mirror of
https://github.com/LonamiWebs/Telethon.git
synced 2024-11-30 05:13:45 +03:00
148 lines
5.1 KiB
Python
148 lines
5.1 KiB
Python
import json
|
|
import re
|
|
from collections import defaultdict
|
|
|
|
from ..utils import snake_to_camel_case
|
|
|
|
# Core base classes depending on the integer error code
|
|
KNOWN_BASE_CLASSES = {
|
|
303: 'InvalidDCError',
|
|
400: 'BadRequestError',
|
|
401: 'UnauthorizedError',
|
|
403: 'ForbiddenError',
|
|
404: 'NotFoundError',
|
|
406: 'AuthKeyError',
|
|
420: 'FloodError',
|
|
500: 'ServerError',
|
|
}
|
|
|
|
# The API doesn't return the code for some (vital) errors. They are
|
|
# all assumed to be 400, except these well-known ones that aren't.
|
|
KNOWN_CODES = {
|
|
'ACTIVE_USER_REQUIRED': 401,
|
|
'AUTH_KEY_UNREGISTERED': 401,
|
|
'USER_DEACTIVATED': 401
|
|
}
|
|
|
|
# Give better semantic names to some captures
|
|
CAPTURE_NAMES = {
|
|
'FloodWaitError': 'seconds',
|
|
'FloodTestPhoneWaitError': 'seconds',
|
|
'TakeoutInitDelayError': 'seconds',
|
|
'FileMigrateError': 'new_dc',
|
|
'NetworkMigrateError': 'new_dc',
|
|
'PhoneMigrateError': 'new_dc',
|
|
'UserMigrateError': 'new_dc',
|
|
'FilePartMissingError': 'which'
|
|
}
|
|
|
|
|
|
def _get_class_name(error_code):
|
|
"""
|
|
Gets the corresponding class name for the given error code,
|
|
this either being an integer (thus base error name) or str.
|
|
"""
|
|
if isinstance(error_code, int):
|
|
return KNOWN_BASE_CLASSES.get(
|
|
error_code, 'RPCError' + str(error_code).replace('-', 'Neg')
|
|
)
|
|
|
|
return snake_to_camel_case(
|
|
error_code.replace('FIRSTNAME', 'FIRST_NAME').lower(), suffix='Error')
|
|
|
|
|
|
class Error:
|
|
def __init__(self, int_code, str_code, description, caused_by):
|
|
# TODO Some errors have the same str_code but different int_code
|
|
# Should these be split into different files or doesn't really matter?
|
|
# Telegram isn't exactly consistent with returned errors anyway.
|
|
self.int_code = int_code
|
|
self.str_code = str_code
|
|
self.subclass = _get_class_name(int_code)
|
|
self.subclass_exists = int_code in KNOWN_BASE_CLASSES
|
|
self.description = description
|
|
self.caused_by = list(sorted(caused_by))
|
|
|
|
self.has_captures = '_X' in str_code
|
|
if self.has_captures:
|
|
self.name = _get_class_name(str_code.replace('_X', ''))
|
|
self.pattern = str_code.replace('_X', r'_(\d+)')
|
|
self.capture_name = CAPTURE_NAMES.get(self.name, 'x')
|
|
else:
|
|
self.name = _get_class_name(str_code)
|
|
self.pattern = str_code
|
|
self.capture_name = None
|
|
|
|
|
|
def parse_errors(json_file, descriptions_file):
|
|
"""
|
|
Parses the given JSON file in the following format:
|
|
{
|
|
"ok": true,
|
|
"human_result": {"int_code": ["descriptions"]},
|
|
"result": {"int_code": {"full_method_name": ["str_error"]}}
|
|
}
|
|
|
|
The descriptions file, which has precedence over the JSON's human_result,
|
|
should have the following format:
|
|
# comment
|
|
str_error=Description
|
|
|
|
The method yields `Error` instances as a result.
|
|
"""
|
|
with open(json_file, 'r', encoding='utf-8') as f:
|
|
data = json.load(f)
|
|
|
|
errors = defaultdict(set)
|
|
error_to_method = defaultdict(set)
|
|
# PWRTelegram's API doesn't return all errors, which we do need here.
|
|
# Add some special known-cases manually first.
|
|
errors[420].update((
|
|
'FLOOD_WAIT_X', 'FLOOD_TEST_PHONE_WAIT_X', 'TAKEOUT_INIT_DELAY_X'
|
|
))
|
|
errors[401].update((
|
|
'AUTH_KEY_INVALID', 'SESSION_EXPIRED', 'SESSION_REVOKED'
|
|
))
|
|
errors[303].update((
|
|
'FILE_MIGRATE_X', 'PHONE_MIGRATE_X',
|
|
'NETWORK_MIGRATE_X', 'USER_MIGRATE_X'
|
|
))
|
|
for int_code, method_errors in data['result'].items():
|
|
for method, error_list in method_errors.items():
|
|
for error in error_list:
|
|
error = re.sub('_\d+', '_X', error).upper()
|
|
errors[int(int_code)].add(error)
|
|
error_to_method[error].add(method)
|
|
|
|
# Some errors are in the human result, but not with a code. Assume 400
|
|
for error in data['human_result']:
|
|
if error[0] != '-' and not error.isdigit():
|
|
error = re.sub('_\d+', '_X', error).upper()
|
|
if not any(error in es for es in errors.values()):
|
|
errors[KNOWN_CODES.get(error, 400)].add(error)
|
|
|
|
# Prefer the descriptions that are related with Telethon way of coding
|
|
# to those that PWRTelegram's API provides.
|
|
telethon_descriptions = {}
|
|
with open(descriptions_file, 'r', encoding='utf-8') as f:
|
|
for line in f:
|
|
line = line.strip()
|
|
if line and not line.startswith('#'):
|
|
equal = line.index('=')
|
|
message, description = line[:equal], line[equal + 1:]
|
|
telethon_descriptions[message.rstrip()] = description.lstrip()
|
|
|
|
for int_code, error_set in errors.items():
|
|
for str_code in sorted(error_set):
|
|
description = telethon_descriptions.get(
|
|
str_code, '\n'.join(data['human_result'].get(
|
|
str_code, ['No description known']
|
|
))
|
|
)
|
|
yield Error(
|
|
int_code=int_code,
|
|
str_code=str_code,
|
|
description=description,
|
|
caused_by=error_to_method[str_code]
|
|
)
|