52 lines
1.5 KiB
Python
Executable File
52 lines
1.5 KiB
Python
Executable File
#!/usr/bin/env python3
|
||
"""从题目文件导入题目。
|
||
|
||
用法：python3 import_quiz.py <file_path> <document_id> <user_id>
|
||
|
||
输出：JSON 格式的 prompt（system_prompt + user_prompt），由 agent 发给 LLM 解析题目。
|
||
LLM 返回题目 JSON 后，agent 应调用 service.import_questions() 导入。
|
||
"""
|
||
from pathlib import Path
|
||
import json
|
||
import sys
|
||
|
||
sys.path.append(str(Path(__file__).resolve().parents[1] / "src"))
|
||
|
||
from quiz_mastery.file_parser import parse_file
|
||
from quiz_mastery.quiz_extractor import build_extraction_prompt
|
||
|
||
|
||
def main() -> None:
|
||
if len(sys.argv) < 4:
|
||
print("Usage: import_quiz.py <file_path> <document_id> <user_id>")
|
||
print(" file_path: Path to question file (.md, .txt, .text)")
|
||
print(" document_id: Identifier for this document")
|
||
print(" user_id: User identifier")
|
||
sys.exit(1)
|
||
|
||
file_path = sys.argv[1]
|
||
document_id = sys.argv[2]
|
||
user_id = sys.argv[3]
|
||
|
||
content = parse_file(file_path)
|
||
prompts = build_extraction_prompt(content)
|
||
|
||
output = {
|
||
"action": "import_questions",
|
||
"document_id": document_id,
|
||
"user_id": user_id,
|
||
"file_path": file_path,
|
||
"prompts": prompts,
|
||
"instructions": (
|
||
"Send the system_prompt and user_prompt to an LLM. "
|
||
"The LLM should return a JSON array of questions. "
|
||
"Then call service.import_questions(document_id, user_id, questions) to import."
|
||
),
|
||
}
|
||
|
||
print(json.dumps(output, ensure_ascii=False, indent=2))
|
||
|
||
|
||
if __name__ == "__main__":
|
||
main()
|