42 lines
1.3 KiB
Python
42 lines
1.3 KiB
Python
#step0: 将文本拆分成问题和答案两部分
|
|
import json
|
|
|
|
input_file_path = '/home/ubuntu/50T/fsy/benchmark/dataset-ours/[Solution]qa_segment_all.json'
|
|
with open(input_file_path, 'r', encoding='utf-8') as infile:
|
|
data = json.load(infile)
|
|
|
|
# 遍历并处理数据
|
|
processed_data = []
|
|
for item in data:
|
|
segment = item.get("segment", "")
|
|
if "Solution" in segment:
|
|
question, answer = segment.split("Solution", 1)
|
|
question = question.strip()
|
|
answer = answer.strip()
|
|
processed_data.append({
|
|
"idx": item.get("idx"),
|
|
"question": question,
|
|
"answer": answer,
|
|
})
|
|
elif "Answer" in segment:
|
|
question, answer = segment.split("Answer", 1)
|
|
question = question.strip()
|
|
answer = answer.strip()
|
|
processed_data.append({
|
|
"idx": item.get("idx"),
|
|
"question": question,
|
|
"answer": answer,
|
|
})
|
|
else:
|
|
processed_data.append({
|
|
"idx": item.get("idx"),
|
|
"question": "000",
|
|
"answer": "000",
|
|
})
|
|
|
|
output_file_path = '[Solution]qa_segment.json'
|
|
with open(output_file_path, 'w', encoding='utf-8') as outfile:
|
|
json.dump(processed_data, outfile, ensure_ascii=False, indent=4)
|
|
|
|
output_file_path
|