1-3古诗内容提取json格式转jsonl

1、首先将json格式转换成jsonl格式文件，代码如下：

import json import jsonlines data = [] with open('poet.song.1000.json', 'r', encoding='utf-8') as file: value = json.load(file) data.append(value) with jsonlines.open('outputgusi.jsonl','w')as file1: file1.write_all(data)

import json
import jsonlines

data = []

with open('poet.song.1000.json', 'r', encoding='utf-8') as file:
    value = json.load(file)
    data.append(value)
with jsonlines.open('outputgusi.jsonl','w')as file1:
    file1.write_all(data)

2、读取jsonl格式文件

  根据文本内容分析，title为古诗的题目，paragraphs为古诗内容，其中title中有部分数据（首字母为句，不是古诗的名字，需要剔除），代码如下：

import json import jsonlines data = [] data_list = [] with open('outputgusi.jsonl', 'r', encoding='utf-8') as file: for line in file: data_list = json.loads(line) print (len(data_list)) for i in range(len(data_list)): if 'title' in data_list[i]: if data_list[i]['title'][0] !='句': value = {} value['instruction'] = '请根据古诗词的题目，回复古诗内容，要求对仗工整，押韵' value['input']=data_list[i]['title'] if 'paragraphs' in data_list[i]: value['output']=data_list[i]['paragraphs'] data.append(value) with jsonlines.open('outputgusi111.jsonl','w') as file: file.write_all(data)

import json
import jsonlines
data = []
data_list = []
with open('outputgusi.jsonl', 'r', encoding='utf-8') as file:
    for line in file:

        data_list = json.loads(line)
        print (len(data_list))
        for i in range(len(data_list)):

            if 'title' in data_list[i]:
                if data_list[i]['title'][0] !='句':
                    value = {}
                    value['instruction'] = '请根据古诗词的题目，回复古诗内容，要求对仗工整，押韵'
                    value['input']=data_list[i]['title']
                    if 'paragraphs' in data_list[i]:
                        value['output']=data_list[i]['paragraphs']
                    data.append(value)



with jsonlines.open('outputgusi111.jsonl','w') as file:
    file.write_all(data)