json文件两种读取方式

json文件两种读取方式 这是一段两个人的对话标注抄本

[
    {
          
   
        "start_time": {
          
   
            "original": "0:00:00.611000"
        },
        "end_time": {
          
   
            "original": "0:00:05.760000"
        },
        "words": "+",
        "speaker": "",
        "location": "",
        "session_id": "MDT_F2F_223"
    },
    {
          
   
        "start_time": {
          
   
            "original": "0:00:05.760000"
        },
        "end_time": {
          
   
            "original": "0:00:09.755000"
        },
        "words": "不想上了，你知道吗？我都烦死了",
        "speaker": "SPK506",
        "location": "",
        "session_id": "MDT_F2F_223"
    },
    ...
    ]

json_file =/data/magic_data/transcription/train/MDT_F2F_223.json
cnt = 0
with open(json_file, r) as f:
    lines = f.readlines()
    print(lines)
    for line in lines:
        print(line)
        if cnt == 3:
            break
        else:
            cnt += 1

[[
,     {
,         "start_time": {
,             "original": "0:00:00.611000"
,         },
,         "end_time": {
,             "original": "0:00:05.760000"
,         },
,         "words": "+",
,         "speaker": "",
,         "location": "",
,         "session_id": "MDT_F2F_223"
,     },
,     {
,         "start_time": {
,             "original": "0:00:05.760000"
,         },
,         "end_time": {
,             "original": "0:00:09.755000"
,         },
,         "words": "不想上了，你知道吗？我都烦死了",
,         "speaker": "SPK506",
,         "location": "",
,         "session_id": "MDT_F2F_223"
,     },
,     {
,         "start_time": {
,             "original": "0:00:09.755000"
,         },
,         "end_time": {
,             "original": "0:00:10.376000"
,         },
,         "words": "[*]",
,         "speaker": "",
,         "location": "",
,         "session_id": "MDT_F2F_223"
,     },
,     {
,         "start_time": {
,             "original": "0:00:10.376000"
,         },
,         "end_time": {
,             "original": "0:00:11.995000"
,         },
,         "words": "诶，那你考研想考哪儿",
,    ]]

[

    {
          
   

        "start_time": {
          
   

            "original": "0:00:00.611000"

这是打印出来效果 原始文件中一行就是list一个item

如何读取json文件

方式1（常用）

import json
json_file =/data/magic_data/transcription/train/MDT_F2F_223.json
with open(json_file, r) as f:
    data1 = json.load(f)

用json模块load方法，load的参数是json文件对象

方式2

json_file =/data/magic_data/transcription/train/MDT_F2F_223.json
with open(json_file, r) as f:
    lines = f.readlines()
    json_str = .join(lines).replace(
, ).replace( , ).replace(,}, })
    data2 = json.loads(json_str)

lines得到一个字符串列表，列表每一项的值就是json文件中一行，调用json的loads方法，传入的参数是字符串, json_str = ‘’.join(lines)可以简写成这样，不用替换换行和空格字符也行

以上两种读取方式都行，一般习惯用第一种，第二种也比较浪费内存。

返回数据

上面两种方式读取的json文件，返回一个字典列表，其实和打印原json文件长得差不多，打印data2或者data1显示如下

[{
          
   start_time: {
          
   original: 0:00:00.611000},
  end_time: {
          
   original: 0:00:05.760000},
  words: +,
  speaker: ,
  location: ,
  session_id: MDT_F2F_223},
 {
          
   start_time: {
          
   original: 0:00:05.760000},
  end_time: {
          
   original: 0:00:09.755000},
  words: 不想上了，你知道吗？我都烦死了,
  speaker: SPK506,
  location: ,
  session_id: MDT_F2F_223},
 {
          
   start_time: {
          
   original: 0:00:09.755000},
  end_time: {
          
   original: 0:00:10.376000},
  words: [*],
  speaker: ,
  location: ,
  session_id: MDT_F2F_223}]