Question

我正在尝试将mongodb集合转换为json文件，稍后将相同的Json文件数据加载到另一个MongoDB集合。该系列有大约60,000行。我写了以下代码：

from pymongo import MongoClient
import json
from bson.json_util import dumps
from bson import json_util

with open("collections/review.json", "w") as f:
    l = list(reviews_collection.find())  
    json.dump(json.dumps(l,default=json_util.default),f,indent = 4)

# reviews_collection_bkp.remove()
reviews_collection_bkp.remove()
with open("collections/review.json") as dataset:
    for line in dataset:
            data = json.loads(line)
            reviews_collection_bkp.insert({
                 "reviewId": data["reviewId"],
                 "business": data["business"],
                 "text": data["text"],
                 "stars": data['stars'],
                 "votes":data["votes"]
             })
print reviews_collection_bkp.find().count()

review_collection是我想用Json文件名review.json编写的集合，后来想要从同一个文件中读取数据以将数据插入到MongoDB集合中。但我认为代码无法创建正确的json文件。因为在读取同一文件时会产生以下错误：

    "reviewId": data["reviewId"],
TypeError: string indices must be integers

为什么创建的Json文件格式不正确？

这是line和data的示例输出：

"[{\"votes\": {\"funny\": 0, \"useful\": 0, \"cool\": 0}, \"business\": \"wqu7ILomIOPSduRwoWp4AQ\", \"text\": \"Went for breakfast on 6/16/14. We received very good service and meal came within a few minutes.Waitress could have smiled more but was friendly. \\nI had a Grand Slam... it was more than enough food. \\nMeal was very tasty... We will definitely go back. \\nIt is a popular Denny's.\", \"reviewId\": \"0GS3S7UsRGI4B7ziy4cd7Q\", \"stars\": 4, \"_id\": {\"$oid\": \"5711d16fe396f81fcb51dc73\"}},...]

[{"votes": {"funny": 0, "useful": 0, "cool": 0}, "business": "wqu7ILomIOPSduRwoWp4AQ", "text": "Went for breakfast on 6/16/14. We received very good service and meal came within a few minutes.Waitress could have smiled more but was friendly. \nI had a Grand Slam... it was more than enough food. \nMeal was very tasty... We will definitely go back. \nIt is a popular Denny's.", "reviewId": "0GS3S7UsRGI4B7ziy4cd7Q", "stars": 4, "_id": {"$oid": "5711d16fe396f81fcb51dc73"}}......]

Answer 1

你确定文件的每一行都是有效的json吗？

我认为这是一种正确的方法：

with open("collections/review.json") as dataset:
    data = json.loads(dataset)
    for line in data:
        reviews_collection_bkp.insert({
             "reviewId": line['reviewId'],
             ...
         })

如果这不起作用，请尝试打印生成的json文件，以了解如何解码。

Answer 2

由于您的数据是字典列表，您需要遍历它。

for line in dataset:
    data = json.loads(line)
    for doc in data:
         reviews_collection_bkp.insert({
                 "reviewId": data["reviewId"],
                 "business": data["business"],
                 "text": data["text"],
                 "stars": data['stars'],
                 "votes":data["votes"]
             })

将MongoDB集合转换为Json文件，并从Python

2 个答案: