兼容单科和总分

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]
-                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["scores"][exam] = _to_score(raw[idx], raw, row_number)
             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"],
     }