get

README

get

A multi-connection file downloader using parallel HTTP range requests.

get -p 4 https://example.com/file.tar.gz
 ████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░  11% 1.79 MiB/s

• Fast: downloads a file using multiple connections
• Resumable: keeps downloaded chunks and continues where it left off
• Mirror-aware: can download from multiple URLs at once
• Cross-platform: builds on Windows, Linux and macOS

Install

Build from source (requires Go 1.27.1+):

go install github.com/setanarut/get/cmd/get@latest

Usage

Download a file with 4 connections:

get -p 4 https://example.com/file.tar.gz

Save to a specific path:

get -o ./downloads/file.tar.gz https://example.com/file.tar.gz

Download from multiple mirrors at the same time:

get -p 2 https://mirror-a.com/file.tar.gz https://mirror-b.com/file.tar.gz

URLs can also be passed through stdin, one per line.

Resume

If a download is interrupted (ctrl+c), run the same command again. The downloaded chunks are stored in a _<filename>.partial directory next to the output file and reused on the next run.

./get https://example.com/file.tar.gz ████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 17% 1.77 MiB/s ^C ./get https://example.com/file.tar.gz ███████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░ 21% 1.74 MiB/s ^C

Because chunks are stored by their byte offset, you can resume with a different -p value and the already downloaded data is still reused.

Notes

• The server must support HTTP range requests (Accept-Ranges: bytes).
• Using many connections to a single URL can put load on the server. In most cases 4 or fewer is enough.

Documentation

Index

• func Download(ctx context.Context, c *DownloadConfig, opts ...DownloadOption) error
• type CheckConfig
• type DownloadConfig
• type DownloadOption
  • func WithReferer(referer string) DownloadOption
  • func WithUserAgent(ua, version string) DownloadOption
• type Get
  • func New() *Get
  • func (g *Get) Run(ctx context.Context, version string, args []string) error
• type Range
  • func (r Range) BytesRange() string
• type Target
  • func Check(ctx context.Context, c *CheckConfig) (*Target, error)

Functions

func Download

func Download(ctx context.Context, c *DownloadConfig, opts ...DownloadOption) error

Types

type CheckConfig

type CheckConfig struct {
	URLs    []string
	Timeout time.Duration
	Client  *http.Client
}

CheckConfig is a configuration to check download target.

type DownloadConfig

type DownloadConfig struct {
	Filename      string
	Dirname       string
	ContentLength int64
	Procs         int
	URLs          []string
	Client        *http.Client
	// contains filtered or unexported fields
}

type DownloadOption

type DownloadOption func(c *DownloadConfig)

func WithReferer

func WithReferer(referer string) DownloadOption

func WithUserAgent

func WithUserAgent(ua, version string) DownloadOption

type Get

type Get struct {
	Output string
	Procs  int
	URLs   []string
	// contains filtered or unexported fields
}

Get struct

func New

func New() *Get

New for Get package

func (*Get) Run

func (g *Get) Run(ctx context.Context, version string, args []string) error

Run parses the command line arguments with cobra and executes the download. With -h / --help, cobra prints the usage and returns nil without running the download, so no "URL is required at least one" error is returned.

type Range

type Range struct {
	// contains filtered or unexported fields
}

Range struct for range access

func (Range) BytesRange

func (r Range) BytesRange() string

type Target

type Target struct {
	Filename      string
	ContentLength int64
	URLs          []string
}

Target represensts download target.

func Check

func Check(ctx context.Context, c *CheckConfig) (*Target, error)

Check checks be able to download from targets