如何使用 Python 3 将 JSON 文件转换为字典

Question

first post here. Total newbie in Python & programming in general.

Basically I'm following this guy's tutorial (not querying the same API though) about extracting data from a JSON response in Python .

Here is my (very simple) code:

#!/usr/bin/env python
import json
import requests

url = 'http://calendar.voyages-sncf.com/cdp/api/proposals/v3/outward/FRPAR/FRGNL/2017-11-01/12-HAPPY_CARD/2/fr/fr?fareCodes=HC16&onlyDirectTrains=true'

json_data = requests.get(url).json()
json_segments = json_data['segments'] 

print (json_segments)

Now, instead of extracting the data marked with 'Segments' in the JSON file. I get this error:

json_segments = json_data['status']
TypeError: list indices must be integers or slices, not str

So, I've tried int(), to make it an integer but does not work.

Thank you very much if you can help.

Answer 1

The data returned by requests.get().json() is a list of data. You can iterate through the list by using a for data in json_data: loop. Then you can access the 'segments' key of the dict directly.

#!/usr/bin/env python
import requests

url = 'http://calendar.voyages-sncf.com/cdp/api/proposals/v3/outward/FRPAR/FRGNL/2017-11-01/12-HAPPY_CARD/2/fr/fr?fareCodes=HC16&onlyDirectTrains=true'

json_data = requests.get(url).json()

for data in json_data:
    print(data['segments'])

Regarding your question about the tutorial code:

import urllib.parse

import requests

address = 'lhr'
main_api = 'http://maps.googleapis.com/maps/api/geocode/json?'
url = main_api + urllib.parse.urlencode({'address': address})
json_data = requests.get(url).json()
json_status = json_data['status']
print(json_status)

The docs for the google geocoding api here (the api he is using in the video) specify that you will get a dictionary with the a specific architecture. However when you are getting your data from http://calendar.voyages-sncf.com/cdp/api they are not returning the same data structure. In their case it is a list of dictionaries.

So every time you get new data from a different api (or a different endpoint in the same api) you will get data in a different structure. Therefor you will need to change your code to work with each piece of data.

JSON is simply a standard way to change a dictionary, list, or other native (native to python, java, or what ever language you are using at the time) to and from a string after it is sent over a network. This can lead to problems. Specifically in python a set() can not be changed directly to JSON. It needs to be changed to a list then to JSON, then back to a list then back to a set on the other side.

The following example is was copied from the google docs here with minor modifications to make it smaller and more readable.

{
   "results" : [
      {
         "address_components" : [
            {
               "long_name" : "1600",
               "short_name" : "1600",
               "types" : [ "street_number" ]
            },
         ],
         "formatted_address" : "1600 Amphitheatre Parkway, Mountain View, CA 94043, USA",
         "geometry" : {
            "location" : {
               "lat" : 37.4224764,
               "lng" : -122.0842499
            },
            "location_type" : "ROOFTOP",
            "viewport" : {
               "northeast" : {
                  "lat" : 37.4238253802915,
                  "lng" : -122.0829009197085
               },
               "southwest" : {
                  "lat" : 37.4211274197085,
                  "lng" : -122.0855988802915
               }
            }
         },
         "place_id" : "ChIJ2eUgeAK6j4ARbn5u_wAGqWA",
         "types" : [ "street_address" ]
      }
   ],
   "status" : "OK"
}

Answer 2

Welcome to SO and glad u chose Python :).

If you are new to Python and intend to work with tables do not hesitate to learn pandas . There are tons of documents and questions about pandas.

Examples:

import pandas as pd
url = 'http://calendar.voyages-sncf.com/cdp/api/proposals/v3/outward/FRPAR/FRGNL/2017-11-01/12-HAPPY_CARD/2/fr/fr?fareCodes=HC16&onlyDirectTrains=true'
df = pd.read_json(url)

And

from pandas.io.json import json_normalize
df = pd.concat(json_normalize(df["segments"].values[i]) for i in range(len(df)))

With pandas you can output to list, dictionaries, csv, json and so on...

Here is a snippet that returns html table with first 3 rows:

print(df.head(3).to_html(index=False))

Result

 <table border="1" class="dataframe"> <thead> <tr style="text-align: right;"> <th>arrivalDate</th> <th>carrierCode</th> <th>departureDate</th> <th>destination.cityLabel</th> <th>destination.label</th> <th>destination.rrcode</th> <th>duration</th> <th>inventory</th> <th>onboardServices</th> <th>origin.cityLabel</th> <th>origin.label</th> <th>origin.rrcode</th> <th>physicalSpace</th> <th>quotations.12-HAPPY_CARD.cos</th> <th>quotations.12-HAPPY_CARD.cosLevel</th> <th>quotations.12-HAPPY_CARD.fareCode</th> <th>quotations.12-HAPPY_CARD.fareCondition.conditions</th> <th>quotations.12-HAPPY_CARD.fareCondition.fareName</th> <th>quotations.12-HAPPY_CARD.fareSequence</th> <th>quotations.12-HAPPY_CARD.fareSpecificRule</th> <th>quotations.12-HAPPY_CARD.passengerType</th> <th>trainNumber</th> <th>trainType</th> <th>transporter</th> <th>travelClass</th> </tr> </thead> <tbody> <tr> <td>2017-11-01T19:46</td> <td>SN</td> <td>2017-11-01T16:41</td> <td>Grenoble</td> <td>Grenoble</td> <td>FRGNB</td> <td>185</td> <td>wdi</td> <td>[BAR, HAN, SMP]</td> <td>Paris</td> <td>Paris Gare de Lyon</td> <td>FRPLY</td> <td>B</td> <td>BN</td> <td>17</td> <td>HC16</td> <td>[Pièce d'identité à présenter à bord du train....</td> <td>TGVmax</td> <td>None</td> <td>None</td> <td>PT00AD</td> <td>6921</td> <td>TGD</td> <td>tgv</td> <td>2</td> </tr> <tr> <td>2017-11-01T17:45</td> <td>SN</td> <td>2017-11-01T14:41</td> <td>Grenoble</td> <td>Grenoble</td> <td>FRGNB</td> <td>184</td> <td>wdi</td> <td>[BAR, HAN, SMP]</td> <td>Paris</td> <td>Paris Gare de Lyon</td> <td>FRPLY</td> <td>B</td> <td>BN</td> <td>17</td> <td>HC16</td> <td>[Pièce d'identité à présenter à bord du train....</td> <td>TGVmax</td> <td>None</td> <td>None</td> <td>PT00AD</td> <td>6919</td> <td>TGD</td> <td>tgv</td> <td>2</td> </tr> <tr> <td>2017-11-01T10:44</td> <td>SN</td> <td>2017-11-01T07:41</td> <td>Grenoble</td> <td>Grenoble</td> <td>FRGNB</td> <td>183</td> <td>wdi</td> <td>[VEP, BAR, HAN, SMP]</td> <td>Paris</td> <td>Paris Gare de Lyon</td> <td>FRPLY</td> <td>B</td> <td>BN</td> <td>17</td> <td>HC16</td> <td>[Pièce d'identité à présenter à bord du train....</td> <td>TGVmax</td> <td>None</td> <td>None</td> <td>PT00AD</td> <td>6905</td> <td>TGS</td> <td>tgv</td> <td>2</td> </tr> </tbody> </table>