric_nodes(included_nodes),
self.semantic_model_nodes(included_nodes),
)

def groupable_nodes(
Expand Down Expand Up @@ -210,8 +224,8 @@ def search(self, included_nodes: Set[UniqueId], selector: str) -> Iterator[Uniqu
:param str selector: The selector or node name
"""
parsed_nodes = list(self.parsed_nodes(included_nodes))
for node, real_node in parsed_nodes:
non_source_nodes = list(self.non_source_nodes(included_nodes))
for node, real_node in non_source_nodes:
if self.node_is_match(selector, real_node.fqn, real_node.is_versioned):
yield node

Expand Down Expand Up @@ -322,6 +336,31 @@ def search(self, included_nodes: Set[UniqueId], selector: str) -> Iterator[Uniqu
yield node


class SemanticModelSelectorMethod(SelectorMethod):
def search(self, included_nodes: Set[UniqueId], selector: str) -> Iterator[UniqueId]:
parts = selector.split(".")
target_package = SELECTOR_GLOB
if len(parts) == 1:
target_name = parts[0]
elif len(parts) == 2:
target_package, target_name = parts
else:
msg = (
'Invalid semantic model selector value "{}". Semantic models must be of '
"the form ${{semantic_model_name}} or "
"${{semantic_model_package.semantic_model_name}}"
).format(selector)
raise DbtRuntimeError(msg)

for node, real_node in self.semantic_model_nodes(included_nodes):
if not fnmatch(real_node.package_name, target_package):
continue
if not fnmatch(real_node.name, target_name):
continue

yield node


class PathSelectorMethod(SelectorMethod):
def search(self, included_nodes: Set[UniqueId], selector: str) -> Iterator[UniqueId]:
"""Yields nodes from included that match the given path."""
Expand Down Expand Up @@ -431,7 +470,7 @@ def search(self, included_nodes: Set[UniqueId], selector: str) -> Iterator[Uniqu
resource_type = NodeType(selector)
except ValueError as exc:
raise DbtRuntimeError(f'Invalid resource_type selector "{selector}"') from exc
for node, real_node in self.parsed_nodes(included_nodes):
for node, real_node in self.all_nodes(included_nodes):
if real_node.resource_type == resource_type:
yield node

Expand Down Expand Up @@ -539,7 +578,7 @@ def check_macros_modified(self, node):
def check_modified_content(
self, old: Optional[SelectorTarget], new: SelectorTarget, adapter_type: str
) -> bool:
if isinstance(new, (SourceDefinition, Exposure, Metric)):
if isinstance(new, (SourceDefinition, Exposure, Metric, SemanticModel)):
# these all overwrite `same_contents`
different_contents = not new.same_contents(old) # type: ignore
else:
Expand Down Expand Up @@ -761,6 +800,7 @@ class MethodManager:
MethodName.Result: ResultSelectorMethod,
MethodName.SourceStatus: SourceStatusSelectorMethod,
MethodName.Version: VersionSelectorMethod,
MethodName.SemanticModel: SemanticModelSelectorMethod,
}

def __init__(
Expand Down
9 changes: 8 additions & 1 deletion core/dbt/task/list.py
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
import json

from dbt.contracts.graph.nodes import Exposure, SourceDefinition, Metric
from dbt.contracts.graph.nodes import Exposure, SourceDefinition, Metric, SemanticModel
from dbt.flags import get_flags
from dbt.graph import ResourceTypeSelector
from dbt.task.runnable import GraphRunnableTask
Expand Down Expand Up @@ -28,6 +28,7 @@ class ListTask(GraphRunnableTask):
NodeType.Source,
NodeType.Exposure,
NodeType.Metric,
NodeType.SemanticModel,
)
)
ALL_RESOURCE_VALUES = DEFAULT_RESOURCE_VALUES | frozenset((NodeType.Analysis,))
Expand Down Expand Up @@ -74,6 +75,8 @@ def _iterate_selected_nodes(self):
yield self.manifest.exposures[node]
elif node in self.manifest.metrics:
yield self.manifest.metrics[node]
elif node in self.manifest.semantic_models:
yield self.manifest.semantic_models[node]
else:
raise DbtRuntimeError(
f'Got an unexpected result from node selection: "{node}"'
Expand All @@ -97,6 +100,10 @@ def generate_selectors(self):
# metrics are searched for by pkg.metric_name
metric_selector = ".".join([node.package_name, node.name])
yield f"metric:{metric_selector}"
elif node.resource_type == NodeType.SemanticModel:
assert isinstance(node, SemanticModel)
semantic_model_selector = ".".join([node.package_name, node.name])
yield f"semantic_model:{semantic_model_selector}"
else:
# everything else is from `fqn`
yield ".".join(node.fqn)
Expand Down
56 changes: 55 additions & 1 deletion tests/functional/list/fixtures.py
Original file line number Diff line number Diff line change
Expand Up @@ -46,7 +46,16 @@
{{ config(materialized='ephemeral') }}
select 1 as id
select
1 as id,
{{ dbt.date_trunc('day', dbt.current_timestamp()) }} as created_at
"""

models__metric_flow = """
select
{{ dbt.date_trunc('day', dbt.current_timestamp()) }} as date_day
"""

Expand Down Expand Up @@ -103,6 +112,38 @@
"""

semantic_models__sm_yml = """
semantic_models:
- name: my_sm
model: ref('outer')
defaults:
agg_time_dimension: created_at
entities:
- name: my_entity
type: primary
expr: id
dimensions:
- name: created_at
type: time
type_params:
time_granularity: day
measures:
- name: total_outer_count
agg: count
expr: 1
"""

metrics__m_yml = """
metrics:
- name: total_outer
type: simple
description: The total count of outer
label: Total Outer
type_params:
measure: total_outer_count
"""


@pytest.fixture(scope="class")
def snapshots():
Expand All @@ -122,6 +163,9 @@ def models():
"incremental.sql": models__incremental_sql,
"docs.md": models__docs_md,
"outer.sql": models__outer_sql,
"metricflow_time_spine.sql": models__metric_flow,
"sm.yml": semantic_models__sm_yml,
"m.yml": metrics__m_yml,
"sub": {"inner.sql": models__sub__inner_sql},
}

Expand All @@ -141,6 +185,16 @@ def analyses():
return {"a.sql": analyses__a_sql}


@pytest.fixture(scope="class")
def semantic_models():
return {"sm.yml": semantic_models__sm_yml}


@pytest.fixture(scope="class")
def metrics():
return {"m.yml": metrics__m_yml}


@pytest.fixture(scope="class")
def project_files(
project_root,
Expand Down
Loading

0 comments on commit c399577

Please sign in to comment.