WeChatMsg/app/DataBase/hard_link.py

409 lines
15 KiB
Python
Raw Normal View History

2023-11-20 22:30:31 +08:00
import binascii
import os.path
import sqlite3
import threading
import xml.etree.ElementTree as ET
2023-11-28 21:51:49 +08:00
from app.log import log
2023-12-07 22:20:38 +08:00
image_db_lock = threading.Lock()
video_db_lock = threading.Lock()
image_db_path = "./app/Database/Msg/HardLinkImage.db"
video_db_path = "./app/Database/Msg/HardLinkVideo.db"
2024-01-03 14:47:53 +08:00
root_path = "FileStorage/MsgAttach/"
video_root_path = "FileStorage/Video/"
2023-11-20 22:30:31 +08:00
@log
2024-01-03 14:47:53 +08:00
def get_md5_from_xml(content, type_="img"):
try:
# 解析XML
root = ET.fromstring(content)
2024-01-03 14:47:53 +08:00
if type_ == "img":
2023-12-07 22:20:38 +08:00
# 提取md5的值
md5_value = root.find(".//img").get("md5")
2024-01-03 14:47:53 +08:00
elif type_ == "video":
2023-12-07 22:20:38 +08:00
md5_value = root.find(".//videomsg").get("md5")
# print(md5_value)
return md5_value
except ET.ParseError:
return None
2023-11-20 22:30:31 +08:00
2023-12-07 22:20:38 +08:00
class tencent_struct:
def __setVals__(self, data, off):
if data:
self.__data = data
if self.__data:
self.__size = len(self.__data)
self.__off = off
def __readString(self):
try:
length = self.__readUleb()
2024-01-03 20:40:04 +08:00
res = self.__data[self.__off: self.__off + length]
self.__add(length)
except:
raise
return res.decode("utf-8")
def __readUleb(self):
try:
i = self.__data[self.__off]
self.__add()
if i & 0x80:
j = self.__data[self.__off]
i = i & 0x7F
i = i | (j << 7)
self.__add()
if i & 0x4000:
j = self.__data[self.__off]
i = i & 0x3FFF
i = i | (j << 14)
self.__add()
if i & 0x200000:
j = self.__data[self.__off]
i = i & 0x1FFFFF
i = i | (j << 21)
self.__add()
if i & 0x10000000:
j = self.__data[self.__off]
i = i & 0xFFFFFFF
i = i | (j << 28)
self.__add()
return i
except:
raise
def __readData(self):
try:
length = self.__readUleb()
2024-01-03 20:40:04 +08:00
data = self.__data[self.__off: self.__off + length]
self.__add(length)
return data
except:
raise
def __init__(self, data=None, off=0):
self.__data = data
self.__off = off
if self.__data:
self.__size = len(self.__data)
else:
self.__size = 0
def __add(self, value=1):
self.__off += value
if self.__off > self.__size:
raise "偏移量超出size"
def readStruct(self, struct_type):
current_dict = None
if isinstance(struct_type, str):
current_dict = getattr(self, struct_type)
else:
current_dict = struct_type
res = {}
try:
while self.__off < self.__size:
key = self.__readUleb()
key = key >> 3
if key == 0:
break
op = None
fieldName = ""
if key in current_dict:
op = current_dict[key][1]
fieldName = current_dict[key][0]
else:
break
if isinstance(op, dict):
if not key in res:
res[key] = []
current_struct = self.__readData()
recursion = tencent_struct(current_struct)
res[key].append((fieldName, recursion.readStruct(op)))
elif op != "":
res[key] = (fieldName, self.__contenttype__[op](self))
else:
break
except:
raise
return res
__struct1__ = {1: ("", "I"), 2: ("", "I")}
__msgInfo__ = {1: ("", "I"), 2: ("msg_info", "s")}
__bytesExtra__ = {
1: ("", __struct1__),
2024-01-03 14:50:02 +08:00
3: ("msg_info_struct", __msgInfo__),
2024-01-03 14:47:53 +08:00
}
__struct2__ = {1: ("", "s"), 2: ("", "s")}
__extraBuf__ = {
1: ("", __struct2__),
}
def get_bytesExta_Content(self, data=None, off=0):
self.__setVals__(data, off)
try:
return self.readStruct("__bytesExtra__")
except:
raise
2024-01-03 14:47:53 +08:00
def get_extraBuf_Content(self, data=None, off=0):
self.__setVals__(data, off)
try:
return self.readStruct("__extraBuf__")
except:
raise
__contenttype__ = {
"s": __readString,
"I": __readUleb,
"P": __readData,
}
def parseBytes(content: bytes):
try:
bytesExtra = tencent_struct().get_bytesExta_Content(content)
return bytesExtra
except:
pass
2024-01-03 14:47:53 +08:00
def parseExtraBuf(content: bytes):
try:
extraBuf = tencent_struct().get_extraBuf_Content(content)
return extraBuf
except:
pass
def decodeExtraBuf(extra_buf_content: bytes):
off = 0
types = [b"\x04", b"\x18", b"\x17", b"\x02", b"\x05"]
trunkName = {
"46CF10C4": "个性签名",
"A4D9024A": "国家",
"E2EAA8D1": "省份",
"1D025BBF": "",
"81AE19B4": "朋友圈背景url",
"F917BCC0": "公司名称",
"4EB96D85": "企业微信属性",
"0E719F13": "备注图片",
"759378AD": "手机号",
}
2024-01-03 20:40:04 +08:00
res = {'手机号': {'18':''}}
2024-01-03 14:47:53 +08:00
while off < len(extra_buf_content):
length = 4 # 块头
2024-01-03 20:40:04 +08:00
trunk_head = extra_buf_content[off: off + length]
2024-01-03 14:47:53 +08:00
off += length
trunk_head = binascii.hexlify(trunk_head).decode().upper()
if trunk_head in trunkName:
trunk_head = trunkName[trunk_head]
res[trunk_head] = {}
2024-01-03 20:40:04 +08:00
char = extra_buf_content[off: off + 1]
2024-01-03 14:47:53 +08:00
off += 1
field = binascii.hexlify(char).decode()
if char == b"\x04": # 四个字节的int小端序
length = 4
2024-01-03 20:40:04 +08:00
intContent = extra_buf_content[off: off + length]
2024-01-03 14:47:53 +08:00
off += 4
intContent = int.from_bytes(intContent, "little")
res[trunk_head][field] = intContent
elif char == b"\x18": # utf-16字符串
length = 4
2024-01-03 20:40:04 +08:00
lengthContent = extra_buf_content[off: off + length]
2024-01-03 14:47:53 +08:00
off += 4
lengthContent = int.from_bytes(lengthContent, "little")
2024-01-03 20:40:04 +08:00
strContent = extra_buf_content[off: off + lengthContent]
2024-01-03 14:47:53 +08:00
off += lengthContent
res[trunk_head][field] = strContent.decode("utf-16").rstrip("\x00")
elif char == b"\x17": # utf-8 protobuf
length = 4
2024-01-03 20:40:04 +08:00
lengthContent = extra_buf_content[off: off + length]
2024-01-03 14:47:53 +08:00
off += 4
lengthContent = int.from_bytes(lengthContent, "little")
2024-01-03 20:40:04 +08:00
strContent = extra_buf_content[off: off + lengthContent]
2024-01-03 14:47:53 +08:00
off += lengthContent
res[trunk_head][field] = parseExtraBuf(strContent)
elif char == b"\x02": # 一个字节的int
2024-01-03 20:40:04 +08:00
content = extra_buf_content[off: off + 1]
2024-01-03 14:47:53 +08:00
off += 1
res[trunk_head][field] = int.from_bytes(content, "little")
2024-01-03 14:52:17 +08:00
elif char == b"\x05": # 暂时不知道有啥用固定8个字节先当int处理
2024-01-03 14:47:53 +08:00
length = 8
2024-01-03 20:40:04 +08:00
content = extra_buf_content[off: off + length]
2024-01-03 14:47:53 +08:00
off += length
res[trunk_head][field] = int.from_bytes(content, "little")
2024-01-03 20:40:04 +08:00
return {
'region': (res['国家']['18'], res['省份']['18'], res['']['18']),
'signature': res['个性签名']['18'],
'telephone': res['手机号']['18'],
}
2024-01-03 14:47:53 +08:00
def singleton(cls):
_instance = {}
def inner():
if cls not in _instance:
_instance[cls] = cls()
return _instance[cls]
return inner
@singleton
class HardLink:
def __init__(self):
2023-12-07 22:20:38 +08:00
self.imageDB = None
self.videoDB = None
self.image_cursor = None
self.video_cursor = None
self.open_flag = False
self.init_database()
def init_database(self):
if not self.open_flag:
2023-12-07 22:20:38 +08:00
if os.path.exists(image_db_path):
self.imageDB = sqlite3.connect(image_db_path, check_same_thread=False)
# '''创建游标'''
self.image_cursor = self.imageDB.cursor()
self.open_flag = True
if image_db_lock.locked():
image_db_lock.release()
if os.path.exists(video_db_path):
self.videoDB = sqlite3.connect(video_db_path, check_same_thread=False)
# '''创建游标'''
2023-12-07 22:20:38 +08:00
self.video_cursor = self.videoDB.cursor()
self.open_flag = True
2023-12-07 22:20:38 +08:00
if video_db_lock.locked():
video_db_lock.release()
def get_image_by_md5(self, md5: bytes):
if not md5:
return None
if not self.open_flag:
return None
2024-01-03 14:47:53 +08:00
sql = """
select Md5Hash,MD5,FileName,HardLinkImageID.Dir as DirName1,HardLinkImageID2.Dir as DirName2
from HardLinkImageAttribute
join HardLinkImageID on HardLinkImageAttribute.DirID1 = HardLinkImageID.DirID
join HardLinkImageID as HardLinkImageID2 on HardLinkImageAttribute.DirID2 = HardLinkImageID2.DirID
where MD5 = ?;
2024-01-03 14:47:53 +08:00
"""
try:
2023-12-07 22:20:38 +08:00
image_db_lock.acquire(True)
try:
2023-12-07 22:20:38 +08:00
self.image_cursor.execute(sql, [md5])
except AttributeError:
self.init_database()
2023-12-07 22:20:38 +08:00
self.image_cursor.execute(sql, [md5])
result = self.image_cursor.fetchone()
return result
finally:
2023-12-07 22:20:38 +08:00
image_db_lock.release()
def get_video_by_md5(self, md5: bytes):
if not md5:
return None
if not self.open_flag:
return None
2024-01-03 14:47:53 +08:00
sql = """
2023-12-07 22:20:38 +08:00
select Md5Hash,MD5,FileName,HardLinkVideoID2.Dir as DirName2
from HardLinkVideoAttribute
join HardLinkVideoID as HardLinkVideoID2 on HardLinkVideoAttribute.DirID2 = HardLinkVideoID2.DirID
where MD5 = ?;
2024-01-03 14:47:53 +08:00
"""
2023-12-07 22:20:38 +08:00
try:
video_db_lock.acquire(True)
try:
self.video_cursor.execute(sql, [md5])
except AttributeError:
self.init_database()
self.video_cursor.execute(sql, [md5])
result = self.video_cursor.fetchone()
return result
finally:
video_db_lock.release()
def get_image(self, content, bytesExtra, thumb=False):
bytesDict = parseBytes(bytesExtra)
for msginfo in bytesDict[3]:
if msginfo[1][1][1] == (3 if thumb else 4):
2024-01-03 14:47:53 +08:00
pathh = msginfo[1][2][1] # wxid\FileStorage\...
pathh = "\\".join(pathh.split("\\")[1:])
return pathh
md5 = get_md5_from_xml(content)
if not md5:
return None
result = self.get_image_by_md5(binascii.unhexlify(md5))
2023-12-07 22:20:38 +08:00
if result:
dir1 = result[3]
dir2 = result[4]
data_image = result[2]
2024-01-03 14:47:53 +08:00
dir0 = "Thumb" if thumb else "Image"
dat_image = os.path.join(root_path, dir1, dir0, dir2, data_image)
return dat_image
2023-11-20 22:30:31 +08:00
def get_video(self, content, bytesExtra, thumb=False):
bytesDict = parseBytes(bytesExtra)
for msginfo in bytesDict[3]:
if msginfo[1][1][1] == (3 if thumb else 4):
2024-01-03 14:47:53 +08:00
pathh = msginfo[1][2][1] # wxid\FileStorage\...
pathh = "\\".join(pathh.split("\\")[1:])
return pathh
2024-01-03 14:47:53 +08:00
md5 = get_md5_from_xml(content, type_="video")
2023-12-07 22:20:38 +08:00
if not md5:
return None
result = self.get_video_by_md5(binascii.unhexlify(md5))
if result:
dir2 = result[3]
2024-01-03 14:47:53 +08:00
data_image = result[2].split(".")[0] + ".jpg" if thumb else result[2]
2023-12-07 22:20:38 +08:00
# dir0 = 'Thumb' if thumb else 'Image'
dat_image = os.path.join(video_root_path, dir2, data_image)
return dat_image
def close(self):
if self.open_flag:
try:
2023-12-07 22:20:38 +08:00
image_db_lock.acquire(True)
video_db_lock.acquire(True)
self.open_flag = False
2023-12-07 22:20:38 +08:00
self.imageDB.close()
self.videoDB.close()
finally:
2023-12-07 22:20:38 +08:00
image_db_lock.release()
video_db_lock.release()
2023-11-20 22:30:31 +08:00
def __del__(self):
self.close()
2023-11-20 22:30:31 +08:00
# 6b02292eecea118f06be3a5b20075afc_t
2024-01-03 14:47:53 +08:00
if __name__ == "__main__":
msg_root_path = "./Msg/"
2023-12-07 22:20:38 +08:00
image_db_path = "./Msg/HardLinkImage.db"
video_db_path = "./Msg/HardLinkVideo.db"
hard_link_db = HardLink()
hard_link_db.init_database()
2023-12-07 22:20:38 +08:00
# content = '''<?xml version="1.0"?><msg>\n\t<img aeskey="bc37a58c32cb203ee9ac587b068e5853" encryver="1" cdnthumbaeskey="bc37a58c32cb203ee9ac587b068e5853" cdnthumburl="3057020100044b30490201000204d181705002032f5405020428a7b4de02046537869d042462313532363539632d663930622d343463302d616636662d333837646434633061626534020401150a020201000405004c4c6d00" cdnthumblength="3097" cdnthumbheight="120" cdnthumbwidth="68" cdnmidheight="0" cdnmidwidth="0" cdnhdheight="0" cdnhdwidth="0" cdnmidimgurl="3057020100044b30490201000204d181705002032f5405020428a7b4de02046537869d042462313532363539632d663930622d343463302d616636662d333837646434633061626534020401150a020201000405004c4c6d00" length="57667" md5="6844b812d5d514eb6878657e0bf4cdbb" originsourcemd5="1dfdfa24922270ea1cb5daba103f45ca" />\n\t<platform_signature></platform_signature>\n\t<imgdatahash></imgdatahash>\n</msg>\n'''
# print(hard_link_db.get_image(content))
# print(hard_link_db.get_image(content, thumb=False))
# result = get_md5_from_xml(content)
# print(result)
2024-01-03 14:47:53 +08:00
content = """<?xml version="1.0"?>
2023-12-07 22:20:38 +08:00
<msg>
<videomsg aeskey="d635d2013d221dbd05a4eab3a8185f5a" cdnvideourl="3057020100044b304902010002040297cead02032f540502042ba7b4de020465673b74042438316562356530652d653764352d343263632d613531642d6464383661313330623965330204052400040201000405004c537500" cdnthumbaeskey="d635d2013d221dbd05a4eab3a8185f5a" cdnthumburl="3057020100044b304902010002040297cead02032f540502042ba7b4de020465673b74042438316562356530652d653764352d343263632d613531642d6464383661313330623965330204052400040201000405004c537500" length="25164270" playlength="60" cdnthumblength="7419" cdnthumbwidth="1920" cdnthumbheight="1080" fromusername="wxid_yt67eeoo4blm22" md5="95558f0e503651375b475636519d2285" newmd5="4ece19bcd92dc5b93b83f397461a1310" isplaceholder="0" rawmd5="d660ba186bb31126d94fa568144face8" rawlength="143850007" cdnrawvideourl="3052020100044b30490201000204d8cd585302032f540502040f6a42b7020465673b85042464666462306634342d653339342d343232302d613534392d3930633030646236306266610204059400040201000400" cdnrawvideoaeskey="5915b14ac8d121e0944d9e444aebb7ed" overwritenewmsgid="0" originsourcemd5="a1a567d8c170bca33d075b787a60dd3f" isad="0" />
</msg>
2024-01-03 14:47:53 +08:00
"""
2023-12-07 22:20:38 +08:00
print(hard_link_db.get_video(content))
print(hard_link_db.get_video(content, thumb=True))