Utilizes the Range HTTP header to download a file with multiple connections in parallel, bypassing cheap bandwidth throttling

Torben

Last update: Apr 12, 2022

Related tags

Networking async-range-downloader

Overview

async-range-downloader

Utilizes the Range HTTP header to download a file with multiple connections in parallel, bypassing cheap bandwidth throttling

The in parallel running workers are lightweight coroutines built on ktor's HttpClient and coroutine wrappers around java.nio

Caveats

Server must support the Range header
Server side throttling must be on a per-connection-basis, not IP address

Usage

Usage: ard [options...] url [file]
Arguments:
    url -> URL of file to download { String }
    file -> Path of the output file (optional) { String }
Options:
    --workers, -w [10] -> Number of async workers { Int }
    --delay, -d [1] -> Delay between opening requests in seconds { Int }
    --help, -h -> Usage info

If a filehost for example only allowed free-tier users to download at 1MB/s but your bandwidth allows for 10MB/s, you'd use 10 workers

TODO

No error handling whatsoever at this time, should at least care for response codes
Adding options for custom headers, cookies, or even auth
Proxy support to bypass IP based throttling

Releases(v0.1.0)

v0.1.0(Jul 19, 2021)

Requires Java 11

Usage: ard [options...] url [file]
Arguments:
    url -> URL of file to download { String }
    file -> Path of the output file (optional) { String }
Options:
    --workers, -w [10] -> Number of async workers { Int }
    --delay, -d [1] -> Delay between opening requests in seconds { Int }
    --help, -h -> Usage info

Source code(tar.gz)
Source code(zip)
ard-0.1.0.zip(5.64 MB)