39 lines
1.2 KiB
Python
39 lines
1.2 KiB
Python
|
import json
|
||
|
|
import six
|
||
|
|
|
||
|
|
from .utils import BaseParser
|
||
|
|
|
||
|
|
|
||
|
|
class Parser(BaseParser):
|
||
|
|
"""Extract all of the string values of a json file (no keys as those
|
||
|
|
are, in some sense, markup). This is useful for parsing content
|
||
|
|
from mongodb dumps, for example.
|
||
|
|
"""
|
||
|
|
|
||
|
|
def extract(self, filename, **kwargs):
|
||
|
|
with open(filename, 'r') as raw:
|
||
|
|
deserialized_json = json.load(raw)
|
||
|
|
return self.get_text(deserialized_json)
|
||
|
|
|
||
|
|
def get_text(self, deserialized_json):
|
||
|
|
"""Recursively get text from subcomponents of a deserialized json. To
|
||
|
|
enforce the same order on the documents, make sure to read keys of
|
||
|
|
deserialized_json in a consistent (alphabetical) order.
|
||
|
|
"""
|
||
|
|
if isinstance(deserialized_json, dict):
|
||
|
|
result = ''
|
||
|
|
for key in sorted(deserialized_json):
|
||
|
|
result += self.get_text(deserialized_json[key]) + ' '
|
||
|
|
return result
|
||
|
|
|
||
|
|
if isinstance(deserialized_json, list):
|
||
|
|
result = ''
|
||
|
|
for item in deserialized_json:
|
||
|
|
result += self.get_text(item) + ' '
|
||
|
|
return result
|
||
|
|
|
||
|
|
if isinstance(deserialized_json, six.string_types):
|
||
|
|
return deserialized_json
|
||
|
|
else:
|
||
|
|
return ''
|