problem extracting backup.db

[motionthings]

Describe the bug
Getting a 'utf-8' error message when trying to extract backup.db with dd command

To Reproduce
$ dd if=backup.ab bs=1 skip=24 | python -c "import zlib,sys;sys.stdout.write(zlib.decompress(sys.stdin.read()))" | tar -xvf -
23+0 records in
23+0 records out
23 bytes copied, 5.1124e-05 s, 450 kB/s
Traceback (most recent call last):
File "", line 1, in
import zlib,sys;sys.stdout.write(zlib.decompress(sys.stdin.read()))
File "", line 325, in decode
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xda in position 1: invalid continuation byte
tar: This does not look like a tar archive
tar: Exiting with failure status due to previous errors
[ble: exit 2]

Platform:
EndeavourOS (Arch)

Am I missing a library of some sort?

[RobHofmann]

ChatGPT says the following:

You're not missing a library — this is the classic Python 3 vs Python 2 text vs bytes problem.

What's happening

Your one-liner worked in Python 2 because sys.stdin.read() returned bytes.
In Python 3 sys.stdin.read() returns a str (text), so when you pass that to zlib.decompress() Python attempts to coerce/encode/decode and that triggers the UnicodeDecodeError you saw. zlib.decompress() expects bytes, and tar expects binary output, so you must read/write the binary buffers.

Quick fix (one-liner that works on Python 3)

dd if=backup.ab bs=1 skip=24 \
  | python3 -c 'import sys,zlib; sys.stdout.buffer.write(zlib.decompress(sys.stdin.buffer.read()))' \
  | tar -xvf -

A slightly faster, simpler variant using tail (avoids bs=1 slowness):

tail -c +25 backup.ab \
  | python3 -c 'import sys,zlib; sys.stdout.buffer.write(zlib.decompress(sys.stdin.buffer.read()))' \
  | tar -xvf -

Extra checks & tips

• Confirm which Python you’re running:
  python --version or python3 --version — if python is Python 2 this explains why the original one-liner worked elsewhere.
• If you want to inspect the header length first:
  head -c 48 backup.ab | xxd (or hexdump -C backup.ab | head) — Android .ab headers can vary; make sure skip=24 is the correct offset for your file.
• dd bs=1 is very slow — tail -c +25 or dd if=backup.ab skip=24 bs=1M (with a calculated offset) are faster options.
• If the tar still complains “not a tar archive”, either the skip offset is wrong or the .ab wasn’t compressed with zlib (or is corrupted). Verify the header bytes and try different skip numbers (or view file with xxd).

If you want, paste the first ~48 bytes of backup.ab (head -c 48 backup.ab | xxd -g 1) and I’ll point at the exact offset and whether it's zlib compressed.

Let me know if this helps.

[motionthings]

That worked
Thank you ChatGPT /s

Maybe update the readme.md?

[RobHofmann]

Done! Thanks for reporting back!