httpie-cli/httpie/output/utils.py
Mickaël Schoentgen e6c5cd3e4b
Improve JSON output when there is leading data before the actual JSON body (#1130)
In some special cases, to prevent against Cross Site Script Inclusion (XSSI)
attacks, the JSON response body starts with a magic prefix line that must be
stripped before feeding the rest of the response body to the JSON parser.
Such prefix is now simply ignored from the parser but still printed in the
terminal.

* Fix Windows tests
2021-09-21 11:15:43 +02:00

37 lines
951 B
Python

import json
import re
from typing import Tuple
from .lexers.json import PREFIX_REGEX
def load_prefixed_json(data: str) -> Tuple[str, json.JSONDecoder]:
"""Simple JSON loading from `data`.
"""
# First, the full data.
try:
return '', json.loads(data)
except ValueError:
pass
# Then, try to find the start of the actual body.
data_prefix, body = parse_prefixed_json(data)
try:
return data_prefix, json.loads(body)
except ValueError:
raise ValueError('Invalid JSON')
def parse_prefixed_json(data: str) -> Tuple[str, str]:
"""Find the potential JSON body from `data`.
Sometimes the JSON body is prefixed with a XSSI magic string, specific to the server.
Return a tuple (data prefix, actual JSON body).
"""
matches = re.findall(PREFIX_REGEX, data)
data_prefix = matches[0] if matches else ''
body = data[len(data_prefix):]
return data_prefix, body