Golang Script to extract file direct download link for Google Drive files with any size.

README.md

Golang Script to extract file direct download link for Google Drive files with any size without encountering the dump Google Drive can't scan this file for viruses webpage.

drive_file_downloader.go

package main

import (
	"context"
	"encoding/base64"
	"flag"
	"fmt"
	"net/http"
	"net/url"
	"os"
	"regexp"
	"strings"

	"golang.org/x/net/html"
	"google.golang.org/api/drive/v2"
	"google.golang.org/api/option"
)

var (
	ctx = context.Background()
)

func main() {
	var shareUri string
	flag.StringVar(&shareUri, "u", "", "share uri")
	flag.Parse()
	if shareUri == "" {
		panic("share uri is required")
	}

	fmt.Print(`
-----------------------------------------------
You are using a beta version of this tool.
The tool is not meant to be distributed or used in production.
The tool is provided as is, without any warranty.

Version: 0.0.2
By: Ahmed Mahmoud
-----------------------------------------------

`)
	apiKey := os.Getenv("GOOGLE_API_KEY")
	if apiKey == "" {
		panic(`
GOOGLE_API_KEY is required, You can get an API key from https://console.developers.google.com/apis/credentials
After creating the API key, make sure to enable the following APIs:
- Google Drive API

Then export the API key as an environment variable:
Example: export GOOGLE_API_KEY=your-api-key
`)
	}

	valid := ValidateShareUrl(shareUri)
	if !valid {
		panic("invalid share uri: " + shareUri)
	}

	fileId, err := ExtractFileId(shareUri)
	if err != nil {
		panic(err)
	}

	srv, err := drive.NewService(ctx, option.WithAPIKey(apiKey))
	if err != nil {
		panic("Unable to retrieve Drive client: %v" + err.Error())
	}

	file, err := srv.Files.Get(fileId).Do()
	if err != nil {
		panic(err)
	}

	client := http.DefaultClient
	downloadUrl, err := GenerateDownloadUrl(client, file)
	if err != nil {
		panic(err)
	}

	fmt.Println("Download URL: " + downloadUrl)
}

func ValidateShareUrl(shareUri string) bool {
	regexPattern := `https:\/\/drive\.google\.com\/file\/d\/[a-zA-Z0-9_-]+\/view`
	match, _ := regexp.MatchString(regexPattern, shareUri)
	return match
}

func ExtractFileId(shareUri string) (string, error) {
	u, err := url.Parse(shareUri)
	if err != nil {
		return "", err
	}
	if u.Host != "drive.google.com" {
		return "", fmt.Errorf("invalid share uri: %s", shareUri)
	}
	s := strings.Split(shareUri, "d/")
	if len(s) != 2 {
		return "", fmt.Errorf("invalid share uri: %s", shareUri)
	}
	ss := strings.Split(s[1], "/")
	if len(ss) < 1 {
		return "", fmt.Errorf("invalid share uri: %s", shareUri)
	}
	return ss[0], nil
}

func DecodeBase64(s string) string {
	s = strings.Trim(s, " ")
	s = strings.Trim(s, "\"")
	d, err := base64.StdEncoding.DecodeString(s)
	if err != nil {
		return ""
	}
	return string(d)
}

func GenerateDownloadUrl(client *http.Client, file *drive.File) (string, error) {
	fileSize := file.FileSize
	fileId := file.Id
	// check if file size is larger than 100MB
	if fileSize < 100*(1<<20) {
		// simple download
		// https://drive.google.com/uc?export=download&id=file-id
		downloadUrl := fmt.Sprintf("https://drive.google.com/uc?export=download&id=%s", fileId)
		return downloadUrl, nil
	}

	// large file download, more complicated download
	// Example: https://drive.google.com/u/0/uc?id=XXXXXXXXBKPx3G5_UZWA79g79ncqEfQ&export=download
	scanFailedUrl := fmt.Sprintf("https://drive.google.com/uc?export=download&id=%s", fileId)
	res, err := client.Get(scanFailedUrl)
	if err != nil {
		return "", err
	}
	defer res.Body.Close()

	if res.StatusCode != http.StatusOK {
		return "", fmt.Errorf("failed to open download url: %s", scanFailedUrl)
	}

	// check content type, must be html
	contentType := res.Header.Get("Content-Type")
	if contentType != "text/html; charset=utf-8" {
		return "", fmt.Errorf("download page content type is not html: %s", contentType)
	}

	downloadPage, err := html.Parse(res.Body)
	if err != nil {
		return "", err
	}

	// TODO: Check if the page is "Google Drive Quota Exceeded or Limit Reached"

	formElem := GetElementByID(downloadPage, "download-form")
	if formElem == nil {
		return "", fmt.Errorf("form element was not found in the download page")
	}

	formAction := GetAttribute(formElem, "action")
	if formAction == "" {
		return "", fmt.Errorf("form action was not found in the download page")
	}

	return formAction, nil
}

func GetElementByID(n *html.Node, id string) *html.Node {
	if n.Type == html.ElementNode {
		if attr := GetAttribute(n, "id"); attr == id {
			return n
		}
	}
	for c := n.FirstChild; c != nil; c = c.NextSibling {
		if n := GetElementByID(c, id); n != nil {
			return n
		}
	}
	return nil
}

func GetAttribute(n *html.Node, key string) string {
	for _, attr := range n.Attr {
		if attr.Key == key {
			return attr.Val
		}
	}
	return ""
}