Read top-level JSON dictionary incrementally using Python ijson

问题

I have the following data in my JSON file:

{
    "first": {
        "name": "James",
        "age": 30
    },
    "second": {
        "name": "Max",
        "age": 30
    },
    "third": {
        "name": "Norah",
        "age": 30
    },
    "fourth": {
        "name": "Sam",
        "age": 30
    }
}

I want to print the top-level key and object as follows:

import json
import ijson

fname = "data.json"

with open(fname) as f:
    raw_data = f.read()

data = json.loads(raw_data)

for k in data.keys():
    print k, data[k]

OUTPUT:

second {u'age': 30, u'name': u'Max'}
fourth {u'age': 30, u'name': u'Sam'}
third {u'age': 30, u'name': u'Norah'}
first {u'age': 30, u'name': u'James'}

So, far so good. However if I want to this same thing for a huge file, I would have to read it all in-memory. This very slow and requires lots of memory.

I want use an incremental JSON parser ( ijson in this case ) to achieve what I described earlier:

The above code was taken from: No access to top level elements with ijson?

with open(fname) as f:
    json_obj = ijson.items(f,'').next()  # '' loads everything as only one object.
    for (key, value) in json_obj.items():
        print key + " -> " + str(value)

This is not suitable either, because it also reads the whole file in memory. This not truly incremental.

How can I do incremental parsing of top-level keys and corresponding objects, of a JSON file in Python?

回答1:

Since essentially json files are text files, consider stripping the top level as string. Basically, use a read file iterable approach where you concatenate a string with each line and then break out of the loop once the string contains the double braces }} signaling the end of the top level. Of course the double brace condition must strip out spaces and line breaks.

toplevelstring = ''

with open('data.json') as f:    
    for line in f:
        if not '}}' in toplevelstring.replace('\n', '').replace('\s+',''):
            toplevelstring = toplevelstring + line
        else:
            break

data = json.loads(toplevelstring)

Now if your larger json is wrapped in square brackets or other braces, still run above routine but add the below line to slice out first character, [, and last two characters for comma and line break after top level's final brace:

[{
    "first": {
        "name": "James",
        "age": 30
    },
    "second": {
        "name": "Max",
        "age": 30
    },
    "third": {
        "name": "Norah",
        "age": 30
    },
    "fourth": {
        "name": "Sam",
        "age": 30
    }
},
{
    "data1": {
        "id": "AAA",
        "type": 55
    },
    "data2": {
        "id": "BBB",
        "type": 1601
    },
    "data3": {
        "id": "CCC",
        "type": 817
    }
}]

...

toplevelstring = toplevelstring[1:-2]
data = json.loads(toplevelstring)

回答2:

Answer from github issue [file name changed]

import ijson
from ijson.common import ObjectBuilder


def objects(file):
    key = '-'
    for prefix, event, value in ijson.parse(file):
        if prefix == '' and event == 'map_key':  # found new object at the root
            key = value  # mark the key value
            builder = ObjectBuilder()
        elif prefix.startswith(key):  # while at this key, build the object
            builder.event(event, value)
            if event == 'end_map':  # found the end of an object at the current key, yield
                yield key, builder.value


for key, value in objects(open('data.json', 'rb')):
    print(key, value)

来源：https://stackoverflow.com/questions/36766532/read-top-level-json-dictionary-incrementally-using-python-ijson

标签

python

json

ijson