Digital Forensics

вівторок, 19 листопада 2024 р.

1000

1000 days of War

0,0001 % of сrimes against Ukraine:

17.11.2024, Sumy

12 people were killed, including two children.

28.09.2024, Sumy

russians attacked one of the largest local hospitals twice, killing 6 people.

08.07.2024, Kyiv

OHMATDYT. The largest children’s hospital in Ukraine. Up to 20,000 children are treated annually. About 10,000 surgeries are performed annually.

13.03.2024, Sumy

russians UAVs attacked residential sector of the city center.

24.02.2022-24.03.2022, Trostianets

30 days of occupation.

07.03.2022, Sumy

Explosions in the area of Romenska Street destroyed dozens of houses, killing 24 people. Four of them are children.

понеділок, 28 лютого 2022 р.

ПРАВДА О ВОЙНЕ В УКРАИНЕ - NO WAR IN UKRAINE

На вашем языке, чтоб вы поняли.

https://t.me/UkrainianWitness (прим. - погляди та думки, висловлені на цих каналах, не обов'язково відображають мої погляди чи позицію; контект сильно змінився станом на 2024 рік)

https://t.me/rf200_now (прим. - канал, схоже, давно видалений, бо відверто демонстрував що і як відбувалося на фронті)

Правда о том, что происходит с вашими детьми в Украине (в Киеве, Житомире, Буче, Гостомеле, Харькове, Чернигове, Сумах, Бердянске и т.д.)

Смотрите на то, куда вы посылаете своих детей.

1000 смертей каждый день!

Т-Ы-С-Я-Ч-И!!

У нас люди выходят с голыми руками против танков, а вы БОИТЕСЬ дубинок ваших ментов!

Как вы нас собрались завоёвывать, нация трусов?! Какая нация такой и руководитель. Лезьте тоже все в бункер, а то мобилизовывать скоро вас будут после окончания первых 150000.

Кстати, матери, ищите своих детей мёртвых и не очень здесь:

https://t.me/rf200_now (The views and opinions expressed in these channels do not necessarily reflect my views or positions).

The truth about what is happening to your children in Ukraine (in Kyiv, Zhytomyr, Bucha, Gostomel, Kharkiv, Chernihiv, Sumy, Berdyansk, etc.)

See where you send your children.

1000 deaths every day!

T-H-O-U-S-A-N-D-S!!!

We have people coming out with their bare hands against tanks, and you are AFRAID of your cops' clubs!

Nation of cowards, how are you going to conquer us?! The people is like as your leader. Climb into the bunker too, otherwise you will be mobilized soon after the end of the first 150,000.

By the way, mothers, look for your children dead and not so here:

https://t.me/rf200_now (note - the channel was probably deleted a long time ago, because it seemed to openly demonstrate what was happening at the front.)

неділя, 1 листопада 2020 р.

VERIFYING OF UFED EXTRACTION IN CASE OF MULTI-VOLUME ZIP-ARCHIVES

Maksym Boiko, mboiko25@gmail.com, Kyiv, 2020

ufed-sha256.py

Some extractions of UFED 4PC are multi-volume ZIP-archives. In order to verify this process it needs to calculate checksum of this whole ZIP-file obtained during export by UFED 4PC software.

The following script calculates the sha256-checksum of each file separately and the entire block of files with an extension starting with the character "z". The hash value of the entire block of files is displayed in the last line.

This script allows to calculate sha256-checksums of files and provides the result in a format compatible with HashCheck, Terracopy, etc.

You should specify two parameters in the command line: 1) the full path to the data export folder, 2) the relative or full path to the file with computed sha256-values:

python.exe ufed-sha256.py <path to data folder> <path to file>

Example:

python.exe ufed-sha256.py "D:\UFED 2020_07_07 (001)\FileSystem Android Backup 01" D:\checksums.sha256

import os, sys
from datetime import datetime
import hashlib

#sha256
def file_as_bytes(f_name):
    with f_name:
        return f_name.read()

dir_ufed=sys.argv[1]
ff=sys.argv[2]
f1=open(ff,'w')

t1=datetime.now()
print ''
print(t1),'\n'
i=0
ln=''
f_sha256={}
sha256=hashlib.sha256()
sha256_all=hashlib.sha256()
for top, dirs, files in os.walk(dir_ufed):
    for nm in files:
        i=i+1
        f=os.path.join(top, nm) # top - filepath, nm - filename
        if f[-3:-2]=='z':
            f_sha256[i]=hashlib.sha256(file_as_bytes(open(f,'rb'))).hexdigest()
	    print f_sha256[i],'*'+f[3:]
            f1.writelines(f_sha256[i])
            f1.writelines(' *')
            f1.writelines(f[3:])
            f1.write('\n')
	    sha256_all.update(file_as_bytes(open(f,'rb')))

t2=datetime.now()
print ''
print(t2)
print(t2-t1),'\n'
ln='[SHA256]=',sha256_all.hexdigest()
print '[SHA256]=',sha256_all.hexdigest()
f1.writelines(ln)
f1.close()

import os, sys
from datetime import datetime
import hashlib

#sha256
def file_as_bytes(f_name):
    with f_name:
        return f_name.read()

dir_ufed=sys.argv[1]
ff=sys.argv[2]
f1=open(ff,'w')

t1=datetime.now()
print ''
print(t1),'\n'
i=0
ln=''
f_sha256={}
sha256=hashlib.sha256()
sha256_all=hashlib.sha256()
for top, dirs, files in os.walk(dir_ufed):
    for nm in files:
        i=i+1
        f=os.path.join(top, nm) # top - filepath, nm - filename
        if f[-3:-2]=='z':
            f_sha256[i]=hashlib.sha256(file_as_bytes(open(f,'rb'))).hexdigest()
	    print f_sha256[i],'*'+f[3:]
            f1.writelines(f_sha256[i])
            f1.writelines(' *')
            f1.writelines(f[3:])
            f1.write('\n')
	    sha256_all.update(file_as_bytes(open(f,'rb')))

t2=datetime.now()
print ''
print(t2)
print(t2-t1),'\n'
ln='[SHA256]=',sha256_all.hexdigest()
print '[SHA256]=',sha256_all.hexdigest()
f1.writelines(ln)
f1.close()

неділя, 17 травня 2020 р.

WORDPROCESSINGML' SINGLE SESSION REVISION SAVE IDs (RSIDs)

Maksym Boiko, mboiko25@gmail.com, Kyiv, 2020

Remark.

Revision Save IDs (RSIDs) are an interesting thing within OOXML-files. These tags can help an investigator to compare several Microsoft Office documents more deeply and to define a parent document for some text.

There were not so many works on this theme. Maybe, the best one is “Forensic Analysis of OOXML Documents” (E. Didriksen). But if you want to comprehend all the subtleties of WordProcessingML document’ structure, you need to have read some parts of “Standard ECMA-376 Office Open XML File Formats”. Of course, you need to investigate a lot of OOXML-files yourself as well.

I hope I will have more time to present a more comprehensive work with practical cases in the future.

Особливості документів WordprocessingML.pdf (ukr., укр.)

WordprocessingML-Boiko.pdf (eng.)

середа, 1 травня 2019 р.

SCRIPTS FOR TELEGRAM CHAT EXTRACTION

Maksym Boiko, mboiko25@gmail.com, Kyiv, 2019

telegram-sqlite-secret.py

telegram-sqlite-v7.py

telegram-sqlite-v7-without-chats.py

telegram-parse-enc.py

After some requests I decided to attach some test scripts. I didn't do them for widespread use. All of them were for my particular cases. These codes were written in 2018. But in general, an internal structure of cache4.db remains the same.

As for telegram-parse-enc.py, sometimes it has problems with transliteration and encoding. It is good for analyzing with unknown, poorly studied structure or raw data (for instance, in cases of searching within RAM dumps).

I have to say that in my opinion this way of analyzing intact cache4.db is not optimal. It is better to analyze it as simple SQLite database and its fields.

For these purposes I use telegram-sqlite-secret.py, telegram-sqlite-v7.py or telegram-sqlite-v7-without-chats.py. Briefly, they perform SQLite queries and then analyze data depending on whether it is chat, secret chat, channel or not. If my memory doesn't fail me, they have similar codes. Important moment - I specify database filename in the beginning of these scripts (for instance, ff='cache4.db-wal' or conn=sqlite3.connect('cache4.db')).

I hope scripts will be useful for you.

пʼятниця, 19 квітня 2019 р.

SOME PRACTICAL ASPECTS OF SQLITE RECORD STRUCTURE ON AN EXAMPLE OF TELEGRAM “CACHE4.DB” FILE

Maksym Boiko, mboiko25@gmail.com, Kyiv, 2019

The subject of study is deeper understanding of the SQLite3 database file on examples of the “cache4.db” file of Telegram for Android.

Software. The study used the following software: operating systems Windows 10, Microsoft Office 365 Home 16.0.10730.20102 (© 2012 Microsoft); sqlite 3.26.0 (https:/www.sqlite.org/copyright.html), SQLite Expert Personal 5.3.0.339 (© 2018 Coral Creek Software), AccessData FTK Imager 4.2.0.13 (© 2016 AccessData Group, Inc.), HxD Hex Editor 2.1.0.0 (© 2002-2018 Maël Hörz).

I’d like to present you a simplified algorithm of SQLite records’ structure analysis. I chose a “cache4.db” database file for this purpose. This file contains Telegram message-related data. It should be said some tools that specialized on mobile forensics doesn’t support an analyzing of secret chats of Telegram for Android (as well as in WAL-files).

The “cache4.db” file is a SQLite database. So it has an internal structure that is described and documented in detail (see the following link https://www.sqlite.org/fileformat2.html).

Every record from a SQLite table contains from 2 main parts: a header and a body. Also payload length and RowID are represented before a record header.

Usually header length consists of a single byte. Values for every column follow the header. If serial type is equal to 0, 8, 9, 12, 13, then a field’s length is 0 bytes. If serial type is equal to 1, 2, 3, 4, 5, 6, 7, then a field’s length can be 1, 2, 3, 4, 6 or 8 bytes.

To correctly interpret header serial types, we will use explanations that are represented below and available by link https://www.sqlite.org/fileformat2.html:

Table 1 Serial Type Codes Of The Record Format

Serial Type	Content Size	Meaning
0	0	Value is a NULL.
1	1	Value is an 8-bit twos-complement integer.
2	2	Value is a big-endian 16-bit twos-complement integer.
3	3	Value is a big-endian 24-bit twos-complement integer.
4	4	Value is a big-endian 32-bit twos-complement integer.
5	6	Value is a big-endian 48-bit twos-complement integer.
6	8	Value is a big-endian 64-bit twos-complement integer.
7	8	Value is a big-endian IEEE 754-2008 64-bit floating point number.
8	0	Value is the integer 0. (Only available for schema format 4 and higher.)
9	0	Value is the integer 1. (Only available for schema format 4 and higher.)
10,11	variable	Reserved for internal use. These serial type codes will never appear in a well-formed database file, but they might be used in transient and temporary database files that SQLite sometimes generates for its own use. The meanings of these codes can shift from one release of SQLite to the next.
N≥12 and even	(N-12)/2	Value is a BLOB that is (N-12)/2 bytes in length.
N≥13 and odd	(N-13)/2	Value is a string in the text encoding and (N-13)/2 bytes in length. The null terminator is not stored.

In cases of large BLOBs and strings, we need to convert serial type’s values for determining of fields’ length.

So, if the 1^st more significant byte (N₁) of some field is more than 0x80 (128 in decimal) then we need to get the next byte (N₂) and to do some mathematical operations. To determine a length of particular field that is specified within record’s header, we need to perform following actions:

	(if (N₁-128)128 + N₂* is even)
or
	(if (N₁-128)128 + N₂* is odd)

Note: To calculate a payload length we only need to convert these values to decimal. In case of determining of a large payload’s length, we need to perform only following actions:

(N₁-128)*128 + N₂

Let’s imagine that we have extracted a Telegram “cache4.db” file from Android device. (see the following link for detailedinformation about location, typical file structure, etc.)

First of all to ease our task we need to determine a table with significant information, its schema, and only then we’ll be able to simply analyze records from early mentioned table we’re interested in. For instance, we’d like to search for message-related data in the “Messages” table.

We can find out a schema of this table in different ways – by using sqlite dot-commands, SQLite database viewers, HEX-viewer, or even from another similar database, etc. Knowing table’s schema, we can disassemble every record of certain table. For example, a schema of the “Messages” table has been received by using sqlite application and .schema dot-command as shown on the following figure 1:

CREATE TABLE messages(

mid INTEGER PRIMARY KEY,

uid INTEGER,

read_state INTEGER,

send_state INTEGER,

date INTEGER,

data BLOB,

out INTEGER,

ttl INTEGER,

media INTEGER,

replydata BLOB,

imp INTEGER,

mention INTEGER

);

Figure 1 “Messages” table’s schema

So, we can make a conclusion that our table has 12 columns with values that we can find by analyzing records’ headers and bodies in way mentioned below.

Let’s consider in details one of the records from the “Messages” table that is shown on the following figure 2. This record is presented in raw on the figure 3. We can see an internal structure of this record in the table 2 and on the following figures 4, 5, 6 as well as descriptions and interpretations of received values.

Figure 2 A record from “Messages” table

Figure 3 A record from “Messages” table (in raw)

Table 2 Record format

№	Offset	Length	Value	Interpretation	Description
1	0x7e83	2	0x827a	378	Payload length
2	0x7e85	1	0x26	38	RowID
3	0x7e86	1	0x0e	14	Record header length
4	0x7e87	1	0x00	NULL	mid
5	0x7e88	1	0x03	3 (24-bit integer)	uid length
6	0x7e89	1	0x01	1 (8-bit integer)	read_state length
7	0x7e8a	1	0x08	0	send_state
8	0x7e8b	1	0x04	4 (32-bit integer)	date length
9	0x7e8c	2	0x8554	356 (BLOB length)	data length
10	0x7e8e	1	0x08	0	out
11	0x7e8f	1	0x08	0	ttl
12	0x7e90	1	0x08	0	media
13	0x7e91	1	0x00	NULL	replydata
14	0x7e92	1	0x08	0	imp
15	0x7e93	1	0x08	0	mention
16	0x7e94	3	0x0bdb28	777000	uid value
17	0x7e97	1	0x03	3	read_state value
18	0x7e98	4	0x5c39f0c0	1547301056	date value
19	0x7e9c	356	Data (BLOB)	Message + metadata	Data (BLOB)