Super fast file downloader with multiple connections

npx ipull http://example.com/file.large

Features

• Download using parallels connections
• Pausing and resuming downloads
• Node.js and browser support
• Smart retry on fail
• CLI Progress bar
• Download statistics (speed, time left, etc.)

NodeJS API

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path', // or 'savePath' for full path
    cliProgress: true, // Show progress bar in the CLI (default: false)
    parallelStreams: 3 // Number of parallel connections (default: 3)
});

await downloader.download();

Browser support

Download a file in the browser using multiple connections

import {downloadFileBrowser} from "ipull/dist/browser.js";

const downloader = await downloadFileBrowser({
    url: 'https://example.com/file.large',
    acceptRangeIsKnown: true // cors origin request will not return the range header, but we can force it to be true (multi-connection download)
});

await downloader.download();
image.src = downloader.writeStream.resultAsBlobURL();

console.log(downloader.writeStream.result); // Uint8Array

Custom stream

You can use a custom stream

import {downloadFileBrowser} from "ipull/dist/browser.js";

const downloader = await downloadFileBrowser({
    url: 'https://example.com/file.large',
    onWrite: (cursor: number, buffer: Uint8Array, options) => {
        console.log(`Writing ${buffer.length} bytes at cursor ${cursor}, with options: ${JSON.stringify(options)}`);
    }
});

await downloader.download();
console.log(downloader.writeStream.result.length === 0); // true, because we write to a custom stream

CLI

Usage: ipull [options] [files...]

Pull/copy files from a remote server/local directory

Arguments:
  files                         Files to pull/copy

Options:
  -s --save [path]              Save location (directory/file)
  -c --connections [number]     Number of parallel connections (default: "4")
  -p --program [type]           The download strategy (choices: "stream", "chunks")
  -t --truncate-name            Truncate file names in the CLI status to make them appear shorter
  -V, --version                 output the version number
  -h, --help                    display help for command

Commands:
  set [options] [path] <value>  Set download locations

Set custom save directory

You can set a custom save directory by using the set command.

ipull set .zip ~/Downloads/zips

(use default to set the default save directory)

Advanced usage

Skip existing files

Skip downloading files that already exist in the save location and have the same size.

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path',
    skipExisting: true
});

Download file from parts

Consolidate multiple files parts into one file. Beneficial for downloading large files from servers that limit file size. (e.g. HuggingFace models)

import {downloadFile} from 'ipull';

const downloadParts = [
    "https://example.com/file.large-part-1",
    "https://example.com/file.large-part-2",
    "https://example.com/file.large-part-3",
];

const downloader = await downloadFile({
    partURLs: downloadParts,
    directory: './this/path',
    filename: 'file.large'
});

await downloader.download();

** The split must be binary and not a zip-split

Custom headers

You can set custom headers for the download request

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    savePath: './this/path/file.large',
    headers: {
        'Authorization': 'Bearer token 1'
    },
    // You can also add alternative headers in case of an 400-499 error
    tryHeaders: [
        {
            Authorization: 'Bearer token 2'
        }
    ]
});

await downloader.download();

Abort download

You can cancel the download by calling the close method (it will not delete the file).

If you want to also delete the file, you can call the closeAndDeleteFile method.

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path'
});

setTimeout(() => {
    downloader.close();
}, 5_000);

await downloader.download();

Pause & Resume download

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path'
});

setInterval(() => {
    downloader.pause();
    setTimeout(() => {
        downloader.resume();
    }, 5_000);
}, 10_000);

await downloader.download();

** The pause may take a few seconds to actually pause the download, because it waits for the current connections to finish

Error handling

If a network/file-system error occurs, the download will automatically retry with async-retry

If the maximum reties was reached the download will fail and an error will be thrown from the download() call:

import {downloadFile} from 'ipull';

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path'
});

try {
    await downloader.download();
} catch (error) {
    console.error(`Download failed: ${error.message}`);
}

Download Stuck

In some edge cases, the re-try mechanism may give the illusion that the download is stuck.

To debug this, disable the re-try mechanism:

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path',
    retry: {
        retries: 0
    }
});

Listening to events

Events are emitted using the EventEmitter pattern and can be listened to using the on method

interface DownloadEngineEvents {
    start: [];
    paused: [];
    resumed: [];
    progress: [FormattedStatus];
    save: [DownloadProgressInfo];
    finished: [];
    closed: [];
}

const downloader = await downloadFile({
    url: 'https://example.com/file.large',
    directory: './this/path'
});

downloader.on("progress", (progress) => {
    console.log(`Downloaded ${progress.transferred} bytes`);
});

Download multiple files

If you want to download multiple files, you can use the downloadSequence function.

By default, it will download files one by one, but you can set the parallel option to download them in parallel. It is better to download one file at a time if you are downloading from the same server (as it may limit the number of connections).

import {downloadFile, downloadSequence} from "ipull";

const downloader = await downloadSequence(
    {
        cliProgress: true,
        // parallelDownloads: 2, download 2 files in parallel, default is 1
    },
    downloadFile({
        url: "https://example.com/file1.large",
        directory: "."
    }),
    downloadFile({
        url: "https://example.com/file2.large",
        directory: "."
    }),
);

console.log(`Downloading ${downloader.downloads.length} files...`);
await downloader.download();

Custom progress bar

import {downloadFile, FormattedStatus} from "ipull";

function createProgressBar({fileName, ...data}: FormattedStatus) {
    return `${fileName} ${JSON.stringify(data)}`;
}

const downloader = await downloadFile({
    url: "https://example.com/file.large",
    directory: "./this/path",
    cliStyle: createProgressBar
});

await downloader.download();