Initialize governed Dataflow module

This commit is contained in:
2026-07-28 01:28:17 +02:00
commit 223c008ff3
44 changed files with 6204 additions and 0 deletions

125
tests/test_graph_and_sql.py Normal file
View File

@@ -0,0 +1,125 @@
from __future__ import annotations
import unittest
from govoplan_dataflow.backend.executor import execute_preview
from govoplan_dataflow.backend.schemas import (
GraphEdge,
GraphNode,
GraphPosition,
PipelineGraph,
)
from govoplan_dataflow.backend.sql_compiler import SqlCompilationError, compile_sql, render_sql
def inline_source() -> GraphNode:
return GraphNode(
id="source",
type="source.inline",
label="Monthly input",
position=GraphPosition(x=40, y=160),
config={
"source_name": "monthly_files",
"rows": [
{"department": "A", "status": "open", "amount": 10},
{"department": "A", "status": "closed", "amount": 5},
{"department": "B", "status": "open", "amount": 20},
],
},
)
class DataflowGraphAndSqlTests(unittest.TestCase):
def test_compiles_renders_and_executes_grouped_query(self) -> None:
sql = """
SELECT department, COUNT(*) AS records, SUM(amount) AS total
FROM monthly_files
WHERE status = 'open'
GROUP BY department
ORDER BY total DESC
LIMIT 50
"""
graph, normalized, diagnostics = compile_sql(sql, source_nodes=[inline_source()])
self.assertEqual([], diagnostics)
self.assertEqual(
["source.inline", "filter", "aggregate", "sort", "limit", "output"],
[node.type for node in graph.nodes],
)
self.assertEqual(normalized, render_sql(graph)[0])
result = execute_preview(graph, row_limit=100)
self.assertEqual(
[
{"department": "B", "records": 1, "total": 20},
{"department": "A", "records": 1, "total": 10},
],
result.rows,
)
self.assertEqual(2, result.total_rows)
self.assertFalse(result.truncated)
self.assertEqual(3, result.input_row_count)
self.assertEqual(6, len(result.node_diagnostics))
self.assertEqual(1, len(result.source_fingerprints))
def test_rejects_effectful_or_multi_statement_sql(self) -> None:
for sql in (
"DELETE FROM monthly_files",
"SELECT * FROM monthly_files; DROP TABLE monthly_files",
"INSERT INTO target SELECT * FROM monthly_files",
):
with self.subTest(sql=sql), self.assertRaises(SqlCompilationError):
compile_sql(sql, source_nodes=[inline_source()])
def test_rejects_join_until_comparison_slice(self) -> None:
with self.assertRaisesRegex(SqlCompilationError, "JOIN support"):
compile_sql(
"SELECT * FROM monthly_files JOIN other ON monthly_files.id = other.id",
source_nodes=[inline_source()],
)
def test_graph_validation_rejects_cycle(self) -> None:
source = inline_source()
output = GraphNode(
id="output",
type="output",
label="Output",
position=GraphPosition(x=300, y=160),
config={},
)
graph = PipelineGraph(
nodes=[source, output],
edges=[
GraphEdge(id="forward", source="source", target="output"),
GraphEdge(id="back", source="output", target="source"),
],
)
from govoplan_dataflow.backend.graph import validate_graph
codes = {item.code for item in validate_graph(graph)}
self.assertIn("graph.cycle", codes)
def test_preview_truncates_response_without_changing_pipeline_limit(self) -> None:
source = inline_source()
output = GraphNode(
id="output",
type="output",
label="Output",
position=GraphPosition(x=300, y=160),
config={},
)
graph = PipelineGraph(
nodes=[source, output],
edges=[GraphEdge(id="edge", source="source", target="output")],
)
result = execute_preview(graph, row_limit=2)
self.assertEqual(2, len(result.rows))
self.assertEqual(3, result.total_rows)
self.assertTrue(result.truncated)
if __name__ == "__main__":
unittest.main()

34
tests/test_manifest.py Normal file
View File

@@ -0,0 +1,34 @@
from __future__ import annotations
import unittest
from govoplan_core.core.modules import ModuleManifest
from govoplan_dataflow.backend.manifest import get_manifest
class DataflowManifestTests(unittest.TestCase):
def test_manifest_declares_independent_dataflow_boundary(self) -> None:
manifest = get_manifest()
self.assertIsInstance(manifest, ModuleManifest)
self.assertEqual(manifest.id, "dataflow")
self.assertEqual(manifest.dependencies, ())
self.assertIn("connectors", manifest.optional_dependencies)
self.assertIn("reporting", manifest.optional_dependencies)
self.assertIn("workflow", manifest.optional_dependencies)
self.assertEqual(manifest.frontend.package_name, "@govoplan/dataflow-webui")
self.assertIsNotNone(manifest.route_factory)
self.assertIsNotNone(manifest.migration_spec)
self.assertEqual(
{
"dataflow.pipeline_catalog",
"dataflow.pipeline_preview",
"dataflow.run_lifecycle",
"dataflow.dataset_output",
},
{item.name for item in manifest.provides_interfaces},
)
if __name__ == "__main__":
unittest.main()

47
tests/test_migrations.py Normal file
View File

@@ -0,0 +1,47 @@
from __future__ import annotations
import tempfile
import unittest
from pathlib import Path
from alembic.runtime.migration import MigrationContext
from sqlalchemy import create_engine, inspect
from govoplan_core.db.migrations import migrate_database
from govoplan_dataflow.backend.manifest import get_manifest
class DataflowMigrationTests(unittest.TestCase):
def test_baseline_creates_dataflow_tables_and_head(self) -> None:
with tempfile.TemporaryDirectory(prefix="govoplan-dataflow-migration-") as directory:
url = f"sqlite:///{Path(directory) / 'dataflow.db'}"
migrate_database(
database_url=url,
enabled_modules=("dataflow",),
manifest_factories=(get_manifest,),
)
engine = create_engine(url)
try:
with engine.connect() as connection:
self.assertIn(
"d4f7a1c8e2b0",
set(MigrationContext.configure(connection).get_current_heads()),
)
self.assertEqual(
{
"dataflow_pipelines",
"dataflow_pipeline_revisions",
"dataflow_runs",
},
{
name
for name in inspect(connection).get_table_names()
if name.startswith("dataflow_")
},
)
finally:
engine.dispose()
if __name__ == "__main__":
unittest.main()

198
tests/test_service.py Normal file
View File

@@ -0,0 +1,198 @@
from __future__ import annotations
import unittest
from sqlalchemy import create_engine, func, select
from sqlalchemy.orm import Session, sessionmaker
from govoplan_core.db.base import Base
from govoplan_dataflow.backend.db.models import (
DataflowPipeline,
DataflowPipelineRevision,
DataflowRun,
)
from govoplan_dataflow.backend.schemas import (
GraphEdge,
GraphNode,
GraphPosition,
PipelineCreateRequest,
PipelineGraph,
PipelinePreviewRequest,
PipelineUpdateRequest,
)
from govoplan_dataflow.backend.service import (
DataflowConflictError,
DataflowNotFoundError,
create_pipeline,
get_pipeline,
preview_pipeline,
update_pipeline,
)
def sample_graph(*, minimum: int = 10) -> PipelineGraph:
return PipelineGraph(
nodes=[
GraphNode(
id="source",
type="source.inline",
label="Monthly input",
position=GraphPosition(x=40, y=160),
config={
"source_name": "monthly_files",
"rows": [
{"id": 1, "amount": 5},
{"id": 2, "amount": 15},
{"id": 3, "amount": 25},
],
},
),
GraphNode(
id="filter",
type="filter",
label="Minimum amount",
position=GraphPosition(x=260, y=160),
config={"column": "amount", "operator": "gte", "value": minimum},
),
GraphNode(
id="output",
type="output",
label="Output",
position=GraphPosition(x=480, y=160),
config={},
),
],
edges=[
GraphEdge(id="source-filter", source="source", target="filter"),
GraphEdge(id="filter-output", source="filter", target="output"),
],
)
class DataflowServiceTests(unittest.TestCase):
def setUp(self) -> None:
self.engine = create_engine("sqlite:///:memory:")
Base.metadata.create_all(
self.engine,
tables=[
DataflowPipeline.__table__,
DataflowPipelineRevision.__table__,
DataflowRun.__table__,
],
)
self.Session = sessionmaker(bind=self.engine)
self.session: Session = self.Session()
def tearDown(self) -> None:
self.session.close()
Base.metadata.drop_all(
self.engine,
tables=[
DataflowRun.__table__,
DataflowPipelineRevision.__table__,
DataflowPipeline.__table__,
],
)
self.engine.dispose()
def _create(self, *, tenant_id: str = "tenant-1") -> DataflowPipeline:
pipeline = create_pipeline(
self.session,
tenant_id=tenant_id,
actor_id="user-1",
payload=PipelineCreateRequest(
name="Monthly comparison",
description="First governed pipeline",
graph=sample_graph(),
editor_mode="graph",
),
)
self.session.commit()
return pipeline
def test_create_and_update_make_immutable_revisions(self) -> None:
pipeline = self._create()
updated = update_pipeline(
self.session,
tenant_id="tenant-1",
pipeline_id=pipeline.id,
actor_id="user-2",
payload=PipelineUpdateRequest(
name="Monthly comparison",
description="Changed threshold",
graph=sample_graph(minimum=20),
editor_mode="graph",
expected_revision=1,
),
)
self.session.commit()
revisions = list(
self.session.scalars(
select(DataflowPipelineRevision)
.where(DataflowPipelineRevision.pipeline_id == pipeline.id)
.order_by(DataflowPipelineRevision.revision)
)
)
self.assertEqual(2, updated.current_revision)
self.assertEqual([1, 2], [item.revision for item in revisions])
self.assertNotEqual(revisions[0].content_hash, revisions[1].content_hash)
self.assertEqual(10, revisions[0].graph["nodes"][1]["config"]["value"])
self.assertEqual(20, revisions[1].graph["nodes"][1]["config"]["value"])
def test_stale_revision_is_rejected(self) -> None:
pipeline = self._create()
with self.assertRaises(DataflowConflictError):
update_pipeline(
self.session,
tenant_id="tenant-1",
pipeline_id=pipeline.id,
actor_id="user-2",
payload=PipelineUpdateRequest(
name="Stale edit",
graph=sample_graph(minimum=30),
editor_mode="graph",
expected_revision=2,
),
)
def test_pipeline_lookup_is_tenant_isolated(self) -> None:
pipeline = self._create()
with self.assertRaises(DataflowNotFoundError):
get_pipeline(
self.session,
tenant_id="tenant-2",
pipeline_id=pipeline.id,
)
def test_saved_preview_records_lineage_but_not_result_rows(self) -> None:
pipeline = self._create()
response = preview_pipeline(
self.session,
tenant_id="tenant-1",
actor_id="user-1",
payload=PipelinePreviewRequest(pipeline_id=pipeline.id, row_limit=1),
)
self.session.commit()
run = self.session.scalar(select(DataflowRun).where(DataflowRun.id == response.run_id))
self.assertEqual("succeeded", response.status)
self.assertEqual([{"id": 2, "amount": 15}], response.rows)
self.assertEqual(2, response.total_rows)
self.assertTrue(response.truncated)
self.assertEqual(2, run.output_row_count)
self.assertEqual(3, run.input_row_count)
self.assertEqual(1, len(run.source_fingerprints))
self.assertFalse(hasattr(run, "result_rows"))
self.assertEqual(
1,
self.session.scalar(select(func.count()).select_from(DataflowRun)),
)
if __name__ == "__main__":
unittest.main()