How can I `adb pull` a large file when the USB connection breaks continuously?

Question

I want to transfer a large file with adb pull, but my USB connection disconnects all the time, interrupting the transfer. How can I make this work?

Answer 1

Here's a Python 3 script that implements a workaround based on adb and dd. It continuously retries and resumes downloads when disconnection happens.

#!/usr/bin/env python3
adb-repull.py

ADB pull emulation for machines with problematic USB ports/cables.
It continuously retries and resumes download when disconnection happens.

Copyright (c) 2018 Alexander Lopatin

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
import errno
import math
import os
import subprocess
import sys
import time
BUFFER_SIZE = 8192
SLEEP_TIMEOUT = 5
def run_with_retries(function):
    while True:
        try:
            return function()
        except ConnectionError as e:
            print(e)
            print('Retrying after {} seconds'.format(SLEEP_TIMEOUT))
            time.sleep(SLEEP_TIMEOUT)
def size_to_blocks(size):
    return math.ceil(size / BUFFER_SIZE)
def get_size(remote_file):
    print('Getting size of {}'.format(remote_file))
    with subprocess.Popen(['adb', 'shell', 'du', '-b', remote_file], stdout=subprocess.PIPE, stderr=subprocess.PIPE) as process:
        out, err = process.communicate()
    if len(err) > 0:
        raise ConnectionError('Disconnected')

    size_and_remote_file = out.decode('utf-8').split()
    if len(size_and_remote_file) > 0:
        return int(size_and_remote_file[0])
    else:
        raise FileNotFoundError(errno.ENOENT, os.strerror(errno.ENOENT), remote_file)



def is_execout_supported():
    print('Checking if exec-out is supported')
    with subprocess.Popen(['adb', 'exec-out', 'echo'], stdout=subprocess.PIPE, stderr=subprocess.PIPE) as process:
        out, err = process.communicate()
    if len(err) > 0 and err.decode('utf-8').strip() != 'error: closed':
        raise ConnectionError('Disconnected')

    result = process.returncode == 0
    print('Yes' if result else 'No')
    return result



def get_size_with_retries(remote_file):
    return run_with_retries(lambda: get_size(remote_file))
def is_execout_supported_with_retries():
    return run_with_retries(is_execout_supported)
def update_progress(remote_file, current_block, last_block, speed):
    if current_block % 1000 == 0 or current_block == last_block:
        progress = (current_block / last_block) * 100.0
        speed_in_mib = speed / (1024 * 1024)
        print('Downloading {} {:.1f}% ({:.1f} MiB/s)'.format(remote_file, progress, speed_in_mib))
def pull(remote_file, local_file, remote_size, execout, output):
    print('Downloading {}'.format(remote_file))
last_block = size_to_blocks(remote_size)
local_size = os.path.getsize(local_file)
current_block = size_to_blocks(local_size)

time_elapsed = 0
bytes_downloaded = 0

dd_command = "dd if={} bs={} skip={} 2>>/dev/null".format(remote_file, BUFFER_SIZE, current_block)
command = ['adb', 'exec-out', dd_command] if execout else ['adb', 'shell', 'busybox stty raw ; {}'.format(dd_command)]

with subprocess.Popen(command, stdout=subprocess.PIPE) as process:
    while current_block < last_block:
        time_start = time.time()

        current_block += 1
        expected_buffer_size = remote_size - local_size if current_block == last_block else BUFFER_SIZE

        buffer = process.stdout.read(expected_buffer_size)
        buffer_size = len(buffer)

        if buffer_size != expected_buffer_size:
            raise ConnectionError('Wrong buffer size {}. Disconnected'.format(buffer_size))

        output.write(buffer)
        local_size += buffer_size

        time_end = time.time()
        time_elapsed += time_end - time_start

        bytes_downloaded += buffer_size
        speed = bytes_downloaded / time_elapsed
        update_progress(remote_file, current_block, last_block, speed)
    print('Done')



def pull_with_retries(remote_file, local_file):
    remote_size = get_size_with_retries(remote_file)
    execout = is_execout_supported_with_retries()
with open(local_file, 'a+b') as output:
    output.seek(os.SEEK_END)
    run_with_retries(lambda: pull(remote_file, local_file, remote_size, execout, output))



def main(argv):
    if len(argv) < 2:
        print('Usage: %s /mnt/sdcard/remote.bin [local.bin]' % argv[0])
    else:
        try:
            if sys.platform != 'linux':
                raise OSError('Unsupported platform')
            remote_file = argv[1]
            local_file = os.path.basename(remote_file) if len(argv) < 3 else argv[2]
            pull_with_retries(remote_file, local_file)
        except OSError as e:
            print(e)
main(sys.argv)

For me, the performance is the same as adb pull. If you experience problems with performance—try to play with BUFFER_SIZE value.

Answer 2

ADB can not pull files partially or resume. Also modifying the PC side of ADB is AFAIK of no use to you, as the provided functionality is provided by the on-device part (which you can't replace).

From my point of view there are two possibilities left:

1. Install an FTP server (app) that allows to automatically start with Android and that is able resume the download. Let is share a directory, the file you want to get is located in, or use your 20 seconds time window to move the file into the FTP shared directory.

2. Another option would be a cloud synchronization app (or command-line tool you could call via adb) that is able to push files automatically block-wise to the server (like the DropBox desktop client does). Unfortunately AFAIK the official Android DropBox client works differently and is of no use for you in this scenario. But some 3rd party Dropbox app or a different cloud sync app may be usable.

Answer 3

You have a few options here if your connection keeps breaking.

1. Use the script by alopatindev to resume using dd command.

2. Use Wireless ADB over Wi-Fi. The connection may be slightly slower. This is easier, but will also stop when disconnected from WiFi because it is the same as ADB. Go to developer options, enable Wireless Debugging, then tap on the words/arrow "Wireless Debugging". Now you can start ADB wirelessly. Look at your IP and port, and enter them with adb connect IP:port

3. Install RSync in Termux (Fdroid). RSync has a resume option. You will need to be able to start the SSH server on one of the devices (either with systemd on the computer, or running sshd directly on Termux), with RSync on both devices. You can do this on Windows using WSL, but the WSL network is quite slow.