Hugging Face's logo

Spaces:
shazhass
/
bacpilot-backend
Runtime error

App Files Community
bacpilot-backend
File size: 4,734 Bytes
d3ee77e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
#!/usr/bin/env python3
"""
Convertit le corpus enseignant JSON vers le CSV du workflow qualité BacPilot.

Usage:
    python3 scripts/convert_corpus_json_to_csv.py <corpus_dir> <output_csv>

Exemple:
    python3 scripts/convert_corpus_json_to_csv.py \
        /home/debpc/corpus_teacher_100 \
        quality/corpus/teacher_evaluation_cases_100.csv

Seules les copies avec status TEACHER_VALIDATED ou READY_FOR_EVAL sont exportées.
Les copies DRAFT sont ignorées.
"""

import csv
import json
import sys
from pathlib import Path


FIELDNAMES = [
    "submission_case_id", "exercise_id", "chapter", "skill",
    "student_answer_anonymized", "teacher_score", "teacher_max_score",
    "teacher_errors", "teacher_comment",
    "bacpilot_score", "bacpilot_max_score", "bacpilot_verdict",
    "bacpilot_confidence", "bacpilot_main_error",
    "absolute_score_error", "false_error_count", "missed_error_count",
    "needs_human_review_expected", "needs_human_review_bacpilot", "decision",
]

VERDICTS_REVIEW = {
    "PARTIELLEMENT_CORRECT",
    "METHODE_CORRECTE_RESULTAT_FAUX",
    "RESULTAT_CORRECT_JUSTIFICATION_INSUFFISANTE",
    "NON_EVALUABLE",
}


def extract_student_answer(path: Path) -> str:
    text = path.read_text(encoding="utf-8")
    if "Réponse :" in text:
        after = text.split("Réponse :")[-1].strip()
        if after.startswith("---"):
            after = after[3:].strip()
        return after
    return text.strip()


def convert(corpus_dir: Path, output_csv: Path) -> None:
    copies_dir = corpus_dir / "copies"
    if not copies_dir.exists():
        print(f"ERREUR: dossier copies introuvable dans {corpus_dir}")
        sys.exit(1)

    rows = []
    converted = 0
    skipped_draft = 0
    errors = []

    for copy_dir in sorted(copies_dir.iterdir()):
        if not copy_dir.is_dir():
            continue

        meta_path = copy_dir / "metadata.json"
        answer_path = copy_dir / "student_answer.md"
        correction_path = copy_dir / "teacher_correction.json"

        if not all(p.exists() for p in [meta_path, answer_path, correction_path]):
            errors.append(f"{copy_dir.name}: fichiers manquants")
            continue

        try:
            meta = json.loads(meta_path.read_text(encoding="utf-8"))
            correction = json.loads(correction_path.read_text(encoding="utf-8"))
        except json.JSONDecodeError as e:
            errors.append(f"{copy_dir.name}: JSON invalide — {e}")
            continue

        status = correction.get("status", "DRAFT")
        if status not in ("TEACHER_VALIDATED", "READY_FOR_EVAL"):
            skipped_draft += 1
            continue

        student_answer = extract_student_answer(answer_path)
        main_errors = correction.get("main_errors", [])
        verdict = correction.get("verdict", "")
        needs_review = "true" if verdict in VERDICTS_REVIEW else "false"

        rows.append({
            "submission_case_id": correction.get("copy_id", copy_dir.name),
            "exercise_id": correction.get("exercise_id", ""),
            "chapter": meta.get("chapter", ""),
            "skill": meta.get("skill", ""),
            "student_answer_anonymized": student_answer,
            "teacher_score": correction.get("teacher_score", ""),
            "teacher_max_score": correction.get("max_score", ""),
            "teacher_errors": ";".join(main_errors),
            "teacher_comment": correction.get("teacher_comment", ""),
            "bacpilot_score": "",
            "bacpilot_max_score": "",
            "bacpilot_verdict": "",
            "bacpilot_confidence": "",
            "bacpilot_main_error": "",
            "absolute_score_error": "",
            "false_error_count": "",
            "missed_error_count": "",
            "needs_human_review_expected": needs_review,
            "needs_human_review_bacpilot": "",
            "decision": "",
        })
        converted += 1

    output_csv.parent.mkdir(parents=True, exist_ok=True)
    with output_csv.open("w", newline="", encoding="utf-8") as f:
        writer = csv.DictWriter(f, fieldnames=FIELDNAMES)
        writer.writeheader()
        writer.writerows(rows)

    print(f"CONVERTED={converted}")
    print(f"SKIPPED_DRAFT={skipped_draft}")
    print(f"ERRORS={len(errors)}")
    print(f"OUTPUT={output_csv}")

    if errors:
        for e in errors:
            print(f"  ERREUR: {e}")

    if converted == 0:
        print("STATUS=NO_READY_COPIES — remplir les copies et passer status à TEACHER_VALIDATED")
    else:
        print("STATUS=OK")


if __name__ == "__main__":
    if len(sys.argv) != 3:
        print(f"Usage: {sys.argv[0]} <corpus_dir> <output_csv>")
        sys.exit(1)

    convert(Path(sys.argv[1]), Path(sys.argv[2]))