3

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?

8
  • I don't believe there a "resume" or continue option...
    – acejavelin
    Mar 23, 2018 at 18:57
  • Is there a way were I can modify the code myself? Is it open source? Mar 23, 2018 at 19:15
  • I don't know that... and that is outside the scope of this forum (developer specific questions are off-topic here).
    – acejavelin
    Mar 23, 2018 at 19:16
  • Ok, I will just wait to see if someone knows a way or a "wrapper" that enables the resume Mar 23, 2018 at 19:19
  • 1
    My loop lasts around 20 secs. It is a Xiaomi, and I have seen that there are more people affected, with similar loop times. If I could resume pulling, then I could pull files little by little in those 20 secs to get them all. I cannot think of any other approach, to be honest. Any suggestion is appreciated Mar 24, 2018 at 2:37

3 Answers 3

3

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.

0

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.

0

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.

0

This site is temporarily in read-only mode and not accepting new answers.

Not the answer you're looking for? Browse other questions tagged .