Cache for reading annotation

copienator/commands/add_final_score.py

@@ -2,11 +2,13 @@ import argparse
 import math
 import shutil
 import sys
+from decimal import Decimal, InvalidOperation, ROUND_CEILING
 from pathlib import Path
 
 import pandas as pd
 from PIL import Image, ImageDraw, ImageFont
 
+from copienator import atomic_write_json, read_json
 from copienator.configuration import (
     FINAL_SCORE_FONT_PATH,
     FINAL_SCORE_HISTOGRAM_PATH,
@@ -72,6 +74,39 @@ def copy_histogram(output_dir: Path, *, copy_merged_histogramme: bool = False) -
         shutil.copy2(source, destination)
         print(f"Copied histogram: {destination}")
 
+
+def write_score_statistics(search_path: Path, output_dir: Path) -> None:
+    """Average available per-question scores, rounding upward to one decimal."""
+    scores_by_label = {}
+    for path in sorted(search_path.glob("*/score.json")):
+        try:
+            scores = read_json(path)
+            if not isinstance(scores, dict):
+                raise ValueError("expected an object of per-label scores")
+        except (OSError, ValueError) as exc:
+            print(f"Warning: Could not read scores from '{path}': {exc}")
+            continue
+        for label, value in scores.items():
+            values = scores_by_label.setdefault(label, [])
+            if value is None or (isinstance(value, str) and not value.strip()):
+                continue
+            try:
+                number = Decimal(str(value))
+                if not number.is_finite():
+                    raise InvalidOperation
+            except InvalidOperation:
+                print(f"Warning: Invalid score '{value}' for '{label}' in '{path}'.")
+                continue
+            values.append(number)
+    statistics = {
+        label: str((sum(values) / len(values)).quantize(Decimal("0.1"), rounding=ROUND_CEILING))
+        if values else ""
+        for label, values in scores_by_label.items()
+    }
+    atomic_write_json(output_dir / "stat.json", statistics)
+    print(f"Wrote per-label averages: {output_dir / 'stat.json'}")
+
+
 def process_images(base_dir, output_dir, *, copy_merged_histogramme=False):
     # 1. Load Data
     try:
@@ -152,6 +187,7 @@ def process_images(base_dir, output_dir, *, copy_merged_histogramme=False):
         for pdf_path in pdf_paths:
             shutil.copy2(pdf_path, student_output / pdf_path.name)
 
+    write_score_statistics(search_path, output_dir)
     copy_histogram(output_dir, copy_merged_histogramme=copy_merged_histogramme)