ktx/python/ktx-daemon/tests/test_cli.py

430 lines
12 KiB
Python
Raw Permalink Normal View History

2026-05-10 23:12:26 +02:00
from __future__ import annotations
import io
import json
import os
import subprocess
import sys
from pathlib import Path
from typing import Any
ORDERS_SOURCE = {
"name": "orders",
"table": "public.orders",
"grain": ["id"],
"columns": [
{"name": "id", "type": "number"},
{"name": "status", "type": "string"},
{"name": "amount", "type": "number"},
],
"joins": [],
"measures": [{"name": "order_count", "expr": "count(*)"}],
}
def run_daemon_command(
command: str, payload: dict[str, object]
) -> subprocess.CompletedProcess[str]:
env = os.environ.copy()
src_path = str(Path(__file__).resolve().parents[1] / "src")
env["PYTHONPATH"] = src_path + os.pathsep + env.get("PYTHONPATH", "")
return subprocess.run(
2026-05-10 23:51:24 +02:00
[sys.executable, "-m", "ktx_daemon", command],
2026-05-10 23:12:26 +02:00
input=json.dumps(payload),
text=True,
capture_output=True,
check=False,
env=env,
)
def test_semantic_query_command_reads_stdin_and_writes_json() -> None:
result = run_daemon_command(
"semantic-query",
{
"sources": [ORDERS_SOURCE],
"dialect": "postgres",
"query": {
"measures": ["orders.order_count"],
"dimensions": ["orders.status"],
},
},
)
assert result.returncode == 0, result.stderr
parsed = json.loads(result.stdout)
assert "public.orders" in parsed["sql"]
assert parsed["columns"][0]["name"] == "orders.status"
def test_semantic_validate_command_reads_stdin_and_writes_json() -> None:
result = run_daemon_command(
"semantic-validate",
{"sources": [ORDERS_SOURCE], "dialect": "postgres"},
)
assert result.returncode == 0, result.stderr
parsed = json.loads(result.stdout)
assert parsed == {
"valid": True,
"errors": [],
"warnings": [],
"per_source_warnings": {},
}
def test_command_returns_nonzero_for_invalid_json() -> None:
env = os.environ.copy()
src_path = str(Path(__file__).resolve().parents[1] / "src")
env["PYTHONPATH"] = src_path + os.pathsep + env.get("PYTHONPATH", "")
result = subprocess.run(
2026-05-10 23:51:24 +02:00
[sys.executable, "-m", "ktx_daemon", "semantic-query"],
2026-05-10 23:12:26 +02:00
input="{",
text=True,
capture_output=True,
check=False,
env=env,
)
assert result.returncode == 1
assert "Expecting property name enclosed in double quotes" in result.stderr
def test_serve_http_command_starts_uvicorn_without_reading_stdin(
monkeypatch,
) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
2026-05-10 23:12:26 +02:00
calls: list[dict[str, object]] = []
class FailingStdin:
def read(self) -> str:
raise AssertionError("serve-http must not read stdin JSON")
def fake_run_http_server(
*,
host: str,
port: int,
log_level: str,
enable_code_execution: bool,
) -> None:
calls.append(
{
"host": host,
"port": port,
"log_level": log_level,
"enable_code_execution": enable_code_execution,
}
)
monkeypatch.setattr(sys, "stdin", FailingStdin())
monkeypatch.setattr(daemon_main, "run_http_server", fake_run_http_server)
assert (
daemon_main.main(
[
"serve-http",
"--host",
"127.0.0.1",
"--port",
"9191",
"--log-level",
"warning",
]
)
== 0
)
assert calls == [
{
"host": "127.0.0.1",
"port": 9191,
"log_level": "warning",
"enable_code_execution": False,
}
]
def test_serve_http_command_defaults_to_loopback(monkeypatch) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
2026-05-10 23:12:26 +02:00
calls: list[dict[str, object]] = []
def fake_run_http_server(
*,
host: str,
port: int,
log_level: str,
enable_code_execution: bool,
) -> None:
calls.append(
{
"host": host,
"port": port,
"log_level": log_level,
"enable_code_execution": enable_code_execution,
}
)
monkeypatch.setattr(daemon_main, "run_http_server", fake_run_http_server)
assert daemon_main.main(["serve-http"]) == 0
assert calls == [
{
"host": "127.0.0.1",
"port": 8765,
"log_level": "info",
"enable_code_execution": False,
}
]
def test_serve_http_command_can_enable_code_execution(monkeypatch) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
2026-05-10 23:12:26 +02:00
calls: list[dict[str, object]] = []
def fake_run_http_server(
*,
host: str,
port: int,
log_level: str,
enable_code_execution: bool,
) -> None:
calls.append(
{
"host": host,
"port": port,
"log_level": log_level,
"enable_code_execution": enable_code_execution,
}
)
monkeypatch.setattr(daemon_main, "run_http_server", fake_run_http_server)
assert daemon_main.main(["serve-http", "--enable-code-execution"]) == 0
assert calls == [
{
"host": "127.0.0.1",
"port": 8765,
"log_level": "info",
"enable_code_execution": True,
}
]
def test_lookml_parse_command_reads_stdin_and_writes_json() -> None:
result = run_daemon_command(
"lookml-parse",
{
"files": [
{
"path": "views/orders.view.lkml",
"content": """
view: orders {
sql_table_name: public.orders ;;
dimension: id {
primary_key: yes
type: number
sql: ${TABLE}.id ;;
}
measure: order_count {
type: count
}
}
""",
}
],
"dialect": "postgres",
},
)
assert result.returncode == 0, result.stderr
parsed = json.loads(result.stdout)
assert parsed["views"][0]["name"] == "orders"
assert parsed["views"][0]["table_ref"] == "public.orders"
assert parsed["views"][0]["measures"][0]["expr"] == "count(*)"
assert parsed["joins"] == []
assert parsed["skipped_views"] == []
assert parsed["warnings"] == []
def test_semantic_generate_sources_command_reads_stdin_and_writes_json() -> None:
result = run_daemon_command(
"semantic-generate-sources",
{
"tables": [
{
"name": "orders",
"db": "public",
"columns": [
{"name": "id", "type": "integer", "primary_key": True},
{"name": "amount", "type": "decimal"},
],
}
],
"links": [],
"dialect": "postgres",
},
)
assert result.returncode == 0, result.stderr
parsed = json.loads(result.stdout)
assert parsed["source_count"] == 1
assert parsed["sources"][0]["name"] == "orders"
assert parsed["sources"][0]["table"] == "public.orders"
assert parsed["sources"][0]["measures"] == [
{
"name": "record_count",
"expr": "count(id)",
"segments": [],
"description": "Count of orders records",
},
{
"name": "total_amount",
"expr": "sum(amount)",
"segments": [],
"description": "Sum of amount",
},
{
"name": "avg_amount",
"expr": "avg(amount)",
"segments": [],
"description": "Average of amount",
},
]
def test_database_introspect_command_reads_stdin_and_writes_json(
monkeypatch, capsys
) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
from ktx_daemon.database_introspection import (
2026-05-10 23:12:26 +02:00
DatabaseIntrospectionResponse,
LiveDatabaseColumn,
LiveDatabaseTable,
)
def fake_introspect(request):
assert request.connection_id == "warehouse"
assert request.driver == "postgres"
assert request.schemas == ["public"]
fix(snowflake): unblock multi-schema ingest and relationship discovery (#204) * feat(setup): drop redundant Snowflake schema prompt; fall back to free-text on listSchemas failure Snowflake setup previously asked for a single schema as free text, then ran a multiselect against the discovered schemas — two schema questions back-to-back, with the first being only a session bootstrap. The SDK's `schema` is optional, so the bootstrap step is unnecessary. - Remove the free-text Snowflake schema prompt; only pass `schema` to snowflake-sdk when one is configured. - When `listSchemas()` fails (e.g. role lacks SHOW SCHEMAS), prompt the user for a comma-separated list, persist it as `schema_names`, and use it as both the table-list filter and the multiselect default. Applies to every driver with a scope-discovery spec, not just Snowflake. - Update docs to lead with `schema_names`; keep `schema_name` as a documented single-schema shorthand. * fix(snowflake): keep introspecting when primary-key discovery is denied The PK query joins INFORMATION_SCHEMA.TABLE_CONSTRAINTS and INFORMATION_SCHEMA.KEY_COLUMN_USAGE, which require grants the connection role may not have. Previously a 'SQL compilation error: Object ANALYTICS.INFORMATION_SCHEMA.KEY_COLUMN_USAGE does not exist or not authorized' aborted the entire introspect — schemas, columns, and row counts were all discarded over a missing nice-to-have. Wrap the constraint query in try/catch, log a one-line warning per schema, and return an empty PK map. Columns end up with primaryKey=false; relationship inference still has FK and profiling to fall back on. * fix(scan): unblock relationship discovery on Snowflake Two adjacent bugs prevented the scan's relationship pipeline from producing any joins on a Snowflake warehouse: - relationship-profiling.ts fell through to a default `GROUP_CONCAT` branch for unknown drivers. Snowflake has no GROUP_CONCAT, so every per-table profile query failed with "Unknown function GROUP_CONCAT". Add an explicit Snowflake branch that uses LISTAGG with a literal '\x1f' delimiter (Snowflake requires the delimiter to be a constant, so CHR(31) is rejected). - description-generation.ts destructured `connector.sampleTable` and `connector.sampleColumn` into bare locals, losing the `this` binding when the class-method connectors (Snowflake, Postgres, MySQL) were invoked. Every sample call threw "Cannot read properties of undefined (reading 'assertConnection')" and degraded LLM descriptions to metadata-only prompts. Call the methods through the connector instead. Without these, even after the primary-key probe is allowed to fail softly, the scan ends up with 0 validated relationships and an empty `joins:` block in every shard YAML. * test(scan): cover table-ref helpers * feat(scan): plumb tableScope through live-database introspection port * feat(scan): apply tableScope during metadata fetch * feat(scan): enforce table scope at fetch boundary * feat(scan): pool Snowflake sessions and batch enrichment for faster ingest (#206) * feat(cli): add RSA key-pair auth option to Snowflake setup wizard Extends the interactive Snowflake setup flow with an authentication-method prompt (password vs RSA/JWT key-pair). The RSA branch collects a private-key path (env/file/absolute) and an optional passphrase; the resulting connection config records `authMethod: 'rsa'` with `privateKey` and `passphrase` instead of `password`. * feat(scan): pool Snowflake sessions * fix(scan): reuse structural snapshots and cleanup connectors * feat(scan): parallelize relationship profiling * feat(scan): batch table description generation * docs: document Snowflake ingest concurrency knobs * fix(scan): close Snowflake ingest perf verification gaps * fix(scan): keep batched description failure bounded * feat(scan): dispatch query-history probes by connection driver Extract historic-sql dialect resolution into a shared helper so the status-project readiness check and the local ingest factory agree on which connections enable query history and which probe to run. The status command now picks the postgres/snowflake/bigquery probe based on the connection's driver instead of always reporting against postgres, which previously caused snowflake connections with queryHistory.enabled to surface a misleading "driver is snowflake" failure. Also drops a noisy console.warn from Snowflake primary-key discovery — INFORMATION_SCHEMA.KEY_COLUMN_USAGE is commonly ungranted for read-only roles and the FK + profiling paths handle the empty PK map already. * fix(llm): allow StructuredOutput tool and raise maxTurns for generateObject The Claude Code agent SDK announces an internal pseudo-tool named StructuredOutput in the system/init message whenever outputFormat is set to { type: 'json_schema' }. The runtime's isolation check built its allowedToolIds set only from MCP tool ids and treated StructuredOutput as an unexpected host-injected tool, so every generateObject call threw "Claude Code runtime isolation failed: tools=StructuredOutput ..." and the table-descriptions and relationship-LLM-proposal enrichment stages recorded null output across the board. Whitelist StructuredOutput specifically in generateObject's allowedToolIds — the check also enforces missing_tools symmetry, so generateText and runAgentLoop, which do not see StructuredOutput, must not require it. generateObject also ran with maxTurns: 1, which the model intermittently breached when it emitted thinking text before the structured response. Raised to 5 to give the schema-bound call enough headroom without allowing unbounded loops. The existing tests now exercise the path with an init message that announces StructuredOutput so the regression cannot slip back in. * chore(scripts): add ktx-reset.sh project-cleanup helper Convenience script for repeatable ingest testing: takes a project directory and prunes everything except ktx.yaml and .ktx/secrets/, so the next ktx setup or ktx ingest run starts from a known-clean state.
2026-05-23 10:41:30 +02:00
assert request.table_scope is not None
assert request.table_scope[0].db == "public"
assert request.table_scope[0].name == "orders"
2026-05-10 23:12:26 +02:00
return DatabaseIntrospectionResponse(
connection_id="warehouse",
extracted_at="2026-04-28T10:00:00+00:00",
metadata={"driver": "postgres", "schemas": ["public"]},
tables=[
LiveDatabaseTable(
catalog="warehouse",
db="public",
name="orders",
columns=[
LiveDatabaseColumn(
name="id",
type="integer",
nullable=False,
primary_key=True,
)
],
)
],
)
monkeypatch.setattr(daemon_main, "introspect_database_response", fake_introspect)
monkeypatch.setattr(
sys,
"stdin",
io.StringIO(
fix(snowflake): unblock multi-schema ingest and relationship discovery (#204) * feat(setup): drop redundant Snowflake schema prompt; fall back to free-text on listSchemas failure Snowflake setup previously asked for a single schema as free text, then ran a multiselect against the discovered schemas — two schema questions back-to-back, with the first being only a session bootstrap. The SDK's `schema` is optional, so the bootstrap step is unnecessary. - Remove the free-text Snowflake schema prompt; only pass `schema` to snowflake-sdk when one is configured. - When `listSchemas()` fails (e.g. role lacks SHOW SCHEMAS), prompt the user for a comma-separated list, persist it as `schema_names`, and use it as both the table-list filter and the multiselect default. Applies to every driver with a scope-discovery spec, not just Snowflake. - Update docs to lead with `schema_names`; keep `schema_name` as a documented single-schema shorthand. * fix(snowflake): keep introspecting when primary-key discovery is denied The PK query joins INFORMATION_SCHEMA.TABLE_CONSTRAINTS and INFORMATION_SCHEMA.KEY_COLUMN_USAGE, which require grants the connection role may not have. Previously a 'SQL compilation error: Object ANALYTICS.INFORMATION_SCHEMA.KEY_COLUMN_USAGE does not exist or not authorized' aborted the entire introspect — schemas, columns, and row counts were all discarded over a missing nice-to-have. Wrap the constraint query in try/catch, log a one-line warning per schema, and return an empty PK map. Columns end up with primaryKey=false; relationship inference still has FK and profiling to fall back on. * fix(scan): unblock relationship discovery on Snowflake Two adjacent bugs prevented the scan's relationship pipeline from producing any joins on a Snowflake warehouse: - relationship-profiling.ts fell through to a default `GROUP_CONCAT` branch for unknown drivers. Snowflake has no GROUP_CONCAT, so every per-table profile query failed with "Unknown function GROUP_CONCAT". Add an explicit Snowflake branch that uses LISTAGG with a literal '\x1f' delimiter (Snowflake requires the delimiter to be a constant, so CHR(31) is rejected). - description-generation.ts destructured `connector.sampleTable` and `connector.sampleColumn` into bare locals, losing the `this` binding when the class-method connectors (Snowflake, Postgres, MySQL) were invoked. Every sample call threw "Cannot read properties of undefined (reading 'assertConnection')" and degraded LLM descriptions to metadata-only prompts. Call the methods through the connector instead. Without these, even after the primary-key probe is allowed to fail softly, the scan ends up with 0 validated relationships and an empty `joins:` block in every shard YAML. * test(scan): cover table-ref helpers * feat(scan): plumb tableScope through live-database introspection port * feat(scan): apply tableScope during metadata fetch * feat(scan): enforce table scope at fetch boundary * feat(scan): pool Snowflake sessions and batch enrichment for faster ingest (#206) * feat(cli): add RSA key-pair auth option to Snowflake setup wizard Extends the interactive Snowflake setup flow with an authentication-method prompt (password vs RSA/JWT key-pair). The RSA branch collects a private-key path (env/file/absolute) and an optional passphrase; the resulting connection config records `authMethod: 'rsa'` with `privateKey` and `passphrase` instead of `password`. * feat(scan): pool Snowflake sessions * fix(scan): reuse structural snapshots and cleanup connectors * feat(scan): parallelize relationship profiling * feat(scan): batch table description generation * docs: document Snowflake ingest concurrency knobs * fix(scan): close Snowflake ingest perf verification gaps * fix(scan): keep batched description failure bounded * feat(scan): dispatch query-history probes by connection driver Extract historic-sql dialect resolution into a shared helper so the status-project readiness check and the local ingest factory agree on which connections enable query history and which probe to run. The status command now picks the postgres/snowflake/bigquery probe based on the connection's driver instead of always reporting against postgres, which previously caused snowflake connections with queryHistory.enabled to surface a misleading "driver is snowflake" failure. Also drops a noisy console.warn from Snowflake primary-key discovery — INFORMATION_SCHEMA.KEY_COLUMN_USAGE is commonly ungranted for read-only roles and the FK + profiling paths handle the empty PK map already. * fix(llm): allow StructuredOutput tool and raise maxTurns for generateObject The Claude Code agent SDK announces an internal pseudo-tool named StructuredOutput in the system/init message whenever outputFormat is set to { type: 'json_schema' }. The runtime's isolation check built its allowedToolIds set only from MCP tool ids and treated StructuredOutput as an unexpected host-injected tool, so every generateObject call threw "Claude Code runtime isolation failed: tools=StructuredOutput ..." and the table-descriptions and relationship-LLM-proposal enrichment stages recorded null output across the board. Whitelist StructuredOutput specifically in generateObject's allowedToolIds — the check also enforces missing_tools symmetry, so generateText and runAgentLoop, which do not see StructuredOutput, must not require it. generateObject also ran with maxTurns: 1, which the model intermittently breached when it emitted thinking text before the structured response. Raised to 5 to give the schema-bound call enough headroom without allowing unbounded loops. The existing tests now exercise the path with an init message that announces StructuredOutput so the regression cannot slip back in. * chore(scripts): add ktx-reset.sh project-cleanup helper Convenience script for repeatable ingest testing: takes a project directory and prunes everything except ktx.yaml and .ktx/secrets/, so the next ktx setup or ktx ingest run starts from a known-clean state.
2026-05-23 10:41:30 +02:00
'{"connection_id":"warehouse","driver":"postgres","url":"postgresql://readonly@example.test/warehouse","schemas":["public"],"table_scope":[{"db":"public","name":"orders"}]}'
2026-05-10 23:12:26 +02:00
),
)
assert daemon_main.main(["database-introspect"]) == 0
captured = capsys.readouterr()
parsed = json.loads(captured.out)
assert parsed["connection_id"] == "warehouse"
assert parsed["metadata"] == {"driver": "postgres", "schemas": ["public"]}
assert parsed["tables"][0]["name"] == "orders"
assert captured.err == ""
def test_embedding_compute_command_reads_stdin_and_writes_json(
monkeypatch, capsys
) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
from ktx_daemon.embeddings import ComputeEmbeddingResponse
2026-05-10 23:12:26 +02:00
def fake_compute(request):
assert request.text == "hello"
return ComputeEmbeddingResponse(embedding=[1.0, 2.0, 3.0])
monkeypatch.setattr(daemon_main, "compute_embedding_response", fake_compute)
monkeypatch.setattr(sys, "stdin", io.StringIO('{"text": "hello"}'))
assert daemon_main.main(["embedding-compute"]) == 0
captured = capsys.readouterr()
assert json.loads(captured.out) == {"embedding": [1.0, 2.0, 3.0]}
assert captured.err == ""
def test_embedding_compute_bulk_command_reads_stdin_and_writes_json(
monkeypatch, capsys
) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
from ktx_daemon.embeddings import ComputeEmbeddingBulkResponse
2026-05-10 23:12:26 +02:00
def fake_compute(request):
assert request.texts == ["hello", "world"]
return ComputeEmbeddingBulkResponse(embeddings=[[1.0, 2.0], [3.0, 4.0]])
monkeypatch.setattr(daemon_main, "compute_embedding_bulk_response", fake_compute)
monkeypatch.setattr(sys, "stdin", io.StringIO('{"texts": ["hello", "world"]}'))
assert daemon_main.main(["embedding-compute-bulk"]) == 0
captured = capsys.readouterr()
assert json.loads(captured.out) == {"embeddings": [[1.0, 2.0], [3.0, 4.0]]}
assert captured.err == ""
def test_code_execute_command_reads_stdin_and_writes_json(monkeypatch, capsys) -> None:
2026-05-10 23:51:24 +02:00
from ktx_daemon import __main__ as daemon_main
from ktx_daemon.code_execution import ExecuteCodeResponse
2026-05-10 23:12:26 +02:00
calls: list[dict[str, Any]] = []
def fake_execute(request, *, nest_api_url, auth_header):
calls.append(
{
"request": request,
"nest_api_url": nest_api_url,
"auth_header": auth_header,
}
)
return ExecuteCodeResponse(
formatted_result="\n\n=== Result ===\n\n7",
result=7,
)
monkeypatch.setattr(daemon_main, "execute_code_response", fake_execute)
monkeypatch.setattr(sys, "stdin", io.StringIO('{"code": "result = 7"}'))
assert daemon_main.main(["code-execute"]) == 0
captured = capsys.readouterr()
assert json.loads(captured.out) == {
"formatted_result": "\n\n=== Result ===\n\n7",
"result": 7,
"console_output": None,
"error": None,
"message": None,
"visualizations": None,
}
assert captured.err == ""
assert calls[0]["request"].code == "result = 7"
assert calls[0]["nest_api_url"] is None
assert calls[0]["auth_header"] is None