sync update from LAN repo (source dhk-1 @ 3e87be55)

This commit is contained in:
gozeros committed 2026-10-21 15:06:08 +08:00
1 parent ab6068a7c1
commit 42f2a26fa8
1482 files changed
+227759 -9584

No files matched your search

@@ -0,0 +1,191 @@
from __future__ import annotations
from pathlib import Path
import pytest
import sqlalchemy as sa
from alembic import command
from alembic.config import Config
_EDGE_INDEXES = {
"ix_llmwiki_wiki_edges_knowledge_base_mapping_id",
"ix_llmwiki_wiki_edges_source_page_id",
"ix_llmwiki_wiki_edges_target_page_id",
"ix_llmwiki_wiki_edge_mapping_source",
"ix_llmwiki_wiki_edge_mapping_target",
}
def test_mysql_manual_column_script_keeps_helper_until_all_calls_finish() -> None:
sql_path = Path(__file__).parents[1] / "scripts" / "sql" / "column_additions_mysql.sql"
sql = sql_path.read_text(encoding="utf-8")
drop_offset = sql.index("DROP PROCEDURE IF EXISTS _add_column_if_missing;", sql.index("CREATE PROCEDURE"))
call_offsets = [
offset
for offset in range(len(sql))
if sql.startswith("CALL _add_column_if_missing(", offset)
]
assert call_offsets
assert max(call_offsets) < drop_offset
assert (
"CALL _add_column_if_missing('llmwiki_wiki_sync_states', "
"'last_sync_result', 'TEXT NULL');"
) in sql
assert (
"CALL _add_column_if_missing('llmwiki_knowledge_bases', "
"'test_questions', 'TEXT NULL');"
) in sql
assert (
"CALL _add_column_if_missing('llmwiki_knowledge_bases', "
"'retrieval_settings', 'TEXT NULL');"
) in sql
def _config(database: Path) -> Config:
migrations = (
Path(__file__).parents[1]
/ "packages"
/ "harness"
/ "deerflow"
/ "persistence"
/ "migrations"
)
config = Config(str(migrations / "alembic.ini"))
config.set_main_option(
"sqlalchemy.url",
f"sqlite+aiosqlite:///{database.resolve().as_posix()}",
)
return config
def _prepare_old_database(
database: Path,
stamp: str,
*,
has_sync_result: bool,
has_retrieval_profile: bool,
) -> None:
engine = sa.create_engine(f"sqlite:///{database.resolve().as_posix()}")
sync_result = ", last_sync_result JSON" if has_sync_result else ""
retrieval_profile = (
", test_questions JSON, retrieval_settings JSON"
if has_retrieval_profile
else ""
)
with engine.begin() as connection:
connection.exec_driver_sql(
"CREATE TABLE llmwiki_knowledge_bases "
f"(id VARCHAR(36) PRIMARY KEY{retrieval_profile})"
)
if has_retrieval_profile:
connection.exec_driver_sql(
"INSERT INTO llmwiki_knowledge_bases "
"(id, test_questions, retrieval_settings) VALUES "
"('mapping-1', '[\"保留问题\"]', '{\"top_k\": 8}')"
)
connection.exec_driver_sql(
"CREATE TABLE llmwiki_wiki_pages (id VARCHAR(36) PRIMARY KEY)"
)
connection.exec_driver_sql(
"CREATE TABLE llmwiki_wiki_sync_states "
f"(id VARCHAR(36) PRIMARY KEY{sync_result})"
)
if has_sync_result:
connection.exec_driver_sql(
"INSERT INTO llmwiki_wiki_sync_states (id, last_sync_result) "
"VALUES ('state-1', '{\"processed\": 3}')"
)
connection.exec_driver_sql(
"CREATE TABLE alembic_version "
"(version_num VARCHAR(32) NOT NULL PRIMARY KEY)"
)
connection.exec_driver_sql(
"INSERT INTO alembic_version (version_num) VALUES (?)",
(stamp,),
)
engine.dispose()
@pytest.mark.parametrize(
("stamp", "has_sync_result", "has_retrieval_profile"),
[
("20260916_01", False, False),
("20260916_02", False, False),
("20260920_01", True, False),
("20260920_01", False, True),
],
)
def test_duplicate_historical_stamps_reconcile_to_single_head(
tmp_path: Path,
stamp: str,
has_sync_result: bool,
has_retrieval_profile: bool,
) -> None:
variant = "sync" if has_sync_result else "profile" if has_retrieval_profile else "base"
database = tmp_path / f"old-{stamp}-{variant}.db"
_prepare_old_database(
database,
stamp,
has_sync_result=has_sync_result,
has_retrieval_profile=has_retrieval_profile,
)
config = _config(database)
command.upgrade(config, "head")
command.upgrade(config, "head")
engine = sa.create_engine(f"sqlite:///{database.resolve().as_posix()}")
with engine.connect() as connection:
inspector = sa.inspect(connection)
assert connection.execute(
sa.text("SELECT version_num FROM alembic_version")
).scalar_one() == "20261020_01"
assert inspector.has_table("task_deeplink_threads")
assert connection.execute(
sa.text("SELECT COUNT(*) FROM task_deeplink_threads")
).scalar_one() == 0
assert inspector.has_table("task_deeplink_details")
assert inspector.has_table("llmwiki_wiki_edges")
assert _EDGE_INDEXES <= {
str(item["name"])
for item in inspector.get_indexes("llmwiki_wiki_edges")
}
assert "uq_llmwiki_wiki_edge_mapping_pages_type" in {
str(item["name"])
for item in inspector.get_unique_constraints("llmwiki_wiki_edges")
}
assert len(inspector.get_foreign_keys("llmwiki_wiki_edges")) == 3
mapping_columns = {
str(item["name"])
for item in inspector.get_columns("llmwiki_knowledge_bases")
}
assert {"test_questions", "retrieval_settings"} <= mapping_columns
if has_retrieval_profile:
assert connection.execute(
sa.text(
"SELECT test_questions FROM llmwiki_knowledge_bases "
"WHERE id = 'mapping-1'"
)
).scalar_one() == '["保留问题"]'
assert connection.execute(
sa.text(
"SELECT retrieval_settings FROM llmwiki_knowledge_bases "
"WHERE id = 'mapping-1'"
)
).scalar_one() == '{"top_k": 8}'
sync_columns = {
str(item["name"])
for item in inspector.get_columns("llmwiki_wiki_sync_states")
}
assert "last_sync_result" in sync_columns
if has_sync_result:
assert connection.execute(
sa.text(
"SELECT last_sync_result FROM llmwiki_wiki_sync_states "
"WHERE id = 'state-1'"
)
).scalar_one() == '{"processed": 3}'
engine.dispose()