Files
AgentLab/backend/tests/fixtures/xlsx/lab.py
T
ThearchyHelios 2d5e50fa88 Excel 导入改造期 2:配方引擎与确认向导——表从哪一行哪一列开始、哪些是合计、日期怎么补年份,都写进一份可校验的配方;首次导入经人确认,之后每期按配方重放并核对
配方语言(agentlab-recipe/2,新 recipe_types、recipe)
- 封闭、不含正则、保存时静态校验:常量只能从表里的候选文字里选,占位符只能映射成空值,单位从固定词表里取
- 交叉表和列表两种版式;分段各自定位、不依赖先后顺序;合计按标签语义区间核对,关系恒等式、口径不同的声明都写在配方里
- 固定的解析器集合(parsers/2):全角、各类横线、千分位、月日写法宽容归一,规范写法存库、原文另记

执行器(新 xlsx_cells、recipe_engine)
- 两遍读取逐行对账,每个非空格都要有去向(格子账);交叉表逆透视成长表,按统计期补年份并断言覆盖整期
- 合计行改作核对并另存一张表,不和明细混在一列;区域外文字、隐藏行列、合并单元格、空行之后的文字都单独报出
- STRICT 加主键写临时库;按锚点段记溯源,供后续证据下钻用;大表流式读取,扫描在 sheetData 处停止

校验与说明(新 recipe_checks、recipe_notes)
- SQL 核对分通过、不一致、无法核对三种结果,失败分结构、数据质量、需确认三类;结构类直接拒收
- 数据质量类可以写理由接受,接受不改库、只改状态和说明
- 表的说明按核对结果生成、不含数字,经过数字检查;没核对过的不写「已核对」

起草(新 recipe_suggest、recipe_ai)
- 先用规则起草:建议卡片加待确认问题(关系要不要登记、占位符表示什么)
- 规则起草不完整时才提示 AI 兜底:界面先展示要发出的全部文字,用户同意才发;只发结构,数字、隐藏行列、遮罩列不发,公式先脱敏;同意绑定模型和服务方;AI 只写配方,有限次修订,记 token 和费用;截断单独识别
- 无论规则还是 AI 起草,都必须人工确认才能启用

导入流程(新 recipe_imports、recipe_confirm、recipe_diff、api/table_imports)
- 暂存 → 起草 → 回答问题或改配方 → 试运行 → 勾确认项、写接受理由 → 确认并启用;接入期 1 的构建、导入记录、快照和运行固定版本
- 上传新一期按已确认的配方重放,不调用模型,出回执和差异卡,手动启用
- 改配方报破坏性变更(breaking:*)、单位变化、工作表改名的兜底,都要逐项勾选确认

前端导入向导(新 pages/import)
- 原始网格(行列号、去向着色)、建议卡片、配方面板(表单,不露正则)、试运行回执、确认清单、差异卡、AI 起草的同意对话框
2026-10-01 22:00:07 +08:00

226 lines
10 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
"""9.3 的实验室难缠用例,按 xlsx-lab/case*.py 的结构重写成合成夹具:地区、产品、人名一律换成假名。
各函数返回 (xlsx 字节, 文件名),不写磁盘。openpyxl 写不出来的形态(公式缓存值、XlsxWriter 式的「缓存为 0 加
打开时重算」)在压缩包层面补。期望(规则草稿 → 试运行)见 P2-SPEC 9.3,WP-7 验收。
"""
from __future__ import annotations
from openpyxl import Workbook
from . import excel_saved, save
ZONES = ("分区甲", "分区乙", "分区丙")
PRODUCTS = ("产品甲", "产品乙", "产品丙")
C01_VARIANTS = ("literal", "uncached", "cached")
C05_VARIANTS = ("uncached", "zero_full_calc", "cached")
C08_VARIANTS = ("stacked", "gap", "side")
C10_MONTHS = (7, 8, 9, 10)
#: 9.3 表里的叫法 → 这里的 variant
_ALIASES = {"写死": "literal", "公式无缓存": "uncached", "公式有缓存": "cached", "有缓存": "cached",
"openpyxl": "uncached", "xlsxwriter": "zero_full_calc"}
def _variant(v: str, allowed: tuple[str, ...]) -> str:
v = _ALIASES.get(v, v)
if v not in allowed:
raise ValueError(f"variant 只能是 {allowed},收到 {v!r}")
return v
def c01(variant: str = "literal") -> tuple[bytes, str]:
"""表头在 C5,上方标题、单位、编制行,下方合计行、空行、备注。
literal:合计写死;uncached:合计是公式、没有缓存值(openpyxl 直接写出的样子);cached:补上缓存值。
期望:列表块 + total_row(pick=合计, keep_as);区域外含数字的文字只有 C1、C3 两格(outside_digits ×2)。
"""
variant = _variant(variant, C01_VARIANTS)
book = Workbook()
ws = book.active
ws.title = "月报"
ws["C1"] = "2026年8月 销售月报"
ws.merge_cells("C1:F1")
ws["C2"] = "单位:万元"
ws["C3"] = "编制:财务部 日期:2026-09-05"
for i, h in enumerate(("地区", "产品", "销量", "金额")):
ws.cell(5, 3 + i, h)
data = [(ZONES[0], PRODUCTS[0], 10, 100), (ZONES[1], PRODUCTS[1], 20, 120), (ZONES[2], PRODUCTS[2], 15, 80)]
for r, row in enumerate(data, start=6):
for i, v in enumerate(row):
ws.cell(r, 3 + i, v)
ws["C9"] = "合计"
cached: dict[str, object] = {}
if variant == "literal":
ws["E9"], ws["F9"] = 45, 300
else:
ws["E9"], ws["F9"] = "=SUM(E6:E8)", "=SUM(F6:F8)"
if variant == "cached":
cached = {"E9": 45, "F9": 300}
# 备注不含数字:这一格若带数字会多出一条 outside_digits,与 9.3 的期望(C1、C3 两格)不符
ws["C11"] = "注:数据来源于业务系统,金额含税。"
ws["C12"] = "制表人:经办甲"
return excel_saved(save(book), "月报", cached), f"c01_{variant}.xlsx"
def c04() -> tuple[bytes, str]:
"""合并单元格:A1:F1 标题;两行表头(地区、产品纵向合并,上下半年横向合并、下层销量 / 金额);地区纵向合并。"""
book = Workbook()
ws = book.active
ws.title = "合并"
ws["A1"] = "分地区销售(万元)"
ws.merge_cells("A1:F1")
ws["A3"], ws["B3"], ws["C3"], ws["E3"] = "地区", "产品", "2026年上半年", "2026年下半年"
for m in ("A3:A4", "B3:B4", "C3:D3", "E3:F3"):
ws.merge_cells(m)
for i, h in enumerate(("销量", "金额", "销量", "金额")):
ws.cell(4, 3 + i, h)
data = [(PRODUCTS[0], 1, 10, 2, 20), (PRODUCTS[1], 3, 30, 4, 40), (PRODUCTS[2], 5, 50, 6, 60),
(PRODUCTS[0], 7, 70, 8, 80), (PRODUCTS[1], 9, 90, 1, 10), (PRODUCTS[2], 2, 20, 3, 30)]
for r, row in enumerate(data, start=5):
for i, v in enumerate(row):
ws.cell(r, 2 + i, v)
ws["A5"], ws["A8"] = ZONES[0], ZONES[1]
ws.merge_cells("A5:A7")
ws.merge_cells("A8:A10")
return excel_saved(save(book), "合并", {}), "c04_merged.xlsx"
def c05(variant: str = "cached") -> tuple[bytes, str]:
"""公式与缓存值:金额列是公式(单价 × 数量),末行合计。
uncached:openpyxl 直接写出,公式没有缓存值;zero_full_calc:模拟 XlsxWriter 的默认写法,缓存值全是 0、
并设打开时重算(fullCalcOnLoad);cached:补上正确的缓存值(模拟 Excel 保存过)。
"""
variant = _variant(variant, C05_VARIANTS)
book = Workbook()
ws = book.active
ws.title = "明细"
rows = [("产品", "单价", "数量", "金额"), (PRODUCTS[0], 2.5, 4), (PRODUCTS[1], 3, 10), (PRODUCTS[2], 10, 1)]
for row in rows:
ws.append(list(row))
for r in (2, 3, 4):
ws.cell(r, 4, f"=B{r}*C{r}")
ws["A5"], ws["D5"] = "合计", "=SUM(D2:D4)"
raw = save(book)
if variant == "uncached":
return excel_saved(raw, "明细", {}), "c05_uncached.xlsx"
if variant == "zero_full_calc":
return excel_saved(raw, "明细", {"D2": 0, "D3": 0, "D4": 0, "D5": 0}, full_calc=True), "c05_zero_full_calc.xlsx"
return excel_saved(raw, "明细", {"D2": 10, "D3": 30, "D4": 10, "D5": 50}), "c05_cached.xlsx"
def c06() -> tuple[bytes, str]:
"""隐藏:明细表第 4 行手动隐藏、C 列隐藏、自动筛选(地区 = 分区甲)隐藏其余行;另有 hidden、veryHidden 工作表。"""
book = Workbook()
ws = book.active
ws.title = "明细"
ws.append(["地区", "产品", "成本", "金额"])
regions = [ZONES[0], ZONES[1], ZONES[0], ZONES[2], ZONES[0], ZONES[1], ZONES[0], ZONES[2], ZONES[0], ZONES[1]]
for i, reg in enumerate(regions):
ws.append([reg, f"型号{'甲乙丙丁戊己庚辛壬癸'[i]}", i, 10 * (i + 1)])
ws.row_dimensions[4].hidden = True
ws.column_dimensions["C"].hidden = True
ws.auto_filter.ref = "A1:D11"
ws.auto_filter.add_filter_column(0, [ZONES[0]])
for r in range(2, 12): # Excel 保存筛选结果时就是给不满足条件的行打 hidden
if ws.cell(r, 1).value != ZONES[0]:
ws.row_dimensions[r].hidden = True
draft = book.create_sheet("草稿")
draft.append(["甲", "乙"])
draft.append([1, 2])
draft.sheet_state = "hidden"
conf = book.create_sheet("配置")
conf.append(["键", "值"])
conf.append(["参数甲", "示例值"])
conf.sheet_state = "veryHidden"
return excel_saved(save(book), "明细", {}), "c06_hidden.xlsx"
def c08(variant: str = "stacked") -> tuple[bytes, str]:
"""同一工作表里的多张表。stacked:上下两张(各有小标题);gap:一张表中间一行空白;side:左右两张、隔一空列。"""
variant = _variant(variant, C08_VARIANTS)
book = Workbook()
ws = book.active
if variant == "stacked":
ws.title = "汇总"
ws["B1"] = "一、销售"
for r, row in enumerate([("地区", "销量", "金额"), (ZONES[0], 10, 100), (ZONES[1], 20, 120),
(ZONES[2], 15, 80)], start=2):
for c, v in enumerate(row, start=2):
ws.cell(r, c, v)
ws["B7"] = "二、费用"
for r, row in enumerate([("部门", "费用"), ("部门甲", 5), ("部门乙", 7), ("部门丙", 8)], start=8):
for c, v in enumerate(row, start=2):
ws.cell(r, c, v)
sheet = "汇总"
elif variant == "gap":
ws.title = "明细"
for row in [("地区", "产品", "金额"), (ZONES[0], PRODUCTS[0], 10), (ZONES[0], PRODUCTS[1], 20),
(None, None, None), (ZONES[1], PRODUCTS[0], 30), (ZONES[1], PRODUCTS[1], 40)]:
ws.append(list(row))
sheet = "明细"
else:
ws.title = "并排"
for r, (a, b) in enumerate([("地区", "金额"), (ZONES[0], 1), (ZONES[1], 2)], start=1):
ws.cell(r, 1, a)
ws.cell(r, 2, b)
for r, (a, b) in enumerate([("部门", "费用"), ("部门甲", 5), ("部门乙", 7)], start=1):
ws.cell(r, 4, a)
ws.cell(r, 5, b)
sheet = "并排"
return excel_saved(save(book), sheet, {}), f"c08_{variant}.xlsx"
_C10_DATA = ((ZONES[0], PRODUCTS[0], 10, 100), (ZONES[1], PRODUCTS[1], 20, 120), (ZONES[2], PRODUCTS[2], 15, 80))
def c10(month: int) -> tuple[bytes, str]:
"""版式漂移(以 7 月起草并启用):
7 月:表头在 C5:地区 产品 销量 金额,下面合计行;
8 月:表头挪到 C7(上面多两行说明),产品和销量之间多「单价」列 → column_extra;
9 月:表头在 B6,外观差异「地区 」「销量\\n」「金 额」→ 通过、差异卡 header_writing;
10 月:「金额」改名「销售额」→ column_missing 加 column_extra。
"""
if month not in C10_MONTHS:
raise ValueError(f"month 只能是 {C10_MONTHS}")
top, left, headers, extra_rows, with_price = {
7: (5, 3, ["地区", "产品", "销量", "金额"], 0, False),
8: (7, 3, ["地区", "产品", "单价", "销量", "金额"], 2, True),
9: (6, 2, ["地区 ", "产品", "销量\n", "金 额"], 1, False),
10: (5, 3, ["地区", "产品", "销量", "销售额"], 0, False),
}[month]
book = Workbook()
ws = book.active
ws.title = "月报"
ws.cell(1, left, "销售月报")
ws.cell(2, left, "单位:万元")
for i in range(extra_rows): # 说明行不含数字:只让表头的变化产生差异
ws.cell(3 + i, left, f"说明{'甲乙丙'[i]}:口径同上月")
for j, h in enumerate(headers):
ws.cell(top, left + j, h)
for r, (zone, prod, qty, amt) in enumerate(_C10_DATA, start=top + 1):
vals = [zone, prod] + ([amt / qty] if with_price else []) + [qty, amt]
for j, v in enumerate(vals):
ws.cell(r, left + j, v)
tot = top + 1 + len(_C10_DATA)
ws.cell(tot, left, "合计")
ws.cell(tot, left + len(headers) - 2, sum(d[2] for d in _C10_DATA))
ws.cell(tot, left + len(headers) - 1, sum(d[3] for d in _C10_DATA))
return excel_saved(save(book), "月报", {}), f"c10_2026-{month:02d}.xlsx"
def kv_form() -> tuple[bytes, str]:
"""键值表单式的版式(规则起草认不出、交给 AI 兜底起草的用例,9.7 第 10 步):两列「项目:值」,没有表头行。"""
book = Workbook()
ws = book.active
ws.title = "基本情况"
ws["A1"] = "基本情况登记表"
items = [("单位名称", "示例单位甲"), ("所在分区", ZONES[0]), ("从业人数", 120), ("营业面积", 860),
("年度客流", 35600), ("负责人", "经办甲")]
for r, (k, v) in enumerate(items, start=3):
ws.cell(r, 1, k)
ws.cell(r, 2, v)
return excel_saved(save(book), "基本情况", {}), "kv_form.xlsx"