在准备环境前提交次全部更改。
This commit is contained in:
1971
tests/test_dataflow_analyzer.py
Normal file
1971
tests/test_dataflow_analyzer.py
Normal file
File diff suppressed because it is too large
Load Diff
572
tests/test_dwd_phase1_properties.py
Normal file
572
tests/test_dwd_phase1_properties.py
Normal file
@@ -0,0 +1,572 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
DWD 第一阶段重构 — 属性测试
|
||||
|
||||
Feature: dwd-phase1-refactor
|
||||
测试位置：monorepo 根目录 tests/
|
||||
|
||||
使用 hypothesis + FakeCursor 验证重构后 DwdLoadTask 的核心正确性属性。
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import re
|
||||
import sys
|
||||
import logging
|
||||
from datetime import datetime, timedelta, timezone
|
||||
from pathlib import Path
|
||||
from types import SimpleNamespace
|
||||
from typing import Any, Dict, List, Sequence, Tuple
|
||||
from unittest.mock import MagicMock
|
||||
|
||||
from hypothesis import given, settings, assume
|
||||
import hypothesis.strategies as st
|
||||
|
||||
# ── 将 ETL 模块加入 sys.path ──
|
||||
_ETL_ROOT = Path(__file__).resolve().parent.parent / "apps" / "etl" / "connectors" / "feiqiu"
|
||||
if str(_ETL_ROOT) not in sys.path:
|
||||
sys.path.insert(0, str(_ETL_ROOT))
|
||||
|
||||
from tasks.dwd.dwd_load_task import DwdLoadTask
|
||||
|
||||
|
||||
# ── 测试用 FakeCursor（扩展自 task_test_utils，增加 fetchmany 支持）──
|
||||
|
||||
class PropertyTestCursor:
|
||||
"""捕获 SQL 语句的极简游标，用于属性测试验证 SQL 结构。"""
|
||||
|
||||
def __init__(self):
|
||||
self.executed: List[Dict[str, Any]] = []
|
||||
self._fetchall_rows: List[Tuple] = []
|
||||
self.rowcount = 0
|
||||
self.connection = SimpleNamespace(encoding="UTF8")
|
||||
|
||||
def execute(self, sql: str, params=None):
|
||||
sql_text = sql.decode("utf-8", errors="ignore") if isinstance(sql, (bytes, bytearray)) else str(sql)
|
||||
self.executed.append({"sql": sql_text.strip(), "params": params})
|
||||
self._fetchall_rows = []
|
||||
|
||||
lowered = sql_text.lower()
|
||||
# information_schema 查询返回假列信息
|
||||
if "from information_schema.columns" in lowered:
|
||||
table_name = params[1] if params and len(params) >= 2 else None
|
||||
self._fetchall_rows = self._fake_columns(table_name)
|
||||
return
|
||||
if "from information_schema.table_constraints" in lowered:
|
||||
self._fetchall_rows = []
|
||||
return
|
||||
self.rowcount = 0
|
||||
|
||||
def fetchone(self):
|
||||
return None
|
||||
|
||||
def fetchall(self):
|
||||
return list(self._fetchall_rows)
|
||||
|
||||
def fetchmany(self, size=None):
|
||||
"""_merge_fact_increment 使用 fetchmany 读取 RETURNING 结果。"""
|
||||
# 返回空列表表示无更多行，终止循环
|
||||
return []
|
||||
|
||||
def mogrify(self, template, args):
|
||||
return b"(?)"
|
||||
|
||||
@staticmethod
|
||||
def _fake_columns(table_name: str | None) -> List[Tuple[str, str, str]]:
|
||||
"""返回假列信息，确保包含 fetched_at 和常见列。"""
|
||||
return [
|
||||
("id", "bigint", "int8"),
|
||||
("site_id", "bigint", "int8"),
|
||||
("record_index", "integer", "int4"),
|
||||
("content_hash", "text", "text"),
|
||||
("fetched_at", "timestamp with time zone", "timestamptz"),
|
||||
("payload", "jsonb", "jsonb"),
|
||||
]
|
||||
|
||||
def __enter__(self):
|
||||
return self
|
||||
|
||||
def __exit__(self, exc_type, exc, tb):
|
||||
return False
|
||||
|
||||
|
||||
# ── 辅助：构造最小可用的 DwdLoadTask 实例 ──
|
||||
|
||||
def _make_task(fact_upsert: bool = True) -> DwdLoadTask:
|
||||
"""构造一个用于测试的 DwdLoadTask，使用 mock config/db/api/logger。"""
|
||||
config = MagicMock()
|
||||
config.get = lambda key, default=None: {
|
||||
"app.store_id": 1,
|
||||
"app.timezone": "Asia/Shanghai",
|
||||
"dwd.fact_upsert": fact_upsert,
|
||||
}.get(key, default)
|
||||
|
||||
db = MagicMock()
|
||||
api = MagicMock()
|
||||
logger = logging.getLogger("test_dwd_property")
|
||||
logger.setLevel(logging.WARNING) # 抑制测试噪音
|
||||
|
||||
return DwdLoadTask(config, db, api, logger)
|
||||
|
||||
|
||||
# ── Hypothesis 策略 ──
|
||||
|
||||
@st.composite
|
||||
def window_pair(draw):
|
||||
"""生成有效的 (window_start, window_end) 时间对，确保 start < end。"""
|
||||
base = draw(st.datetimes(
|
||||
min_value=datetime(2020, 1, 1),
|
||||
max_value=datetime(2030, 12, 31),
|
||||
timezones=st.just(timezone.utc),
|
||||
))
|
||||
delta_minutes = draw(st.integers(min_value=1, max_value=10080)) # 1 分钟 ~ 7 天
|
||||
end = base + timedelta(minutes=delta_minutes)
|
||||
return base, end
|
||||
|
||||
|
||||
# 从 TABLE_MAP 中提取所有事实表（dwd_ 前缀）
|
||||
_FACT_TABLES = [
|
||||
(dwd, ods)
|
||||
for dwd, ods in DwdLoadTask.TABLE_MAP.items()
|
||||
if dwd.split(".")[-1].startswith("dwd_")
|
||||
]
|
||||
|
||||
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
# Property 1: 事实表增量 SQL 始终使用窗口范围条件
|
||||
# Feature: dwd-phase1-refactor, Property 1: 事实表增量 SQL 始终使用窗口范围条件
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
|
||||
# 匹配双边窗口条件：fetched_at >= %s AND fetched_at < %s
|
||||
_WINDOW_PATTERN = re.compile(
|
||||
r'"fetched_at"\s*>=\s*%s\s+AND\s+"fetched_at"\s*<\s*%s',
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
# 匹配单边水位线条件：fetched_at > %s（不跟 AND fetched_at < %s）
|
||||
_SINGLE_WATERMARK_PATTERN = re.compile(
|
||||
r'"fetched_at"\s*>\s*%s(?!\s+AND\s+"fetched_at"\s*<\s*%s)',
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
|
||||
@given(
|
||||
table_pair=st.sampled_from(_FACT_TABLES),
|
||||
window=window_pair(),
|
||||
)
|
||||
@settings(max_examples=100)
|
||||
def test_fact_increment_sql_uses_window_range(
|
||||
table_pair: Tuple[str, str],
|
||||
window: Tuple[datetime, datetime],
|
||||
):
|
||||
"""
|
||||
Property 1: 事实表增量 SQL 始终使用窗口范围条件
|
||||
|
||||
对于任意事实表和任意 window_start/window_end 时间对，
|
||||
_merge_fact_increment() 生成的 SQL 的 WHERE 子句：
|
||||
- 必须包含 fetched_at >= %s AND fetched_at < %s
|
||||
- 不得包含单边水位线条件 fetched_at > %s
|
||||
|
||||
**Validates: Requirements 1.1, 1.2**
|
||||
"""
|
||||
dwd_table, ods_table = table_pair
|
||||
window_start, window_end = window
|
||||
|
||||
task = _make_task()
|
||||
cur = PropertyTestCursor()
|
||||
|
||||
# 获取 FakeCursor 返回的列名（模拟 information_schema 查询结果）
|
||||
fake_cols = [row[0] for row in PropertyTestCursor._fake_columns(None)]
|
||||
|
||||
# 获取该表的 FACT_MAPPINGS 中的目标列
|
||||
mapping_entries = task.FACT_MAPPINGS.get(dwd_table) or []
|
||||
mapping_dest = [dst for dst, _, _ in mapping_entries]
|
||||
|
||||
# 构建 dwd_cols 和 ods_cols（合并映射列和假列）
|
||||
dwd_cols = list(set(fake_cols + mapping_dest))
|
||||
ods_cols = list(fake_cols)
|
||||
|
||||
# 构建类型字典
|
||||
fake_types = {row[0]: row[1] for row in PropertyTestCursor._fake_columns(None)}
|
||||
dwd_types = dict(fake_types)
|
||||
ods_types = dict(fake_types)
|
||||
# 为映射目标列补充类型
|
||||
for col in mapping_dest:
|
||||
if col not in dwd_types:
|
||||
dwd_types[col] = "text"
|
||||
|
||||
result = task._merge_fact_increment(
|
||||
cur=cur,
|
||||
dwd_table=dwd_table,
|
||||
ods_table=ods_table,
|
||||
dwd_cols=dwd_cols,
|
||||
ods_cols=ods_cols,
|
||||
dwd_types=dwd_types,
|
||||
ods_types=ods_types,
|
||||
window_start=window_start,
|
||||
window_end=window_end,
|
||||
)
|
||||
|
||||
# 找到主 INSERT SQL（排除 information_schema 查询）
|
||||
insert_sqls = [
|
||||
entry["sql"]
|
||||
for entry in cur.executed
|
||||
if entry["sql"].upper().startswith("INSERT INTO")
|
||||
]
|
||||
|
||||
assert len(insert_sqls) == 1, (
|
||||
f"期望恰好 1 条 INSERT SQL，实际 {len(insert_sqls)} 条，"
|
||||
f"表: {dwd_table}"
|
||||
)
|
||||
|
||||
sql = insert_sqls[0]
|
||||
|
||||
# 验证：包含双边窗口条件
|
||||
assert _WINDOW_PATTERN.search(sql), (
|
||||
f"INSERT SQL 缺少双边窗口条件 (fetched_at >= %s AND fetched_at < %s)，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 验证：不包含单边水位线条件
|
||||
assert not _SINGLE_WATERMARK_PATTERN.search(sql), (
|
||||
f"INSERT SQL 包含单边水位线条件 (fetched_at > %s)，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 验证：SQL 参数包含 window_start 和 window_end
|
||||
insert_entry = [
|
||||
entry for entry in cur.executed
|
||||
if entry["sql"].upper().startswith("INSERT INTO")
|
||||
][0]
|
||||
params = insert_entry["params"]
|
||||
assert params is not None, f"INSERT SQL 参数为 None，表: {dwd_table}"
|
||||
assert window_start in params, (
|
||||
f"SQL 参数缺少 window_start={window_start}，"
|
||||
f"表: {dwd_table}, params={params}"
|
||||
)
|
||||
assert window_end in params, (
|
||||
f"SQL 参数缺少 window_end={window_end}，"
|
||||
f"表: {dwd_table}, params={params}"
|
||||
)
|
||||
|
||||
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
# Property 2: 事实表增量不执行回补 SQL
|
||||
# Feature: dwd-phase1-refactor, Property 2: 事实表增量不执行回补 SQL
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
|
||||
# 匹配回补 SQL 特征：LEFT JOIN ... WHERE ... IS NULL
|
||||
_BACKFILL_PATTERN = re.compile(
|
||||
r"LEFT\s+JOIN\b.*?\bWHERE\b.*?\bIS\s+NULL\b",
|
||||
re.IGNORECASE | re.DOTALL,
|
||||
)
|
||||
|
||||
|
||||
@given(
|
||||
table_pair=st.sampled_from(_FACT_TABLES),
|
||||
window=window_pair(),
|
||||
)
|
||||
@settings(max_examples=100)
|
||||
def test_fact_increment_no_backfill(
|
||||
table_pair: Tuple[str, str],
|
||||
window: Tuple[datetime, datetime],
|
||||
):
|
||||
"""
|
||||
Property 2: 事实表增量不执行回补 SQL
|
||||
|
||||
对于任意事实表和任意 window_start/window_end 时间对，
|
||||
_merge_fact_increment() 执行的所有 SQL 语句中不得包含
|
||||
LEFT JOIN ... WHERE ... IS NULL 模式（回补特征）。
|
||||
|
||||
**Validates: Requirements 2.3**
|
||||
"""
|
||||
dwd_table, ods_table = table_pair
|
||||
window_start, window_end = window
|
||||
|
||||
task = _make_task()
|
||||
cur = PropertyTestCursor()
|
||||
|
||||
# 构建列信息（复用 Property 1 的逻辑）
|
||||
fake_cols = [row[0] for row in PropertyTestCursor._fake_columns(None)]
|
||||
mapping_entries = task.FACT_MAPPINGS.get(dwd_table) or []
|
||||
mapping_dest = [dst for dst, _, _ in mapping_entries]
|
||||
|
||||
dwd_cols = list(set(fake_cols + mapping_dest))
|
||||
ods_cols = list(fake_cols)
|
||||
|
||||
fake_types = {row[0]: row[1] for row in PropertyTestCursor._fake_columns(None)}
|
||||
dwd_types = dict(fake_types)
|
||||
ods_types = dict(fake_types)
|
||||
for col in mapping_dest:
|
||||
if col not in dwd_types:
|
||||
dwd_types[col] = "text"
|
||||
|
||||
result = task._merge_fact_increment(
|
||||
cur=cur,
|
||||
dwd_table=dwd_table,
|
||||
ods_table=ods_table,
|
||||
dwd_cols=dwd_cols,
|
||||
ods_cols=ods_cols,
|
||||
dwd_types=dwd_types,
|
||||
ods_types=ods_types,
|
||||
window_start=window_start,
|
||||
window_end=window_end,
|
||||
)
|
||||
|
||||
# 检查所有执行的 SQL（排除 information_schema 元数据查询）
|
||||
all_sqls = [
|
||||
entry["sql"]
|
||||
for entry in cur.executed
|
||||
if "information_schema" not in entry["sql"].lower()
|
||||
]
|
||||
|
||||
for sql in all_sqls:
|
||||
assert not _BACKFILL_PATTERN.search(sql), (
|
||||
f"检测到回补 SQL（LEFT JOIN ... IS NULL 模式），"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
# Property 4: 维度表始终走 SCD2 路径
|
||||
# Feature: dwd-phase1-refactor, Property 4: 维度表始终走 SCD2 路径
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
|
||||
# 从 TABLE_MAP 中提取所有维度表（dim_ 前缀）
|
||||
_DIM_TABLES = [
|
||||
(dwd, ods)
|
||||
for dwd, ods in DwdLoadTask.TABLE_MAP.items()
|
||||
if dwd.split(".")[-1].startswith("dim_")
|
||||
]
|
||||
|
||||
|
||||
@given(
|
||||
table_pair=st.sampled_from(_DIM_TABLES),
|
||||
now=st.datetimes(
|
||||
min_value=datetime(2020, 1, 1),
|
||||
max_value=datetime(2030, 12, 31),
|
||||
timezones=st.just(timezone.utc),
|
||||
),
|
||||
)
|
||||
@settings(max_examples=100)
|
||||
def test_dim_always_scd2(
|
||||
table_pair: Tuple[str, str],
|
||||
now: datetime,
|
||||
):
|
||||
"""
|
||||
Property 4: 维度表始终走 SCD2 路径
|
||||
|
||||
对于任意维度表（TABLE_MAP 中 dim_ 前缀），调用 _merge_dim()
|
||||
始终委托给 _merge_dim_scd2()，不经过任何条件分支判断。
|
||||
|
||||
**Validates: Requirements 3.8, 5.1**
|
||||
"""
|
||||
from unittest.mock import patch, sentinel
|
||||
|
||||
dwd_table, ods_table = table_pair
|
||||
|
||||
task = _make_task()
|
||||
cur = PropertyTestCursor()
|
||||
|
||||
# 构建最小列集合
|
||||
fake_cols = [row[0] for row in PropertyTestCursor._fake_columns(None)]
|
||||
dwd_cols = list(fake_cols)
|
||||
ods_cols = list(fake_cols)
|
||||
|
||||
# mock _merge_dim_scd2，返回 sentinel 以验证委托关系
|
||||
expected_return = {"processed": 0, "inserted": 0, "updated": 0, "skipped": 0}
|
||||
|
||||
with patch.object(task, "_merge_dim_scd2", return_value=expected_return) as mock_scd2:
|
||||
result = task._merge_dim(cur, dwd_table, ods_table, dwd_cols, ods_cols, now)
|
||||
|
||||
# 验证 _merge_dim_scd2 被调用了恰好一次
|
||||
assert mock_scd2.call_count == 1, (
|
||||
f"期望 _merge_dim_scd2 被调用 1 次，实际 {mock_scd2.call_count} 次，"
|
||||
f"表: {dwd_table}"
|
||||
)
|
||||
|
||||
# 验证传入参数正确
|
||||
call_args = mock_scd2.call_args
|
||||
assert call_args[0] == (cur, dwd_table, ods_table, dwd_cols, ods_cols, now), (
|
||||
f"_merge_dim_scd2 调用参数不匹配，表: {dwd_table}\n"
|
||||
f"期望: (cur, {dwd_table}, {ods_table}, {dwd_cols}, {ods_cols}, {now})\n"
|
||||
f"实际: {call_args[0]}"
|
||||
)
|
||||
|
||||
# 验证返回值直接透传
|
||||
assert result is expected_return, (
|
||||
f"_merge_dim() 返回值未直接透传 _merge_dim_scd2() 的结果，"
|
||||
f"表: {dwd_table}"
|
||||
)
|
||||
|
||||
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
# Property 3: 事实表主增量 SQL 结构等价性
|
||||
# Feature: dwd-phase1-refactor, Property 3: 事实表主增量 SQL 结构等价性
|
||||
# ══════════════════════════════════════════════════════════════
|
||||
|
||||
|
||||
class _PkAwareCursor(PropertyTestCursor):
|
||||
"""扩展 PropertyTestCursor，对 table_constraints 查询返回主键信息。
|
||||
|
||||
_merge_fact_increment 通过 _get_primary_keys 查询 information_schema
|
||||
获取主键列。基础 PropertyTestCursor 返回空列表，导致无法生成
|
||||
ON CONFLICT 子句。此子类注入 pk_cols 以完整测试 SQL 结构。
|
||||
"""
|
||||
|
||||
def __init__(self, pk_cols: Sequence[str]):
|
||||
super().__init__()
|
||||
self._pk_cols = list(pk_cols)
|
||||
|
||||
def execute(self, sql: str, params=None):
|
||||
sql_text = sql.decode("utf-8", errors="ignore") if isinstance(sql, (bytes, bytearray)) else str(sql)
|
||||
lowered = sql_text.lower()
|
||||
# 拦截主键查询，返回注入的 pk_cols
|
||||
if "from information_schema.table_constraints" in lowered:
|
||||
self.executed.append({"sql": sql_text.strip(), "params": params})
|
||||
self._fetchall_rows = [
|
||||
{"column_name": col} for col in self._pk_cols
|
||||
]
|
||||
return
|
||||
super().execute(sql, params)
|
||||
|
||||
|
||||
# INSERT INTO "dwd"."<table>" (...) — 实际 SQL 使用引号包裹 schema 和表名
|
||||
_INSERT_INTO_PATTERN = re.compile(
|
||||
r'INSERT\s+INTO\s+"?dwd"?\."?\w+"?\s*\(',
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
# SELECT ... FROM "ods"."<table>"
|
||||
_SELECT_FROM_ODS_PATTERN = re.compile(
|
||||
r'SELECT\s+.+?\s+FROM\s+"?ods"?\."?\w+"?',
|
||||
re.IGNORECASE | re.DOTALL,
|
||||
)
|
||||
|
||||
# ON CONFLICT (...) DO UPDATE SET 或 DO NOTHING
|
||||
_ON_CONFLICT_PATTERN = re.compile(
|
||||
r"ON\s+CONFLICT\s*\([^)]+\)\s+DO\s+(UPDATE\s+SET|NOTHING)",
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
# IS DISTINCT FROM（变更检测）
|
||||
_IS_DISTINCT_FROM_PATTERN = re.compile(
|
||||
r"IS\s+DISTINCT\s+FROM",
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
# RETURNING (xmax = 0) AS inserted
|
||||
_RETURNING_PATTERN = re.compile(
|
||||
r"RETURNING\s+\(xmax\s*=\s*0\)\s+AS\s+inserted",
|
||||
re.IGNORECASE,
|
||||
)
|
||||
|
||||
|
||||
@given(
|
||||
table_pair=st.sampled_from(_FACT_TABLES),
|
||||
window=window_pair(),
|
||||
fact_upsert=st.booleans(),
|
||||
)
|
||||
@settings(max_examples=100)
|
||||
def test_fact_increment_sql_structure(
|
||||
table_pair: Tuple[str, str],
|
||||
window: Tuple[datetime, datetime],
|
||||
fact_upsert: bool,
|
||||
):
|
||||
"""
|
||||
Property 3: 事实表主增量 SQL 结构等价性
|
||||
|
||||
对于任意事实表、窗口参数和 fact_upsert 配置，
|
||||
_merge_fact_increment() 生成的主增量 SQL 应包含：
|
||||
1. INSERT INTO <dwd_table> 结构
|
||||
2. SELECT ... FROM <ods_table> 结构
|
||||
3. ON CONFLICT (<pk_cols>) DO UPDATE SET 或 DO NOTHING 结构
|
||||
4. IS DISTINCT FROM 变更检测（当有 DO UPDATE SET 时）
|
||||
5. RETURNING (xmax = 0) AS inserted 结构
|
||||
|
||||
**Validates: Requirements 5.2, 5.5**
|
||||
"""
|
||||
dwd_table, ods_table = table_pair
|
||||
window_start, window_end = window
|
||||
|
||||
task = _make_task(fact_upsert=fact_upsert)
|
||||
|
||||
# 构建列信息
|
||||
fake_cols = [row[0] for row in PropertyTestCursor._fake_columns(None)]
|
||||
mapping_entries = task.FACT_MAPPINGS.get(dwd_table) or []
|
||||
mapping_dest = [dst for dst, _, _ in mapping_entries]
|
||||
|
||||
dwd_cols = list(set(fake_cols + mapping_dest))
|
||||
ods_cols = list(fake_cols)
|
||||
|
||||
fake_types = {row[0]: row[1] for row in PropertyTestCursor._fake_columns(None)}
|
||||
dwd_types = dict(fake_types)
|
||||
ods_types = dict(fake_types)
|
||||
for col in mapping_dest:
|
||||
if col not in dwd_types:
|
||||
dwd_types[col] = "text"
|
||||
|
||||
# 使用注入主键的 cursor，确保 ON CONFLICT 子句被生成
|
||||
# 选择 "id" 作为主键——它存在于 _fake_columns 返回的列中
|
||||
cur = _PkAwareCursor(pk_cols=["id"])
|
||||
|
||||
result = task._merge_fact_increment(
|
||||
cur=cur,
|
||||
dwd_table=dwd_table,
|
||||
ods_table=ods_table,
|
||||
dwd_cols=dwd_cols,
|
||||
ods_cols=ods_cols,
|
||||
dwd_types=dwd_types,
|
||||
ods_types=ods_types,
|
||||
window_start=window_start,
|
||||
window_end=window_end,
|
||||
)
|
||||
|
||||
# 提取主 INSERT SQL
|
||||
insert_sqls = [
|
||||
entry["sql"]
|
||||
for entry in cur.executed
|
||||
if entry["sql"].upper().startswith("INSERT INTO")
|
||||
]
|
||||
|
||||
assert len(insert_sqls) == 1, (
|
||||
f"期望恰好 1 条 INSERT SQL，实际 {len(insert_sqls)} 条，"
|
||||
f"表: {dwd_table}"
|
||||
)
|
||||
|
||||
sql = insert_sqls[0]
|
||||
|
||||
# 1. 验证 INSERT INTO dwd.<table> 结构
|
||||
assert _INSERT_INTO_PATTERN.search(sql), (
|
||||
f"SQL 缺少 INSERT INTO dwd.<table> 结构，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 2. 验证 SELECT ... FROM ods.<table> 结构
|
||||
assert _SELECT_FROM_ODS_PATTERN.search(sql), (
|
||||
f"SQL 缺少 SELECT ... FROM ods.<table> 结构，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 3. 验证 ON CONFLICT 结构
|
||||
assert _ON_CONFLICT_PATTERN.search(sql), (
|
||||
f"SQL 缺少 ON CONFLICT (...) DO UPDATE SET / DO NOTHING 结构，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 4. 验证 IS DISTINCT FROM 变更检测逻辑
|
||||
# 有 DO UPDATE SET 时，必须有 IS DISTINCT FROM 变更检测
|
||||
has_do_update = re.search(
|
||||
r"ON\s+CONFLICT\s*\([^)]+\)\s+DO\s+UPDATE\s+SET",
|
||||
sql, re.IGNORECASE,
|
||||
)
|
||||
if has_do_update:
|
||||
assert _IS_DISTINCT_FROM_PATTERN.search(sql), (
|
||||
f"SQL 有 DO UPDATE SET 但缺少 IS DISTINCT FROM 变更检测，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
|
||||
# 5. 验证 RETURNING (xmax = 0) AS inserted 结构
|
||||
assert _RETURNING_PATTERN.search(sql), (
|
||||
f"SQL 缺少 RETURNING (xmax = 0) AS inserted 结构，"
|
||||
f"表: {dwd_table}\nSQL: {sql[:500]}"
|
||||
)
|
||||
325
tests/test_etl_refactor_properties.py
Normal file
325
tests/test_etl_refactor_properties.py
Normal file
@@ -0,0 +1,325 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""Monorepo 级属性测试 — ETL DWS/Flow 重构
|
||||
|
||||
集中存放跨模块的属性测试，验证重构后的正确性属性。
|
||||
"""
|
||||
import sys
|
||||
import string
|
||||
from pathlib import Path
|
||||
|
||||
import pytest
|
||||
from hypothesis import given, settings, assume
|
||||
from hypothesis import strategies as st
|
||||
|
||||
# 将 ETL feiqiu 模块加入 sys.path，以便直接导入
|
||||
_ETL_ROOT = Path(__file__).resolve().parent.parent / "apps" / "etl" / "connectors" / "feiqiu"
|
||||
if str(_ETL_ROOT) not in sys.path:
|
||||
sys.path.insert(0, str(_ETL_ROOT))
|
||||
|
||||
from cli.main import parse_layers, VALID_LAYERS # noqa: E402
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Property 5: --layers 解析正确性
|
||||
# Feature: etl-dws-flow-refactor, Property 5: --layers 解析正确性
|
||||
# Validates: Requirements 6.1, 6.2
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
_ALL_LAYERS = sorted(VALID_LAYERS)
|
||||
|
||||
_layer_subsets = st.lists(
|
||||
st.sampled_from(_ALL_LAYERS),
|
||||
min_size=1,
|
||||
max_size=4,
|
||||
unique=True,
|
||||
)
|
||||
|
||||
_invalid_layer = st.text(
|
||||
alphabet=string.ascii_letters,
|
||||
min_size=1,
|
||||
max_size=10,
|
||||
).filter(lambda s: s.upper() not in VALID_LAYERS)
|
||||
|
||||
|
||||
class TestParseLayersProperty:
|
||||
"""Property 5: --layers 解析正确性
|
||||
|
||||
# Feature: etl-dws-flow-refactor, Property 5: --layers 解析正确性
|
||||
"""
|
||||
|
||||
@given(subset=_layer_subsets)
|
||||
@settings(max_examples=100)
|
||||
def test_valid_subset_parsed_correctly(self, subset: list[str]):
|
||||
"""任意合法层子集经逗号拼接后，parse_layers 返回恰好该子集（大写）。
|
||||
|
||||
**Validates: Requirements 6.1, 6.2**
|
||||
"""
|
||||
raw = ",".join(subset)
|
||||
result = parse_layers(raw)
|
||||
assert set(result) == set(subset)
|
||||
assert all(l == l.upper() for l in result)
|
||||
assert len(result) == len(subset)
|
||||
|
||||
@given(
|
||||
valid=_layer_subsets,
|
||||
bad=_invalid_layer,
|
||||
)
|
||||
@settings(max_examples=100)
|
||||
def test_invalid_layer_raises(self, valid: list[str], bad: str):
|
||||
"""包含无效层名时必须抛出 ValueError。
|
||||
|
||||
**Validates: Requirements 6.1, 6.2**
|
||||
"""
|
||||
assume(bad.upper() not in VALID_LAYERS)
|
||||
raw = ",".join(valid + [bad])
|
||||
with pytest.raises(ValueError, match="无效的层名"):
|
||||
parse_layers(raw)
|
||||
|
||||
|
||||
from orchestration.flow_runner import FlowRunner # noqa: E402
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Property 6: 配置优先级——配置值优先于 Registry
|
||||
# Feature: etl-dws-flow-refactor, Property 6: 配置优先级
|
||||
# Validates: Requirements 7.2
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
_LAYER_CONFIG_KEY = {
|
||||
"ODS": "run.ods_tasks",
|
||||
"DWD": "run.dwd_tasks",
|
||||
"DWS": "run.dws_tasks",
|
||||
"INDEX": "run.index_tasks",
|
||||
}
|
||||
|
||||
|
||||
class _DictConfig:
|
||||
"""模拟 AppConfig.get()，支持点号路径查找。"""
|
||||
|
||||
def __init__(self, data: dict):
|
||||
self._data = data
|
||||
|
||||
def get(self, key: str, default=None):
|
||||
keys = key.split(".")
|
||||
node = self._data
|
||||
for k in keys:
|
||||
if isinstance(node, dict) and k in node:
|
||||
node = node[k]
|
||||
else:
|
||||
return default
|
||||
return node
|
||||
|
||||
|
||||
def _make_runner(config_data: dict, registry_by_layer: dict) -> FlowRunner:
|
||||
"""构造最小 FlowRunner 实例，仅满足 _resolve_tasks 所需。"""
|
||||
import logging
|
||||
from unittest.mock import MagicMock
|
||||
|
||||
config = _DictConfig(config_data)
|
||||
registry = MagicMock()
|
||||
registry.get_tasks_by_layer = MagicMock(
|
||||
side_effect=lambda layer: registry_by_layer.get(layer.upper(), [])
|
||||
)
|
||||
runner = object.__new__(FlowRunner)
|
||||
runner.config = config
|
||||
runner.task_registry = registry
|
||||
runner.logger = logging.getLogger("test_prop6")
|
||||
return runner
|
||||
|
||||
|
||||
_task_code_p6 = st.from_regex(r"[A-Z][A-Z0-9_]{2,20}", fullmatch=True)
|
||||
_task_list_p6 = st.lists(_task_code_p6, min_size=1, max_size=8, unique=True)
|
||||
_layer_p6 = st.sampled_from(["ODS", "DWD", "DWS", "INDEX"])
|
||||
|
||||
|
||||
class TestConfigPriorityProperty:
|
||||
"""Property 6: 配置优先级——配置值优先于 Registry
|
||||
|
||||
# Feature: etl-dws-flow-refactor, Property 6: 配置优先级
|
||||
"""
|
||||
|
||||
@given(layer=_layer_p6, config_tasks=_task_list_p6, registry_tasks=_task_list_p6)
|
||||
@settings(max_examples=100)
|
||||
def test_config_overrides_registry(
|
||||
self, layer: str, config_tasks: list[str], registry_tasks: list[str]
|
||||
):
|
||||
"""任意层名 + 非空配置任务列表 → 返回配置值，忽略 Registry。
|
||||
|
||||
**Validates: Requirements 7.2**
|
||||
"""
|
||||
assume(set(config_tasks) != set(registry_tasks))
|
||||
key = _LAYER_CONFIG_KEY[layer]
|
||||
parts = key.split(".")
|
||||
config_data = {parts[0]: {parts[1]: config_tasks}}
|
||||
runner = _make_runner(config_data, {layer: registry_tasks})
|
||||
result = runner._resolve_tasks([layer])
|
||||
assert result == config_tasks
|
||||
|
||||
@given(layer=_layer_p6, registry_tasks=_task_list_p6)
|
||||
@settings(max_examples=100)
|
||||
def test_empty_config_falls_back_to_registry(
|
||||
self, layer: str, registry_tasks: list[str]
|
||||
):
|
||||
"""配置为空时回退到 Registry 结果。
|
||||
|
||||
**Validates: Requirements 7.2**
|
||||
"""
|
||||
runner = _make_runner({}, {layer: registry_tasks})
|
||||
result = runner._resolve_tasks([layer])
|
||||
assert result == registry_tasks
|
||||
|
||||
|
||||
from orchestration.topological_sort import topological_sort # noqa: E402
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Property 7 & 8 辅助：轻量 TaskMeta 替身 + 简易 Registry
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
from dataclasses import dataclass, field as dc_field
|
||||
|
||||
|
||||
@dataclass
|
||||
class _FakeMeta:
|
||||
depends_on: list[str] = dc_field(default_factory=list)
|
||||
|
||||
|
||||
class _FakeRegistry:
|
||||
"""最小 Registry 替身，仅提供 get_metadata()。"""
|
||||
|
||||
def __init__(self, deps: dict[str, list[str]]):
|
||||
self._deps = deps
|
||||
|
||||
def get_metadata(self, code: str):
|
||||
if code in self._deps:
|
||||
return _FakeMeta(depends_on=self._deps[code])
|
||||
return _FakeMeta()
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Hypothesis 策略：DAG 生成 & 含环图生成
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
@st.composite
|
||||
def dag_strategy(draw):
|
||||
"""生成随机 DAG：(task_codes, deps_dict)。
|
||||
|
||||
通过仅允许依赖索引更小的任务来保证无环。
|
||||
"""
|
||||
n = draw(st.integers(min_value=1, max_value=10))
|
||||
codes = [f"TASK_{i}" for i in range(n)]
|
||||
deps: dict[str, list[str]] = {}
|
||||
for i, code in enumerate(codes):
|
||||
if i > 0:
|
||||
possible = codes[:i]
|
||||
chosen = draw(
|
||||
st.lists(
|
||||
st.sampled_from(possible),
|
||||
max_size=min(3, i),
|
||||
unique=True,
|
||||
)
|
||||
)
|
||||
deps[code] = chosen
|
||||
else:
|
||||
deps[code] = []
|
||||
return codes, deps
|
||||
|
||||
|
||||
@st.composite
|
||||
def cyclic_graph_strategy(draw):
|
||||
"""生成一定包含至少一个环的有向图：(task_codes, deps_dict, cycle_tasks)。"""
|
||||
n = draw(st.integers(min_value=2, max_value=8))
|
||||
codes = [f"TASK_{i}" for i in range(n)]
|
||||
deps: dict[str, list[str]] = {code: [] for code in codes}
|
||||
# 选取 2-4 个任务构成环
|
||||
cycle_size = draw(st.integers(min_value=2, max_value=min(4, n)))
|
||||
cycle_tasks = draw(
|
||||
st.lists(
|
||||
st.sampled_from(codes),
|
||||
min_size=cycle_size,
|
||||
max_size=cycle_size,
|
||||
unique=True,
|
||||
)
|
||||
)
|
||||
for i in range(len(cycle_tasks)):
|
||||
next_task = cycle_tasks[(i + 1) % len(cycle_tasks)]
|
||||
if next_task not in deps[cycle_tasks[i]]:
|
||||
deps[cycle_tasks[i]].append(next_task)
|
||||
return codes, deps, cycle_tasks
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Property 7: 拓扑排序正确性
|
||||
# Feature: etl-dws-flow-refactor, Property 7: 拓扑排序正确性
|
||||
# Validates: Requirements 8.3
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
class TestTopologicalSortProperty:
|
||||
"""Property 7: 拓扑排序正确性
|
||||
|
||||
# Feature: etl-dws-flow-refactor, Property 7: 拓扑排序正确性
|
||||
"""
|
||||
|
||||
@given(data=dag_strategy())
|
||||
@settings(max_examples=100)
|
||||
def test_dependencies_precede_dependents(self, data):
|
||||
"""对于任意 DAG，排序结果中每个任务的依赖都排在它之前。
|
||||
|
||||
**Validates: Requirements 8.3**
|
||||
"""
|
||||
codes, deps = data
|
||||
registry = _FakeRegistry(deps)
|
||||
result = topological_sort(codes, registry)
|
||||
|
||||
# 结果包含且仅包含输入任务（无丢失、无重复）
|
||||
assert set(result) == set(codes)
|
||||
assert len(result) == len(codes)
|
||||
|
||||
# 每个任务的依赖（在列表内的）都排在它之前
|
||||
pos = {code: idx for idx, code in enumerate(result)}
|
||||
for code in codes:
|
||||
for dep in deps.get(code, []):
|
||||
if dep in pos:
|
||||
assert pos[dep] < pos[code], (
|
||||
f"依赖 {dep} 应排在 {code} 之前，"
|
||||
f"但位置为 {pos[dep]} vs {pos[code]}"
|
||||
)
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Property 8: 循环依赖检测
|
||||
# Feature: etl-dws-flow-refactor, Property 8: 循环依赖检测
|
||||
# Validates: Requirements 8.4
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
class TestCyclicDependencyProperty:
|
||||
"""Property 8: 循环依赖检测
|
||||
|
||||
# Feature: etl-dws-flow-refactor, Property 8: 循环依赖检测
|
||||
"""
|
||||
|
||||
@given(data=cyclic_graph_strategy())
|
||||
@settings(max_examples=100)
|
||||
def test_cycle_raises_value_error(self, data):
|
||||
"""对于任意含环有向图，topological_sort 应抛出 ValueError。
|
||||
|
||||
**Validates: Requirements 8.4**
|
||||
"""
|
||||
codes, deps, cycle_tasks = data
|
||||
registry = _FakeRegistry(deps)
|
||||
with pytest.raises(ValueError, match="循环依赖"):
|
||||
topological_sort(codes, registry)
|
||||
|
||||
@given(data=cyclic_graph_strategy())
|
||||
@settings(max_examples=100)
|
||||
def test_cycle_error_mentions_involved_tasks(self, data):
|
||||
"""错误信息中应包含环中涉及的至少一个任务代码。
|
||||
|
||||
**Validates: Requirements 8.4**
|
||||
"""
|
||||
codes, deps, cycle_tasks = data
|
||||
registry = _FakeRegistry(deps)
|
||||
with pytest.raises(ValueError) as exc_info:
|
||||
topological_sort(codes, registry)
|
||||
msg = str(exc_info.value)
|
||||
# 至少一个环中的任务出现在错误信息中
|
||||
assert any(t in msg for t in cycle_tasks), (
|
||||
f"错误信息 '{msg}' 中未包含环任务 {cycle_tasks} 中的任何一个"
|
||||
)
|
||||
@@ -1,12 +1,25 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""
|
||||
Property 5: 文件迁移完整性
|
||||
Property 5: 文件迁移完整性（已归档）
|
||||
|
||||
对于任意源-目标目录映射关系（ETL 业务代码、database 文件、tests 目录），
|
||||
原始用途：验证从 C:\\ZQYY\\FQ-ETL\\ 迁移到 NeoZQYY monorepo 时，
|
||||
源目录中的每个文件在目标目录的对应位置都应存在且内容一致。
|
||||
|
||||
归档原因：迁移已于 2025 年完成，后续多轮重构（dwd-phase1-refactor、
|
||||
etl-dws-flow-refactor、ods-dedup-standardize 等）对目标代码做了大量
|
||||
结构性修改，源-目标 1:1 对比前提不再成立。扫描显示 50+ 个文件已合理分化。
|
||||
|
||||
如需重新启用，可移除模块级 skip 标记。
|
||||
|
||||
**Validates: Requirements 5.1, 5.2, 5.3**
|
||||
"""
|
||||
import pytest
|
||||
|
||||
# 迁移已完成且目标代码经多轮重构已合理分化，此测试模块整体跳过
|
||||
pytestmark = pytest.mark.skip(
|
||||
reason="文件迁移已完成，后续重构导致源-目标合理分化（50+ 文件），测试使命结束"
|
||||
)
|
||||
|
||||
import hashlib
|
||||
import os
|
||||
from typing import List, Tuple
|
||||
@@ -16,24 +29,20 @@ from hypothesis.strategies import sampled_from
|
||||
|
||||
# 源-目标目录映射（需求 5.1: ETL 业务代码，5.2: database，5.3: tests）
|
||||
MIGRATION_MAPPINGS: List[Tuple[str, str]] = [
|
||||
# ETL 业务代码目录（需求 5.1）
|
||||
(r"C:\ZQYY\FQ-ETL\api", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\api"),
|
||||
(r"C:\ZQYY\FQ-ETL\cli", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\cli"),
|
||||
(r"C:\ZQYY\FQ-ETL\config", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\config"),
|
||||
(r"C:\ZQYY\FQ-ETL\loaders", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\loaders"),
|
||||
(r"C:\ZQYY\FQ-ETL\models", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\models"),
|
||||
(r"C:\ZQYY\FQ-ETL\orchestration", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\orchestration"),
|
||||
(r"C:\ZQYY\FQ-ETL\scd", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\scd"),
|
||||
(r"C:\ZQYY\FQ-ETL\tasks", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\tasks"),
|
||||
(r"C:\ZQYY\FQ-ETL\utils", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\utils"),
|
||||
(r"C:\ZQYY\FQ-ETL\quality", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\quality"),
|
||||
# tests 子目录（需求 5.3）— 只映射 ETL 自身的 unit/integration，
|
||||
# Monorepo 级属性测试（test_property_*.py）按设计放在 C:\NeoZQYY\tests\
|
||||
(r"C:\ZQYY\FQ-ETL\tests\unit", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\tests\unit"),
|
||||
(r"C:\ZQYY\FQ-ETL\tests\integration", r"C:\NeoZQYY\apps\etl\pipelines\feiqiu\tests\integration"),
|
||||
(r"C:\ZQYY\FQ-ETL\api", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\api"),
|
||||
(r"C:\ZQYY\FQ-ETL\cli", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\cli"),
|
||||
(r"C:\ZQYY\FQ-ETL\config", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\config"),
|
||||
(r"C:\ZQYY\FQ-ETL\loaders", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\loaders"),
|
||||
(r"C:\ZQYY\FQ-ETL\models", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\models"),
|
||||
(r"C:\ZQYY\FQ-ETL\orchestration", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\orchestration"),
|
||||
(r"C:\ZQYY\FQ-ETL\scd", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\scd"),
|
||||
(r"C:\ZQYY\FQ-ETL\tasks", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\tasks"),
|
||||
(r"C:\ZQYY\FQ-ETL\utils", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\utils"),
|
||||
(r"C:\ZQYY\FQ-ETL\quality", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\quality"),
|
||||
(r"C:\ZQYY\FQ-ETL\tests\unit", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\tests\unit"),
|
||||
(r"C:\ZQYY\FQ-ETL\tests\integration", r"C:\NeoZQYY\apps\etl\connectors\feiqiu\tests\integration"),
|
||||
]
|
||||
|
||||
# 排除模式：__pycache__ 等不参与比较
|
||||
EXCLUDE_DIRS = {"__pycache__", ".pytest_cache", ".hypothesis"}
|
||||
|
||||
|
||||
@@ -67,7 +76,6 @@ def test_all_source_files_exist_in_target(mapping: Tuple[str, str]) -> None:
|
||||
**Validates: Requirements 5.1, 5.2, 5.3**
|
||||
"""
|
||||
src_dir, dst_dir = mapping
|
||||
|
||||
assert os.path.isdir(src_dir), f"源目录不存在: {src_dir}"
|
||||
assert os.path.isdir(dst_dir), f"目标目录不存在: {dst_dir}"
|
||||
|
||||
@@ -96,7 +104,6 @@ def test_source_and_target_file_content_identical(mapping: Tuple[str, str]) -> N
|
||||
**Validates: Requirements 5.1, 5.2, 5.3**
|
||||
"""
|
||||
src_dir, dst_dir = mapping
|
||||
|
||||
assert os.path.isdir(src_dir), f"源目录不存在: {src_dir}"
|
||||
assert os.path.isdir(dst_dir), f"目标目录不存在: {dst_dir}"
|
||||
|
||||
@@ -106,17 +113,13 @@ def test_source_and_target_file_content_identical(mapping: Tuple[str, str]) -> N
|
||||
for rel_path in src_files:
|
||||
src_path = os.path.join(src_dir, rel_path)
|
||||
dst_path = os.path.join(dst_dir, rel_path)
|
||||
|
||||
if not os.path.isfile(dst_path):
|
||||
continue
|
||||
|
||||
src_hash = _file_hash(src_path)
|
||||
dst_hash = _file_hash(dst_path)
|
||||
if src_hash != dst_hash:
|
||||
if _file_hash(src_path) != _file_hash(dst_path):
|
||||
mismatched.append(rel_path)
|
||||
|
||||
assert not mismatched, (
|
||||
f"源目录 {src_dir} 与目标目录 {dst_dir} 中 {len(mismatched)} 个文件内容不一致:\n"
|
||||
+ "\n".join(f" - {f}" for f in mismatched[:10])
|
||||
+ (f"\n ... 及其他 {len(mismatched) - 10} 个" if len(mismatched) > 10 else "")
|
||||
)
|
||||
)
|
||||
|
||||
@@ -15,10 +15,9 @@ from hypothesis.strategies import sampled_from
|
||||
|
||||
# uv workspace 声明的 Python 子项目成员
|
||||
WORKSPACE_MEMBERS = [
|
||||
"apps/etl/pipelines/feiqiu",
|
||||
"apps/etl/connectors/feiqiu",
|
||||
"apps/backend",
|
||||
"packages/shared",
|
||||
"gui",
|
||||
]
|
||||
|
||||
MONOREPO_ROOT = r"C:\NeoZQYY"
|
||||
|
||||
@@ -17,9 +17,9 @@ from hypothesis.strategies import sampled_from
|
||||
MONOREPO_ROOT = r"C:\NeoZQYY"
|
||||
|
||||
# 一级目录列表（需求 1.5 定义）
|
||||
# 仅包含实际存在的一级目录
|
||||
TOP_LEVEL_DIRS = [
|
||||
"apps",
|
||||
"gui",
|
||||
"packages",
|
||||
"db",
|
||||
"docs",
|
||||
|
||||
Reference in New Issue
Block a user