chg: [telegram importer] add username correlation

pull/586/head
Terrtia 2022-05-05 16:58:05 +02:00
parent c81942d5dd
commit 3cafc5eae7
No known key found for this signature in database
GPG Key ID: 1E1B1F50D84613D0
2 changed files with 65 additions and 1 deletions

View File

@ -0,0 +1,59 @@
#!/usr/bin/env python3
# -*-coding:UTF-8 -*
"""
The JSON Receiver Module
================
Recieve Json Items (example: Twitter feeder)
"""
import os
import json
import sys
import datetime
sys.path.append(os.path.join(os.environ['AIL_BIN'], 'lib'))
import item_basic
import Username
sys.path.append(os.path.join(os.environ['AIL_BIN'], 'import', 'ail_json_importer'))
from Default_json import Default_json
class Ail_feeder_telegram(Default_json):
"""Twitter Feeder functions"""
def __init__(self, name, json_item):
super().__init__(name, json_item)
def get_feeder_name(self):
return 'telegram'
# define item id
def get_item_id(self):
# use twitter timestamp ?
item_date = datetime.date.today().strftime("%Y/%m/%d")
channel_id = str(self.json_item['meta']['channel_id'])
message_id = str(self.json_item['meta']['message_id'])
item_id = f'{channel_id}_{message_id}'
return os.path.join('telegram', item_date, item_id) + '.gz'
def process_json_meta(self, process, item_id):
'''
Process JSON meta filed.
'''
channel_id = str(self.json_item['meta']['channel_id'])
message_id = str(self.json_item['meta']['message_id'])
telegram_id = f'{channel_id}_{message_id}'
item_basic.add_map_obj_id_item_id(telegram_id, item_id, 'telegram_id')
#print(self.json_item['meta'])
username = None
if self.json_item['meta'].get('user'):
username = str(self.json_item['meta']['user'])
else:
if self.json_item['meta'].get('channel'):
username = str(self.json_item['meta']['channel']['username'])
if username:
#print(username)
item_date = item_basic.get_item_date(item_id)
Username.save_item_correlation('telegram', username, item_id, item_date)
return None

View File

@ -116,6 +116,9 @@ def is_domain_root(item_id):
else:
return True
def get_item_url(item_id):
return r_serv_metadata.hget(f'paste_metadata:{item_id}', 'real_link')
def get_nb_children(item_id):
return r_serv_metadata.scard('paste_children:{}'.format(item_id))
@ -166,7 +169,7 @@ def add_item_parent(parent_item_id, item_id):
#### UNKNOW SECTION ####
def get_obj_id_item_id(parent_type, parent_id):
all_parents_type = ['twitter_id', 'jabber_id']
all_parents_type = ['twitter_id', 'jabber_id', 'telegram_id']
if parent_type in all_parents_type:
return r_serv_metadata.hget('map:{}:item_id'.format(parent_type), parent_id)
else:
@ -177,6 +180,8 @@ def add_map_obj_id_item_id(obj_id, item_id, obj_type):
r_serv_metadata.hset('map:twitter_id:item_id', obj_id, item_id)
if obj_type == 'jabber_id':
r_serv_metadata.hset('map:jabber_id:item_id', obj_id, item_id)
if obj_type == 'telegram_id':
r_serv_metadata.hset('map:telegram_id:item_id', obj_id, item_id)
# delete twitter id