B-42: 导入功能加固——逐工作表确认、流式上传与诊断证据
This commit is contained in:
腾讯WorkBuddy
@@ -0,0 +1,587 @@
|
||||
"""HTTP integration tests for the hardened import API (B-42).
|
||||
|
||||
Covers streaming multipart limits, per-worksheet parse results with the four
|
||||
diagnostic evidence items, the confirm/ignore review lifecycle with tenant
|
||||
scoping and idempotency, unconfirmed-sheet export gating, and concurrent
|
||||
duplicate uploads. Uses a real ``ThreadingHTTPServer`` like
|
||||
``test_server_auth``.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import io
|
||||
import itertools
|
||||
import json
|
||||
import os
|
||||
from pathlib import Path
|
||||
import tempfile
|
||||
import threading
|
||||
import unittest
|
||||
|
||||
from openpyxl import Workbook
|
||||
|
||||
from bank_importer import auth
|
||||
from bank_importer.db import connect, migrate
|
||||
|
||||
import server
|
||||
from test_server_auth import Client, as_json
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[1]
|
||||
SAMPLES = ROOT / "流水模板"
|
||||
CCB_SAMPLE = SAMPLES / "中国建设银行账户流水.xls"
|
||||
CITIC_SAMPLE = SAMPLES / "中信银行账户流水.xlsx"
|
||||
|
||||
BOOTSTRAP_PASSWORD = "BootAdmin123"
|
||||
ADMIN_PASSWORD = "AdminPass123"
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Fixtures
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
CCB_HEADER = [
|
||||
"客户账号", "账户名称", "交易时间", "借方发生额（支取）", "贷方发生额（收入）",
|
||||
"余额", "币种", "对方户名", "对方账号", "对方开户机构", "摘要", "备注",
|
||||
]
|
||||
CCB_DATA = [
|
||||
["6228480000000000", "测试公司", "2026-01-05 10:00:00", "100.00", "", "99900.00", "RMB", "供应商", "1002003004", "某银行", "货款", ""],
|
||||
["6228480000000000", "测试公司", "2026-01-06 11:00:00", "", "200.00", "100100.00", "RMB", "客户", "2003004005", "某银行", "收款", ""],
|
||||
]
|
||||
|
||||
|
||||
def workbook_bytes(sheets):
|
||||
"""Build an xlsx in memory.
|
||||
|
||||
``sheets`` is a list of ``(name, rows)``; an empty ``rows`` list means an
|
||||
empty worksheet.
|
||||
"""
|
||||
workbook = Workbook()
|
||||
workbook.remove(workbook.active)
|
||||
for name, rows in sheets:
|
||||
worksheet = workbook.create_sheet(name)
|
||||
for row in rows:
|
||||
worksheet.append(row)
|
||||
buffer = io.BytesIO()
|
||||
workbook.save(buffer)
|
||||
return buffer.getvalue()
|
||||
|
||||
|
||||
# Workbook fixtures are content-hashed for dedupe, so every helper produces
|
||||
# distinct bytes per call (each test uploads a genuinely new file).
|
||||
_counter = itertools.count()
|
||||
|
||||
|
||||
def multi_sheet_bytes():
|
||||
"""Valid CCB sheet + unknown-header sheet + empty sheet in one workbook."""
|
||||
return workbook_bytes(
|
||||
[
|
||||
("正常流水", [CCB_HEADER, *CCB_DATA]),
|
||||
("未知模板", [["日期", "金额", "备注"], ["2026-01-01", next(_counter), "x"]]),
|
||||
("空表", []),
|
||||
]
|
||||
)
|
||||
|
||||
|
||||
def empty_workbook_bytes():
|
||||
return workbook_bytes([("Sheet1", [])])
|
||||
|
||||
|
||||
def single_cell_bytes():
|
||||
return workbook_bytes([("候选", [[f"只有一个单元格 {next(_counter)}"]])])
|
||||
|
||||
|
||||
def unknown_header_bytes():
|
||||
return workbook_bytes([("流水", [["日期", "金额", "备注"], ["2026-01-01", next(_counter), "x"]])])
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Test server
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
|
||||
class ImportApiServerTests(unittest.TestCase):
|
||||
@classmethod
|
||||
def setUpClass(cls) -> None:
|
||||
cls.temp_dir = tempfile.TemporaryDirectory()
|
||||
root = Path(cls.temp_dir.name)
|
||||
cls.db_path = root / "app.db"
|
||||
cls.storage = root / "files"
|
||||
|
||||
cls._old_db_path = server.DB_PATH
|
||||
cls._old_storage = server.STORAGE_DIR
|
||||
server.DB_PATH = cls.db_path
|
||||
server.STORAGE_DIR = cls.storage
|
||||
|
||||
os.environ["APP_BOOTSTRAP_ADMIN_PASSWORD"] = BOOTSTRAP_PASSWORD
|
||||
connection = connect(cls.db_path)
|
||||
migrate(connection)
|
||||
generated = server.ensure_bootstrap_admin(connection)
|
||||
assert generated is None
|
||||
connection.close()
|
||||
|
||||
class QuietHandler(server.AppHandler):
|
||||
def log_message(self, *args) -> None:
|
||||
pass
|
||||
|
||||
cls.httpd = server.ThreadingHTTPServer(("127.0.0.1", 0), QuietHandler)
|
||||
cls.port = cls.httpd.server_address[1]
|
||||
cls.thread = threading.Thread(target=cls.httpd.serve_forever, daemon=True)
|
||||
cls.thread.start()
|
||||
|
||||
cls.admin = Client("127.0.0.1", cls.port)
|
||||
status, _, data = cls.admin.post_json(
|
||||
"/api/login",
|
||||
{"username": "group-admin", "password": BOOTSTRAP_PASSWORD, "portal": "admin"},
|
||||
)
|
||||
assert status == 200, data
|
||||
status, _, data = cls.admin.post_json(
|
||||
"/api/password/change",
|
||||
{"old_password": BOOTSTRAP_PASSWORD, "new_password": ADMIN_PASSWORD},
|
||||
)
|
||||
assert status == 200, data
|
||||
|
||||
status, _, data = cls.admin.post_json("/api/admin/companies", {"name": "甲公司"})
|
||||
assert status == 200, data
|
||||
cls.company_a = as_json(data)["company_id"]
|
||||
status, _, data = cls.admin.post_json("/api/admin/companies", {"name": "乙公司"})
|
||||
assert status == 200, data
|
||||
cls.company_b = as_json(data)["company_id"]
|
||||
|
||||
cls.cashier_a, cls.cashier_a_password = cls.create_company_user(
|
||||
cls.admin, cls.port, "cashier-a"
|
||||
)
|
||||
cls.cashier_b, cls.cashier_b_password = cls.create_company_user(
|
||||
cls.admin, cls.port, "cashier-b"
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def create_company_user(cls, admin, port, username):
|
||||
status, _, data = admin.post_json(
|
||||
"/api/admin/users", {"username": username, "company_id": cls.company_a}
|
||||
)
|
||||
assert status == 200, data
|
||||
initial = as_json(data)["initial_password"]
|
||||
client = Client("127.0.0.1", port)
|
||||
status, _, data = client.post_json(
|
||||
"/api/login",
|
||||
{"username": username, "password": initial, "portal": "company"},
|
||||
)
|
||||
assert status == 200, data
|
||||
new_password = "Changed456"
|
||||
status, _, data = client.post_json(
|
||||
"/api/password/change",
|
||||
{"old_password": initial, "new_password": new_password},
|
||||
)
|
||||
assert status == 200, data
|
||||
return client, new_password
|
||||
|
||||
@classmethod
|
||||
def tearDownClass(cls) -> None:
|
||||
cls.httpd.shutdown()
|
||||
cls.httpd.server_close()
|
||||
server.DB_PATH = cls._old_db_path
|
||||
server.STORAGE_DIR = cls._old_storage
|
||||
os.environ.pop("APP_BOOTSTRAP_ADMIN_PASSWORD", None)
|
||||
cls.temp_dir.cleanup()
|
||||
|
||||
def fresh_client(self) -> Client:
|
||||
return Client("127.0.0.1", self.port)
|
||||
|
||||
def upload(self, client, content: bytes, filename: str = "语句.xlsx", fields=None):
|
||||
return client.post_multipart(
|
||||
"/api/parse", fields or {}, filename, content
|
||||
)
|
||||
|
||||
def upload_for_company(self, content: bytes, company_id: int, filename: str = "语句.xlsx"):
|
||||
return self.admin.post_multipart(
|
||||
"/api/parse", {"company_id": str(company_id)}, filename, content
|
||||
)
|
||||
|
||||
def upload_for_a(self, content: bytes, filename: str = "语句.xlsx"):
|
||||
return self.cashier_a.post_multipart("/api/parse", {}, filename, content)
|
||||
|
||||
def batch_sheets(self, client, batch_id: int):
|
||||
status, _, data = client.get(f"/api/batches/{batch_id}/sheets")
|
||||
self.assertEqual(200, status, data)
|
||||
return as_json(data)["sheets"]
|
||||
|
||||
def confirm(self, client, batch_id: int, sheets, reason=None):
|
||||
payload = {"sheets": sheets}
|
||||
if reason is not None:
|
||||
payload["reason"] = reason
|
||||
return client.post_json(f"/api/batches/{batch_id}/confirm", payload)
|
||||
|
||||
def ignore(self, client, batch_id: int, sheets, reason):
|
||||
return client.post_json(
|
||||
f"/api/batches/{batch_id}/ignore", {"sheets": sheets, "reason": reason}
|
||||
)
|
||||
|
||||
# ------------------------------------------------------------------
|
||||
# Multipart validation
|
||||
# ------------------------------------------------------------------
|
||||
|
||||
def test_oversized_body_returns_stable_error(self) -> None:
|
||||
content = b"x" * (server.MAX_UPLOAD_BYTES + 1)
|
||||
status, _, data = self.upload_for_a(content)
|
||||
self.assertEqual(422, status)
|
||||
payload = as_json(data)
|
||||
self.assertIn("20 MB", payload["message"])
|
||||
self.assertNotIn(server.STORAGE_DIR.as_posix(), payload["message"])
|
||||
self.assertNotIn(str(self.storage), payload["message"])
|
||||
|
||||
def test_missing_boundary_returns_stable_error(self) -> None:
|
||||
status, _, data = self.cashier_a.request(
|
||||
"POST",
|
||||
"/api/parse",
|
||||
body=b"whatever",
|
||||
headers={"Content-Type": "multipart/form-data"},
|
||||
)
|
||||
self.assertEqual(422, status)
|
||||
self.assertIn("边界", as_json(data)["message"])
|
||||
|
||||
def test_wrong_content_type_returns_stable_error(self) -> None:
|
||||
status, _, data = self.cashier_a.request(
|
||||
"POST",
|
||||
"/api/parse",
|
||||
body=CCB_SAMPLE.read_bytes(),
|
||||
headers={"Content-Type": "application/octet-stream"},
|
||||
)
|
||||
self.assertEqual(422, status)
|
||||
|
||||
def test_missing_content_length_returns_stable_error(self) -> None:
|
||||
boundary = "----missingcl"
|
||||
body = f"--{boundary}--\r\n".encode()
|
||||
status, _, data = self.cashier_a.request(
|
||||
"POST",
|
||||
"/api/parse",
|
||||
body=body,
|
||||
headers={"Content-Type": f"multipart/form-data; boundary={boundary}"},
|
||||
)
|
||||
self.assertEqual(422, status)
|
||||
|
||||
def test_bad_extension_returns_stable_error(self) -> None:
|
||||
status, _, data = self.upload_for_a(CCB_SAMPLE.read_bytes(), filename="statement.pdf")
|
||||
self.assertEqual(422, status)
|
||||
self.assertIn(".xls", as_json(data)["message"])
|
||||
|
||||
def test_signature_mismatch_returns_stable_error(self) -> None:
|
||||
# Valid .xls content renamed with a .xlsx extension must be rejected.
|
||||
status, _, data = self.upload_for_a(CCB_SAMPLE.read_bytes(), filename="fake.xlsx")
|
||||
self.assertEqual(422, status)
|
||||
self.assertIn("损坏", as_json(data)["message"])
|
||||
|
||||
def test_corrupt_file_returns_stable_error_without_temp_path(self) -> None:
|
||||
status, _, data = self.upload_for_a(b"PK\x03\x04 not a real zip at all")
|
||||
self.assertEqual(422, status)
|
||||
payload = as_json(data)
|
||||
text = json.dumps(payload, ensure_ascii=False)
|
||||
self.assertNotIn(str(self.storage), text)
|
||||
self.assertNotIn("temp", text.lower())
|
||||
self.assertNotIn("data/files", text)
|
||||
|
||||
def test_two_file_parts_rejected(self) -> None:
|
||||
boundary = "----twofilesboundary"
|
||||
content = CCB_SAMPLE.read_bytes()
|
||||
crlf = b"\r\n"
|
||||
body = (
|
||||
b"--" + boundary.encode() + crlf
|
||||
+ b'Content-Disposition: form-data; name="file"; filename="a.xls"' + crlf
|
||||
+ b"Content-Type: application/octet-stream" + crlf + crlf
|
||||
+ content + crlf
|
||||
+ b"--" + boundary.encode() + crlf
|
||||
+ b'Content-Disposition: form-data; name="file"; filename="b.xls"' + crlf
|
||||
+ b"Content-Type: application/octet-stream" + crlf + crlf
|
||||
+ content + crlf
|
||||
+ b"--" + boundary.encode() + b"--" + crlf
|
||||
)
|
||||
status, _, data = self.cashier_a.request(
|
||||
"POST",
|
||||
"/api/parse",
|
||||
body=body,
|
||||
headers={"Content-Type": f"multipart/form-data; boundary={boundary}"},
|
||||
)
|
||||
self.assertEqual(422, status)
|
||||
|
||||
def test_no_file_part_rejected(self) -> None:
|
||||
boundary = "----nofile"
|
||||
crlf = b"\r\n"
|
||||
body = (
|
||||
b"--" + boundary.encode() + crlf
|
||||
+ b'Content-Disposition: form-data; name="company_id"' + crlf + crlf
|
||||
+ b"1" + crlf
|
||||
+ b"--" + boundary.encode() + b"--" + crlf
|
||||
)
|
||||
status, _, data = self.cashier_a.request(
|
||||
"POST",
|
||||
"/api/parse",
|
||||
body=body,
|
||||
headers={"Content-Type": f"multipart/form-data; boundary={boundary}"},
|
||||
)
|
||||
self.assertEqual(422, status)
|
||||
self.assertIn("文件", as_json(data)["message"])
|
||||
|
||||
def test_upload_trailing_bytes_preserved_exactly(self) -> None:
|
||||
# The stored source file must be byte-identical to what was uploaded
|
||||
# even when it ends with CRLF/LF; the parser must never trim tail
|
||||
# bytes, and the upload temp must not appear anywhere in the payload.
|
||||
content = CCB_SAMPLE.read_bytes()
|
||||
uploaded = content + b"\r\n\r\n\x00\x01"
|
||||
status, _, data = self.upload_for_a(uploaded, filename="尾部字节.xls")
|
||||
self.assertEqual(200, status, data)
|
||||
batch_id = as_json(data)["batch_id"]
|
||||
connection = connect(self.db_path)
|
||||
try:
|
||||
row = connection.execute(
|
||||
"""
|
||||
SELECT f.storage_path FROM source_files f
|
||||
JOIN import_batches b ON b.source_file_id = f.id
|
||||
WHERE b.id = ?
|
||||
""",
|
||||
(batch_id,),
|
||||
).fetchone()
|
||||
finally:
|
||||
connection.close()
|
||||
self.assertIsNotNone(row)
|
||||
self.assertEqual(uploaded, Path(row["storage_path"]).read_bytes())
|
||||
self.assertNotIn("upload-", json.dumps(as_json(data), ensure_ascii=False))
|
||||
|
||||
# ------------------------------------------------------------------
|
||||
# Per-sheet results and four-item evidence
|
||||
# ------------------------------------------------------------------
|
||||
|
||||
def test_multi_sheet_workbook_returns_every_sheet_result(self) -> None:
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="多工作表.xlsx")
|
||||
self.assertEqual(200, status, data)
|
||||
payload = as_json(data)
|
||||
self.assertEqual("parsed", payload["status"])
|
||||
self.assertEqual("多工作表.xlsx", payload.get("original_filename"))
|
||||
sheets = payload["sheets"]
|
||||
self.assertEqual(3, len(sheets))
|
||||
|
||||
by_name = {sheet["sheet_name"]: sheet for sheet in sheets}
|
||||
self.assertEqual("parsed", by_name["正常流水"]["outcome"])
|
||||
self.assertEqual("pending", by_name["正常流水"]["review_status"])
|
||||
self.assertEqual("中国建设银行", by_name["正常流水"]["bank"])
|
||||
self.assertEqual(2, by_name["正常流水"]["transactions"])
|
||||
|
||||
unknown = by_name["未知模板"]
|
||||
self.assertEqual("exception", unknown["outcome"])
|
||||
self.assertIn("未识别到受支持的银行表头", unknown["message"])
|
||||
self.assertGreaterEqual(unknown["scanned_rows"], 2)
|
||||
self.assertTrue(unknown["candidate_headers"])
|
||||
self.assertIn("日期、金额、备注", "；".join(unknown["candidate_headers"]))
|
||||
|
||||
empty = by_name["空表"]
|
||||
self.assertEqual("ignored", empty["outcome"])
|
||||
self.assertIn("为空", empty["message"])
|
||||
self.assertEqual(0, empty["scanned_rows"])
|
||||
|
||||
def test_batch_detail_matches_parse_response(self) -> None:
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="详情.xlsx")
|
||||
batch_id = as_json(data)["batch_id"]
|
||||
sheets = self.batch_sheets(self.cashier_a, batch_id)
|
||||
self.assertEqual(3, len(sheets))
|
||||
status, _, data = self.cashier_a.get(f"/api/batches/{batch_id}/sheets")
|
||||
self.assertEqual(200, status)
|
||||
detail = as_json(data)
|
||||
self.assertEqual("详情.xlsx", detail["original_filename"])
|
||||
|
||||
def test_unknown_header_file_returns_four_evidence_items(self) -> None:
|
||||
status, _, data = self.upload_for_a(unknown_header_bytes(), filename="未知银行.xlsx")
|
||||
self.assertEqual(422, status, data)
|
||||
payload = as_json(data)
|
||||
self.assertEqual("exception", payload["status"])
|
||||
self.assertEqual("未知银行.xlsx", payload.get("original_filename"))
|
||||
sheet = payload["sheets"][0]
|
||||
self.assertEqual("exception", sheet["outcome"])
|
||||
self.assertEqual("流水", sheet["sheet_name"])
|
||||
self.assertGreaterEqual(sheet["scanned_rows"], 1)
|
||||
self.assertTrue(sheet["candidate_headers"])
|
||||
text = json.dumps(payload, ensure_ascii=False)
|
||||
for evidence in (str(self.storage), "data/files", ".uploads", "upload-"):
|
||||
self.assertNotIn(evidence, text)
|
||||
|
||||
def test_empty_workbook_returns_evidence(self) -> None:
|
||||
status, _, data = self.upload_for_a(empty_workbook_bytes(), filename="空工作簿.xlsx")
|
||||
self.assertEqual(422, status, data)
|
||||
payload = as_json(data)
|
||||
self.assertEqual("空工作簿.xlsx", payload.get("original_filename"))
|
||||
sheet = payload["sheets"][0]
|
||||
self.assertEqual("ignored", sheet["outcome"])
|
||||
self.assertEqual("Sheet1", sheet["sheet_name"])
|
||||
self.assertEqual(0, sheet["scanned_rows"])
|
||||
|
||||
def test_single_cell_sheet_returns_candidate_evidence(self) -> None:
|
||||
status, _, data = self.upload_for_a(single_cell_bytes(), filename="单格.xlsx")
|
||||
self.assertEqual(422, status, data)
|
||||
sheet = as_json(data)["sheets"][0]
|
||||
self.assertEqual("exception", sheet["outcome"])
|
||||
self.assertEqual("候选", sheet["sheet_name"])
|
||||
self.assertEqual(1, sheet["scanned_rows"])
|
||||
self.assertTrue(sheet["candidate_headers"])
|
||||
self.assertIn("只有一个单元格", "；".join(sheet["candidate_headers"]))
|
||||
|
||||
# ------------------------------------------------------------------
|
||||
# Confirm / ignore lifecycle
|
||||
# ------------------------------------------------------------------
|
||||
|
||||
def test_confirm_and_ignore_lifecycle(self) -> None:
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="确认.xlsx")
|
||||
payload = as_json(data)
|
||||
batch_id = payload["batch_id"]
|
||||
names = [s["sheet_name"] for s in payload["sheets"]]
|
||||
self.assertIn("正常流水", names)
|
||||
|
||||
# Confirm the parsed sheet; exception/ignored cannot be confirmed.
|
||||
status, _, data = self.confirm(self.cashier_a, batch_id, ["正常流水"])
|
||||
self.assertEqual(200, status, data)
|
||||
result = as_json(data)
|
||||
self.assertEqual(["正常流水"], result["updated"])
|
||||
sheet = next(s for s in result["sheets"] if s["sheet_name"] == "正常流水")
|
||||
self.assertEqual("confirmed", sheet["review_status"])
|
||||
|
||||
# Confirming an exception sheet is a conflict.
|
||||
status, _, data = self.confirm(self.cashier_a, batch_id, ["未知模板"])
|
||||
self.assertEqual(409, status, data)
|
||||
|
||||
# Ignore the exception sheet with a reason.
|
||||
status, _, data = self.ignore(self.cashier_a, batch_id, ["未知模板"], "模板待补充")
|
||||
self.assertEqual(200, status, data)
|
||||
sheet = next(s for s in as_json(data)["sheets"] if s["sheet_name"] == "未知模板")
|
||||
self.assertEqual("ignored", sheet["review_status"])
|
||||
self.assertEqual("模板待补充", sheet["review_reason"])
|
||||
|
||||
# Idempotent repeat of an already-applied decision.
|
||||
status, _, data = self.confirm(self.cashier_a, batch_id, ["正常流水"])
|
||||
self.assertEqual(200, status, data)
|
||||
result = as_json(data)
|
||||
self.assertEqual([], result["updated"])
|
||||
self.assertEqual(["正常流水"], result["already"])
|
||||
|
||||
# Changing a settled decision is a conflict.
|
||||
status, _, data = self.ignore(self.cashier_a, batch_id, ["正常流水"], "改主意")
|
||||
self.assertEqual(409, status, data)
|
||||
|
||||
def test_ignore_requires_reason(self) -> None:
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="忽略原因.xlsx")
|
||||
batch_id = as_json(data)["batch_id"]
|
||||
status, _, data = self.cashier_a.post_json(
|
||||
f"/api/batches/{batch_id}/ignore", {"sheets": ["未知模板"], "reason": " "}
|
||||
)
|
||||
self.assertEqual(400, status, data)
|
||||
self.assertIn("原因", as_json(data)["message"])
|
||||
|
||||
def test_unknown_sheet_name_is_rejected_and_atomic(self) -> None:
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="原子.xlsx")
|
||||
batch_id = as_json(data)["batch_id"]
|
||||
# One valid + one unknown sheet in the same request must not leave a
|
||||
# half-applied confirmation behind.
|
||||
status, _, data = self.confirm(self.cashier_a, batch_id, ["正常流水", "不存在"])
|
||||
self.assertEqual(400, status, data)
|
||||
sheet = next(
|
||||
s for s in self.batch_sheets(self.cashier_a, batch_id)
|
||||
if s["sheet_name"] == "正常流水"
|
||||
)
|
||||
self.assertEqual("pending", sheet["review_status"])
|
||||
|
||||
def test_company_cannot_confirm_another_companys_batch(self) -> None:
|
||||
status, _, data = self.upload_for_company(multi_sheet_bytes(), self.company_b)
|
||||
batch_b = as_json(data)["batch_id"]
|
||||
status, _, data = self.confirm(self.cashier_a, batch_b, ["正常流水"])
|
||||
self.assertEqual(404, status, data)
|
||||
status, _, data = self.cashier_a.get(f"/api/batches/{batch_b}/sheets")
|
||||
self.assertEqual(404, status, data)
|
||||
|
||||
def test_export_excludes_unconfirmed_and_ignore_keeps_excluded(self) -> None:
|
||||
def export_line_count() -> int:
|
||||
status, _, data = self.cashier_a.get("/api/export.csv")
|
||||
self.assertEqual(200, status)
|
||||
return len([line for line in data.decode("utf-8-sig").splitlines() if line])
|
||||
|
||||
baseline = export_line_count()
|
||||
|
||||
# Unconfirmed upload: export stays unchanged (no downstream leak).
|
||||
status, _, data = self.upload_for_a(CCB_SAMPLE.read_bytes(), filename="导出.xls")
|
||||
batch_id = as_json(data)["batch_id"]
|
||||
self.assertEqual("parsed", as_json(data)["status"])
|
||||
self.assertEqual(baseline, export_line_count())
|
||||
|
||||
# Confirm → the sheet's rows enter the export.
|
||||
sheets = self.batch_sheets(self.cashier_a, batch_id)
|
||||
name = sheets[0]["sheet_name"]
|
||||
status, _, data = self.confirm(self.cashier_a, batch_id, [name])
|
||||
self.assertEqual(200, status, data)
|
||||
self.assertGreater(export_line_count(), baseline)
|
||||
|
||||
# A confirmed sheet cannot be silently revoked: the settled decision
|
||||
# is a conflict, not a silent data removal.
|
||||
status, _, data = self.ignore(self.cashier_a, batch_id, [name], "取消")
|
||||
self.assertEqual(409, status, data)
|
||||
self.assertGreater(export_line_count(), baseline)
|
||||
|
||||
# A sheet that was only ever ignored never enters the export.
|
||||
status, _, data = self.upload_for_a(multi_sheet_bytes(), filename="忽略.xlsx")
|
||||
batch2 = as_json(data)["batch_id"]
|
||||
parsed = [
|
||||
sheet for sheet in self.batch_sheets(self.cashier_a, batch2)
|
||||
if sheet["outcome"] == "parsed"
|
||||
]
|
||||
self.assertTrue(parsed)
|
||||
before_ignore = export_line_count()
|
||||
status, _, data = self.ignore(
|
||||
self.cashier_a, batch2, [parsed[0]["sheet_name"]], "该表复核后不采用"
|
||||
)
|
||||
self.assertEqual(200, status, data)
|
||||
self.assertEqual(before_ignore, export_line_count())
|
||||
|
||||
# ------------------------------------------------------------------
|
||||
# Concurrency
|
||||
# ------------------------------------------------------------------
|
||||
|
||||
def test_concurrent_duplicate_upload_keeps_one_fact_set(self) -> None:
|
||||
content = CITIC_SAMPLE.read_bytes()
|
||||
results: list[tuple[int, dict]] = []
|
||||
errors: list[Exception] = []
|
||||
barrier = threading.Barrier(2)
|
||||
|
||||
def upload() -> None:
|
||||
try:
|
||||
client = self.fresh_client()
|
||||
client.cookies.update(self.cashier_a.cookies)
|
||||
barrier.wait(timeout=10)
|
||||
status, _, data = client.post_multipart(
|
||||
"/api/parse", {}, "并发.xlsx", content
|
||||
)
|
||||
results.append((status, as_json(data)))
|
||||
except Exception as exc: # pragma: no cover
|
||||
errors.append(exc)
|
||||
|
||||
threads = [threading.Thread(target=upload) for _ in range(2)]
|
||||
for thread in threads:
|
||||
thread.start()
|
||||
for thread in threads:
|
||||
thread.join(timeout=30)
|
||||
|
||||
self.assertEqual([], errors)
|
||||
self.assertEqual(2, len(results))
|
||||
statuses = sorted(payload["status"] for _, payload in results)
|
||||
self.assertEqual(["duplicate", "parsed"], statuses)
|
||||
|
||||
connection = connect(self.db_path)
|
||||
try:
|
||||
# The concurrent upload adds exactly one source file: identical
|
||||
# bytes never create a second fact set.
|
||||
count = connection.execute(
|
||||
"SELECT COUNT(*) AS n FROM source_files WHERE original_filename = '并发.xlsx'"
|
||||
).fetchone()["n"]
|
||||
self.assertEqual(1, count)
|
||||
duplicates = connection.execute(
|
||||
"SELECT COUNT(*) AS n FROM import_batches WHERE status = 'duplicate'"
|
||||
).fetchone()["n"]
|
||||
self.assertGreaterEqual(duplicates, 1)
|
||||
finally:
|
||||
connection.close()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
unittest.main()
|
||||
Reference in New Issue
Block a user