兼容单科和总分

README.md

@@ -1,6 +1,6 @@
 # 指标测算 · 相邻层次法
 
-B/S 系统：上传历次大考总分，按方向内名次把学生分到固定宽度层次；每次只比较相邻两次考试的“上一层次 → 本次到线”命中人数；汇总后做拟合概率，再用最后一次考试各班层次人数摊派尖子线和特控线指标。
+B/S 系统：上传历次大考总分或单科分，按方向内名次把学生分到固定宽度层次；每次只比较相邻两次考试的“上一层次 → 本次到线”命中人数；汇总后做拟合概率，再用最后一次考试各班层次人数摊派尖子线和特控线指标。
 
 ## 启动
 
@@ -18,12 +18,13 @@ python3 -m app
 ## 成绩表要求
 
 - Excel `.xlsx`
-- 至少包含：学号/姓名、班级、多次考试总分
+- 至少包含：学号/姓名、班级、多次考试成绩列
+- 成绩列可以是总分（如 `高一下期末-总分`），也可以是单科分（如 `高一下期末-数学分`、`高三分班合成-语文分`）
 - 建议包含：方向、选科、学籍、异动
 - 物理/历史分表可直接识别；同分按同名次，后续名次递增
 - 未排除的缺考学生按 0 分参加同方向排名
 
-同目录 `basedata.xlsx` 上传后可作为按新算法复算的示例。
+同目录 `basedata.xlsx` 是总分示例；`2024级高三指标数核算表：数学（六次）.xlsx`、`2024级高三指标数核算表：语文（六次）.xlsx` 可作为单科示例。
 
 ## Web 设置
 

app/excel_parse.py

@@ -1,10 +1,12 @@
 from __future__ import annotations
 
+import ast
 import re
 from io import BytesIO
 from pathlib import Path
 
 from openpyxl import load_workbook
+from openpyxl.utils import column_index_from_string
 
 
 META_ALIASES = {
@@ -18,9 +20,12 @@ META_ALIASES = {
 }
 
 EXAM_HINT = re.compile(
-    r"(总分|期末|联考|月考|模考|开学|学考|诊断|调研|高考)",
+    r"(总分|期中|期末|联考|月考|模考|开学|学考|诊断|调研|高考|分班合成)",
     re.I,
 )
+SUBJECT_SUFFIX = re.compile(
+    r"-(语文|数学|英语|外语|物理|化学|生物|历史|地理|政治|思想政治)分$"
+)
 SKIP_SHEETS = ("说明", "readme", "目录")
 
 
@@ -32,6 +37,8 @@ def _is_skip_sheet(name: str) -> bool:
 
 ART_RE = re.compile(r"艺术|体育|艺体")
 LEAVE_RE = re.compile(r"请假|出国|休学|国际|不在校|港澳|港籍.*假")
+CELL_REF_RE = re.compile(r"\$?([A-Za-z]+)\$?(\d+)")
+MISSING_TOKENS = {"", "缺考", "—", "-", "/", "none", "nan"}
 
 
 def _norm(v) -> str:
@@ -42,13 +49,97 @@ def _norm(v) -> str:
     return str(v).strip()
 
 
+def _safe_arith(expr: str) -> float:
+    tree = ast.parse(expr, mode="eval")
+    allowed = (
+        ast.Expression,
+        ast.BinOp,
+        ast.UnaryOp,
+        ast.Add,
+        ast.Sub,
+        ast.Mult,
+        ast.Div,
+        ast.USub,
+        ast.UAdd,
+        ast.Constant,
+        ast.Load,
+        ast.Mod,
+    )
+    for node in ast.walk(tree):
+        if not isinstance(node, allowed):
+            raise ValueError("unsupported formula")
+    return float(eval(compile(tree, "<score-formula>", "eval"), {"__builtins__": {}}, {}))
+
+
+def _eval_simple_formula(formula: str, raw, row_number: int) -> float | None:
+    if not isinstance(formula, str) or not formula.startswith("="):
+        return None
+    expr = formula[1:].strip()
+    if re.search(r"[^0-9A-Za-z+\-*/().$,\s]", expr):
+        return None
+
+    def repl(match: re.Match) -> str:
+        col = column_index_from_string(match.group(1).upper()) - 1
+        src_row = int(match.group(2))
+        if src_row != row_number or col < 0 or col >= len(raw):
+            raise ValueError("formula refers outside the current row")
+        val = raw[col]
+        if isinstance(val, str) and val.startswith("="):
+            raise ValueError("nested formula")
+        if val is None or str(val).strip().lower() in MISSING_TOKENS:
+            raise ValueError("missing ref")
+        return str(float(val))
+
+    try:
+        filled = CELL_REF_RE.sub(repl, expr)
+        return _safe_arith(filled)
+    except (TypeError, ValueError, SyntaxError, ZeroDivisionError):
+        return None
+
+
+def _to_score(val, raw, row_number: int):
+    if val is None:
+        return None
+    if isinstance(val, str):
+        text = val.strip()
+        if text.lower() in MISSING_TOKENS:
+            return None
+        if text.startswith("="):
+            return _eval_simple_formula(text, raw, row_number)
+        val = text
+    try:
+        return float(val)
+    except (TypeError, ValueError):
+        return None
+
+
 def _is_exam_header(text: str) -> bool:
     t = text.strip()
-    if not t or t in {"考试次数"}:
+    if not t or t in {"考试次数"} or "次数" in t:
         return False
-    if t.endswith("-总分") or t.endswith("总分"):
+    # 总分、数学分、语文分、分班合成-数学分 等都视为成绩列。
+    if t.endswith("分"):
         return True
-    return bool(EXAM_HINT.search(t)) and "次数" not in t
+    return bool(EXAM_HINT.search(t))
+
+
+def infer_score_kind(exam_names: list[str], source_name: str = "") -> str:
+    found: list[str] = []
+    for name in exam_names:
+        m = SUBJECT_SUFFIX.search(str(name).strip())
+        if m:
+            found.append(m.group(1))
+        elif str(name).endswith("总分"):
+            found.append("总分")
+    uniq = list(dict.fromkeys(found))
+    if len(uniq) == 1:
+        return uniq[0]
+    if uniq and "总分" not in uniq:
+        return uniq[0]
+    for subj in ("语文", "数学", "英语", "外语", "化学", "生物", "地理", "政治"):
+        if subj in source_name:
+            return subj
+    return "总分"
 
 
 def _map_meta(header: str) -> str | None:
@@ -76,11 +167,12 @@ def _sheet_kind(name: str) -> str:
 
 
 def parse_workbook(path: Path | BytesIO) -> dict:
-    wb = load_workbook(path, data_only=True, read_only=True)
+    wb = load_workbook(path, data_only=False, read_only=True)
     sheets = []
     students: list[dict] = []
     exam_order: list[str] = []
     warnings: list[str] = []
+    source_name = path.name if isinstance(path, Path) else ""
 
     for ws in wb.worksheets:
         kind = _sheet_kind(ws.title)
@@ -112,7 +204,7 @@ def parse_workbook(path: Path | BytesIO) -> dict:
             continue
 
         count = 0
-        for raw in rows:
+        for row_number, raw in enumerate(rows, start=2):
             if raw is None or all(v is None or str(v).strip() == "" for v in raw):
                 continue
             rec = {
@@ -143,14 +235,7 @@ def parse_workbook(path: Path | BytesIO) -> dict:
             for exam, idx in exam_idx.items():
                 if idx >= len(raw):
                     continue
-                val = raw[idx]
+                rec["scores"][exam] = _to_score(raw[idx], raw, row_number)
-                if val is None or str(val).strip() in ("", "缺考", "—", "-", "/"):
-                    rec["scores"][exam] = None
-                    continue
-                try:
-                    rec["scores"][exam] = float(val)
-                except (TypeError, ValueError):
-                    rec["scores"][exam] = None
             rec["art"] = bool(ART_RE.search(rec["move"]) or ART_RE.search(ws.title))
             rec["leave"] = bool(LEAVE_RE.search(rec["move"]))
             students.append(rec)
@@ -195,12 +280,19 @@ def parse_workbook(path: Path | BytesIO) -> dict:
         warnings.append(f"按学号去重 {dups} 条")
 
     if not exam_order:
-        warnings.append("未识别到考试成绩列，请确认表头含「总分」或「期末/联考/月考」")
+        warnings.append("未识别到考试成绩列，请确认表头含「总分」或单科分（如「高一下期末-数学分」）")
+    else:
+        for exam in exam_order:
+            present = sum(1 for s in students if s["scores"].get(exam) is not None)
+            if present == 0:
+                warnings.append(f"考试列「{exam}」没有读到数值，若是未计算的公式请先在 Excel 中打开保存后再上传")
 
+    score_kind = infer_score_kind(exam_order, source_name)
     return {
         "sheets": sheets,
         "students": students,
         "exams": exam_order,
+        "score_kind": score_kind,
         "warnings": warnings,
     }
 
@@ -253,6 +345,7 @@ def summarize_parse(parsed: dict) -> dict:
         "score_scale": score_scale,
         "sheets": parsed["sheets"],
         "exams": exams,
+        "score_kind": parsed.get("score_kind") or infer_score_kind(exams),
         "warnings": parsed["warnings"],
     }
 

app/export_xlsx.py

@@ -36,6 +36,16 @@ def _safe_title(text: str) -> str:
     return text.replace(":", "-").replace("/", "-").replace("\\", "-").replace("?", "-")[:31]
 
 
+def _score_kind(result_or_group: dict) -> str:
+    return (result_or_group.get("score_kind") or "总分").strip() or "总分"
+
+
+def _line_title(group: dict, line_name: str) -> str:
+    kind = _score_kind(group)
+    prefix = "" if kind == "总分" else f"{kind}——"
+    return f"{prefix}{group['group']}方向{line_name}指标分配"
+
+
 def _raw_sheet(wb: Workbook, group: dict, students: list[dict]):
     title = _safe_title(f"{group['group']}方向人员")
     ws = wb.create_sheet(title)
@@ -149,7 +159,7 @@ def _metric_sheet(wb: Workbook, group: dict):
     ws = wb.create_sheet(title)
     r = 1
     for line in group["targets"]:
-        ws.cell(r, 1, f"{group['group']}方向{line['name']}指标分配")
+        ws.cell(r, 1, _line_title(group, line["name"]))
         r += 1
         headers = ["班级"] + [f"{i+1}层人数" for i in range(group["n_layers"])] + ["加权计算", "摊派指标"]
         ws.cell(r, 1, headers[0])
@@ -182,7 +192,11 @@ def _metric_sheet(wb: Workbook, group: dict):
         ws.cell(r, len(headers) - 1, line["sum_expected"])
         ws.cell(r, len(headers), line["sum_assigned"])
         r += 1
-        ws.cell(r, 1, "注：加权计算=Σ(各层人数×拟合概率)；摊派指标=加权计算四舍五入。")
+        last_exam = (group.get("_exams") or [None])[-1]
+        note = "注：加权计算=Σ(各层人数×拟合概率)；摊派指标=加权计算四舍五入。"
+        if last_exam:
+            note = f"注：数据源以最后一次考试「{last_exam}」为准。" + note[2:]
+        ws.cell(r, 1, note)
         r += 3
     ws.column_dimensions["A"].width = 14
     for c in range(2, ws.max_column + 1):
@@ -197,9 +211,11 @@ def build_export(result: dict) -> bytes:
     wb.remove(wb.active)
     # Keep the same familiar order as basedata.xlsx: 物理 raw/prob/metric, then 历史.
     groups = sorted(result["groups"], key=lambda g: _group_order(g["group"]))
+    score_kind = _score_kind(result)
     for i, group in enumerate(groups):
         students = [s for s in result["students"] if s["group"] == group["group"]]
         group["_exams"] = result["exams"]
+        group["score_kind"] = score_kind
         _raw_sheet(wb, group, students)
     for group in groups:
         _probability_sheet(wb, group)

app/layer_method.py

@@ -347,8 +347,10 @@ def analyze(parsed: dict, settings: dict) -> dict:
             }
         )
 
+    score_kind = parsed.get("score_kind") or "总分"
     return {
         "algorithm": "adjacent-layer-markov-v2",
+        "score_kind": score_kind,
         "exams": exams,
         "layer_size": LAYER_SIZE,
         "pair_count": len(probability_exams) - 1 if len(probability_exams) > 1 else 0,

app/templates/analysis.html

@@ -40,11 +40,12 @@
     <section class="card panel">
       {% if step == 1 %}
       <h1>第一步：成绩表体检</h1>
-      <p class="muted">这里只检查源表识别结果。缺考学生在后续测算中会按 0 分参加同方向排名。</p>
+      <p class="muted">这里只检查源表识别结果。缺考学生在后续测算中会按 0 分参加同方向排名。单科成绩与总分使用同一套相邻层次算法。</p>
       <div class="kpis">
         <div class="kpi"><em>识别学生</em><strong>{{ summary.n }}</strong></div>
         <div class="kpi"><em>班级</em><strong>{{ summary.n_class }}</strong></div>
         <div class="kpi"><em>考试次数</em><strong>{{ summary.exams|length }}</strong></div>
+        <div class="kpi"><em>成绩口径</em><strong>{{ summary.score_kind or '总分' }}</strong></div>
         <div class="kpi"><em>异动备注</em><strong>{{ summary.moves }}</strong></div>
       </div>
       {% for w in summary.warnings %}
@@ -82,7 +83,7 @@
       <h1>第二步：口径、层次与线</h1>
       <form method="post" action="/analysis/{{ item.id }}/settings">
         <h3>考试范围</h3>
-        <p class="tiny">考试列只来自本次上传工作簿的物理/历史人员表；只比较这里勾选后的相邻两次考试。例如勾选 7 次，会生成 6 组转移样本；最后一次既进入相邻概率计算，也用于最终摊派。</p>
+        <p class="tiny">考试列只来自本次上传工作簿的物理/历史人员表，总分或单科分均可；只比较这里勾选后的相邻两次考试。例如勾选 7 次，会生成 6 组转移样本；最后一次既进入相邻概率计算，也用于最终摊派。</p>
         <div class="explain" style="margin-bottom:10px"><b>本次源表</b>：{{ item.filename }}；方向人员表识别到 {{ summary.exams|length }} 个考试列。</div>
         {% for e in summary.exams %}
         <label class="check"><input type="checkbox" name="exams" value="{{ e }}" {% if e in settings.exams %}checked{% endif %}> {{ e }}（有分 {{ summary.exam_present[e] }} 人）</label>

app/templates/home.html

@@ -25,7 +25,7 @@
       </div>
       <form class="drop" action="/upload" method="post" enctype="multipart/form-data">
         <h3>上传成绩表</h3>
-        <p class="muted">xlsx，需含学号/姓名/班级，以及多次考试总分列。物理/历史分表会分开测算。</p>
+        <p class="muted">xlsx，需含学号/姓名/班级，以及多次考试成绩列（总分或单科分，如「高一下期末-数学分」）。物理/历史分表会分开测算。</p>
         <input type="file" name="file" accept=".xlsx,.xlsm" required>
         <div class="actions">
           <button class="primary" type="submit">开始逐步测算</button>