Files
WhatsApp-Chat-Exporter/Whatsapp_Chat_Exporter/extract.py
2022-05-23 20:12:50 +08:00

537 lines
20 KiB
Python

#!/usr/bin/python3
import sqlite3
import json
import jinja2
import os
import shutil
import re
import io
import hmac
from pathlib import Path
from bleach import clean as sanitize
from markupsafe import Markup
from datetime import datetime
from enum import Enum
from mimetypes import MimeTypes
from hashlib import sha256
try:
import zlib
from Crypto.Cipher import AES
except ModuleNotFoundError:
support_backup = False
else:
support_backup = True
try:
import javaobj
except ModuleNotFoundError:
support_crypt15 = False
else:
support_crypt15 = True
def sanitize_except(html):
return Markup(sanitize(html, tags=["br"]))
def determine_day(last, current):
last = datetime.fromtimestamp(last).date()
current = datetime.fromtimestamp(current).date()
if last == current:
return None
else:
return current
CRYPT14_OFFSETS = [
{"iv": 67, "db": 191},
{"iv": 67, "db": 190},
{"iv": 66, "db": 99}
]
class Crypt(Enum):
CRYPT15 = 15
CRYPT14 = 14
CRYPT12 = 12
def brute_force_offset():
for iv in range(0, 200):
for db in range(0, 200):
yield iv, iv + 16, db
def _generate_hmac_of_hmac(key_stream):
key = hmac.new(
hmac.new(
b'\x00' * 32,
key_stream,
sha256
).digest(),
b"backup encryption\x01",
sha256
)
return key.digest()
def _extract_encrypted_key(keyfile):
key_stream = b""
for byte in javaobj.loads(keyfile):
key_stream += byte.to_bytes(1, "big", signed=True)
return _generate_hmac_of_hmac(key_stream)
def decrypt_backup(database, key, output, crypt=Crypt.CRYPT14):
if not support_backup:
return 1
if isinstance(key, io.IOBase):
key = key.read()
if crypt is not Crypt.CRYPT15:
t1 = key[30:62]
if crypt is not Crypt.CRYPT15 and len(key) != 158:
raise ValueError("The key file must be 158 bytes")
if crypt == Crypt.CRYPT14:
if len(database) < 191:
raise ValueError("The crypt14 file must be at least 191 bytes")
current_try = 0
offsets = CRYPT14_OFFSETS[current_try]
t2 = database[15:47]
iv = database[offsets["iv"]:offsets["iv"] + 16]
db_ciphertext = database[offsets["db"]:]
elif crypt == Crypt.CRYPT12:
if len(database) < 67:
raise ValueError("The crypt12 file must be at least 67 bytes")
t2 = database[3:35]
iv = database[51:67]
db_ciphertext = database[67:-20]
elif crypt == Crypt.CRYPT15:
if not support_crypt15:
return 1
if len(database) < 131:
raise ValueError("The crypt15 file must be at least 131 bytes")
t1 = t2 = None
iv = database[8:24]
db_ciphertext = database[131:]
if t1 != t2:
raise ValueError("The signature of key file and backup file mismatch")
if crypt == Crypt.CRYPT15:
if len(key) == 32:
main_key = _generate_hmac_of_hmac(key)
else:
main_key = _extract_encrypted_key(key)
else:
main_key = key[126:]
decompressed = False
while not decompressed:
cipher = AES.new(main_key, AES.MODE_GCM, iv)
db_compressed = cipher.decrypt(db_ciphertext)
try:
db = zlib.decompress(db_compressed)
except zlib.error:
if crypt == Crypt.CRYPT14:
current_try += 1
if current_try < len(CRYPT14_OFFSETS):
offsets = CRYPT14_OFFSETS[current_try]
iv = database[offsets["iv"]:offsets["iv"] + 16]
db_ciphertext = database[offsets["db"]:]
continue
else:
print("Common offsets are not applicable to "
"your backup. Trying to brute force it...")
for start_iv, end_iv, start_db in brute_force_offset():
iv = database[start_iv:end_iv]
db_ciphertext = database[start_db:]
cipher = AES.new(main_key, AES.MODE_GCM, iv)
db_compressed = cipher.decrypt(db_ciphertext)
try:
db = zlib.decompress(db_compressed)
except zlib.error:
continue
else:
decompressed = True
print(
f"The offsets of your IV and database are {start_iv} and "
f"{start_db}, respectively. To include your offsets in the "
"program, please report it by creating an issue on GitHub: "
"https://github.com/KnugiHK/Whatsapp-Chat-Exporter/issues/new"
)
break
if not decompressed:
return 2
else:
return 3
else:
decompressed = True
if db[0:6].upper() == b"SQLITE":
with open(output, "wb") as f:
f.write(db)
return 0
else:
raise ValueError("The plaintext is not a SQLite database. Did you use the key to encrypt something...")
def contacts(db, data):
# Get contacts
c = db.cursor()
c.execute("""SELECT count() FROM wa_contacts""")
total_row_number = c.fetchone()[0]
print(f"Gathering contacts...({total_row_number})")
c.execute("""SELECT jid, display_name FROM wa_contacts; """)
row = c.fetchone()
while row is not None:
data[row[0]] = {"name": row[1], "messages": {}}
row = c.fetchone()
def messages(db, data):
# Get message history
c = db.cursor()
c.execute("""SELECT count() FROM messages""")
total_row_number = c.fetchone()[0]
print(f"Gathering messages...(0/{total_row_number})", end="\r")
phone_number_re = re.compile(r"[0-9]+@s.whatsapp.net")
c.execute("""SELECT messages.key_remote_jid,
messages._id,
messages.key_from_me,
messages.timestamp,
messages.data,
messages.status,
messages.edit_version,
messages.thumb_image,
messages.remote_resource,
messages.media_wa_type,
messages.latitude,
messages.longitude,
messages_quotes.key_id as quoted,
messages.key_id,
messages_quotes.data,
messages.media_caption
FROM messages
LEFT JOIN messages_quotes
ON messages.quoted_row_id = messages_quotes._id
WHERE messages.key_remote_jid <> '-1';""")
i = 0
content = c.fetchone()
while content is not None:
if content[0] not in data:
data[content[0]] = {"name": None, "messages": {}}
data[content[0]]["messages"][content[1]] = {
"from_me": bool(content[2]),
"timestamp": content[3]/1000,
"time": datetime.fromtimestamp(content[3]/1000).strftime("%H:%M"),
"media": False,
"key_id": content[13],
"meta": False,
"data": None
}
if "-" in content[0] and content[2] == 0:
name = None
if content[8] in data:
name = data[content[8]]["name"]
if "@" in content[8]:
fallback = content[8].split('@')[0]
else:
fallback = None
else:
fallback = None
data[content[0]]["messages"][content[1]]["sender"] = name or fallback
else:
data[content[0]]["messages"][content[1]]["sender"] = None
if content[12] is not None:
data[content[0]]["messages"][content[1]]["reply"] = content[12]
data[content[0]]["messages"][content[1]]["quoted_data"] = content[14]
else:
data[content[0]]["messages"][content[1]]["reply"] = None
if content[15] is not None:
data[content[0]]["messages"][content[1]]["caption"] = content[15]
else:
data[content[0]]["messages"][content[1]]["caption"] = None
if content[5] == 6:
if "-" in content[0]:
# Is Group
if content[4] is not None:
try:
int(content[4])
except ValueError:
msg = f"The group name changed to {content[4]}"
data[content[0]]["messages"][content[1]]["data"] = msg
data[content[0]]["messages"][content[1]]["meta"] = True
else:
del data[content[0]]["messages"][content[1]]
else:
thumb_image = content[7]
if thumb_image is not None:
if b"\x00\x00\x01\x74\x00\x1A" in thumb_image:
# Add user
added = phone_number_re.search(
thumb_image.decode("unicode_escape"))[0]
if added in data:
name_right = data[added]["name"]
else:
name_right = added.split('@')[0]
if content[8] is not None:
if content[8] in data:
name_left = data[content[8]]["name"]
else:
name_left = content[8].split('@')[0]
msg = f"{name_left} added {name_right or 'You'}"
else:
msg = f"Added {name_right or 'You'}"
elif b"\xac\xed\x00\x05\x74\x00" in thumb_image:
# Changed number
original = content[8].split('@')[0]
changed = thumb_image[7:].decode().split('@')[0]
msg = f"{original} changed to {changed}"
data[content[0]]["messages"][content[1]]["data"] = msg
data[content[0]]["messages"][content[1]]["meta"] = True
else:
if content[4] is None:
del data[content[0]]["messages"][content[1]]
else:
# Private chat
if content[4] is None and content[7] is None:
del data[content[0]]["messages"][content[1]]
else:
if content[2] == 1:
if content[5] == 5 and content[6] == 7:
msg = "Message deleted"
data[content[0]]["messages"][content[1]]["meta"] = True
else:
if content[9] == "5":
msg = f"Location shared: {content[10], content[11]}"
data[content[0]]["messages"][content[1]]["meta"] = True
else:
msg = content[4]
if msg is not None:
if "\r\n" in msg:
msg = msg.replace("\r\n", "<br>")
if "\n" in msg:
msg = msg.replace("\n", "<br>")
else:
if content[5] == 0 and content[6] == 7:
msg = "Message deleted"
data[content[0]]["messages"][content[1]]["meta"] = True
else:
if content[9] == "5":
msg = f"Location shared: {content[10], content[11]}"
data[content[0]]["messages"][content[1]]["meta"] = True
else:
msg = content[4]
if msg is not None:
if "\r\n" in msg:
msg = msg.replace("\r\n", "<br>")
if "\n" in msg:
msg = msg.replace("\n", "<br>")
data[content[0]]["messages"][content[1]]["data"] = msg
i += 1
if i % 1000 == 0:
print(f"Gathering messages...({i}/{total_row_number})", end="\r")
content = c.fetchone()
print(f"Gathering messages...({total_row_number}/{total_row_number})", end="\r")
def media(db, data, media_folder):
# Get media
c = db.cursor()
c.execute("""SELECT count() FROM message_media""")
total_row_number = c.fetchone()[0]
print(f"\nGathering media...(0/{total_row_number})", end="\r")
i = 0
c.execute("""SELECT messages.key_remote_jid,
message_row_id,
file_path,
message_url,
mime_type,
media_key
FROM message_media
INNER JOIN messages
ON message_media.message_row_id = messages._id
ORDER BY messages.key_remote_jid ASC""")
content = c.fetchone()
mime = MimeTypes()
while content is not None:
file_path = f"{media_folder}/{content[2]}"
data[content[0]]["messages"][content[1]]["media"] = True
if os.path.isfile(file_path):
data[content[0]]["messages"][content[1]]["data"] = file_path
if content[4] is None:
guess = mime.guess_type(file_path)[0]
if guess is not None:
data[content[0]]["messages"][content[1]]["mime"] = guess
else:
data[content[0]]["messages"][content[1]]["mime"] = "data/data"
else:
data[content[0]]["messages"][content[1]]["mime"] = content[4]
else:
# if "https://mmg" in content[4]:
# try:
# r = requests.get(content[3])
# if r.status_code != 200:
# raise RuntimeError()
# except:
# data[content[0]]["messages"][content[1]]["data"] = "{The media is missing}"
# data[content[0]]["messages"][content[1]]["media"] = True
# data[content[0]]["messages"][content[1]]["mime"] = "media"
# else:
data[content[0]]["messages"][content[1]]["data"] = "The media is missing"
data[content[0]]["messages"][content[1]]["mime"] = "media"
data[content[0]]["messages"][content[1]]["meta"] = True
i += 1
if i % 100 == 0:
print(f"Gathering media...({i}/{total_row_number})", end="\r")
content = c.fetchone()
print(
f"Gathering media...({total_row_number}/{total_row_number})", end="\r")
def vcard(db, data):
c = db.cursor()
c.execute("""SELECT message_row_id,
messages.key_remote_jid,
vcard,
messages.media_name
FROM messages_vcards
INNER JOIN messages
ON messages_vcards.message_row_id = messages._id
ORDER BY messages.key_remote_jid ASC;""")
rows = c.fetchall()
total_row_number = len(rows)
print(f"\nGathering vCards...(0/{total_row_number})", end="\r")
base = "WhatsApp/vCards"
if not os.path.isdir(base):
Path(base).mkdir(parents=True, exist_ok=True)
for index, row in enumerate(rows):
media_name = row[3] if row[3] else ""
file_name = "".join(x for x in media_name if x.isalnum())
file_path = f"{base}/{file_name}.vcf"
if not os.path.isfile(file_path):
with open(file_path, "w", encoding="utf-8") as f:
f.write(row[2])
data[row[1]]["messages"][row[0]]["data"] = media_name + \
"The vCard file cannot be displayed here, " \
f"however it should be located at {file_path}"
data[row[1]]["messages"][row[0]]["mime"] = "text/x-vcard"
data[row[1]]["messages"][row[0]]["meta"] = True
print(f"Gathering vCards...({index + 1}/{total_row_number})", end="\r")
def create_html(data, output_folder, template=None, embedded=False):
if template is None:
template_dir = os.path.dirname(__file__)
template_file = "whatsapp.html"
else:
template_dir = os.path.dirname(template)
template_file = os.path.basename(template)
templateLoader = jinja2.FileSystemLoader(searchpath=template_dir)
templateEnv = jinja2.Environment(loader=templateLoader)
templateEnv.globals.update(determine_day=determine_day)
templateEnv.filters['sanitize_except'] = sanitize_except
template = templateEnv.get_template(template_file)
total_row_number = len(data)
print(f"\nCreating HTML...(0/{total_row_number})", end="\r")
if not os.path.isdir(output_folder):
os.mkdir(output_folder)
for current, contact in enumerate(data):
if len(data[contact]["messages"]) == 0:
continue
phone_number = contact.split('@')[0]
if "-" in contact:
file_name = ""
else:
file_name = phone_number
if data[contact]["name"] is not None:
if file_name != "":
file_name += "-"
file_name += data[contact]["name"].replace("/", "-")
name = data[contact]["name"]
else:
name = phone_number
safe_file_name = ''
safe_file_name = "".join(x for x in file_name if x.isalnum() or x in "- ")
with open(f"{output_folder}/{safe_file_name}.html", "w", encoding="utf-8") as f:
f.write(
template.render(
name=name,
msgs=data[contact]["messages"].values(),
my_avatar=None,
their_avatar=f"WhatsApp/Avatars/{contact}.j"
)
)
if current % 10 == 0:
print(f"Creating HTML...({current}/{total_row_number})", end="\r")
print(f"Creating HTML...({total_row_number}/{total_row_number})", end="\r")
if __name__ == "__main__":
from optparse import OptionParser
parser = OptionParser()
parser.add_option(
"-w",
"--wa",
dest="wa",
default="wa.db",
help="Path to contact database")
parser.add_option(
"-m",
"--media",
dest="media",
default="WhatsApp",
help="Path to WhatsApp media folder"
)
# parser.add_option(
# "-t",
# "--template",
# dest="html",
# default="wa.db",
# help="Path to HTML template")
(options, args) = parser.parse_args()
msg_db = "msgstore.db"
output_folder = "temp"
contact_db = options.wa
media_folder = options.media
if len(args) == 1:
msg_db = args[0]
elif len(args) == 2:
msg_db = args[0]
output_folder = args[1]
data = {}
if os.path.isfile(contact_db):
with sqlite3.connect(contact_db) as db:
contacts(db, data)
if os.path.isfile(msg_db):
with sqlite3.connect(msg_db) as db:
messages(db, data)
media(db, data, media_folder)
vcard(db, data)
create_html(data, output_folder)
if not os.path.isdir(f"{output_folder}/WhatsApp"):
shutil.move(media_folder, f"{output_folder}/")
with open("result.json", "w") as f:
data = json.dumps(data)
print(f"\nWriting JSON file...({int(len(data)/1024/1024)}MB)")
f.write(data)
print("Everything is done!")