Resume Example

Resume interrupted downloads.

Overview

Demonstrates:

• Resume configuration
• Detecting resumed downloads
• Fallback to fresh download
• File exists handling

Running

zig build run-resume

Configuration

var config = downloader.Config.default();
config.resume_downloads = true;
config.file_exists_action = .resume_or_overwrite;

Detecting Resume

Check is_resumed in progress callback:

fn progressCallback(p: downloader.Progress) bool {
    if (p.is_resumed) {
        std.debug.print("Resuming from byte {d}\n", .{p.start_offset});
    }
    // Show progress...
    return true;
}

Example

const std = @import("std");
const downloader = @import("downloader");

const url = "https://example.com/large-file.iso";
const output = "large-file.iso";

pub fn main() !void {
    var gpa = std.heap.GeneralPurposeAllocator(.{}){};
    defer _ = gpa.deinit();

    var config = downloader.Config.default();
    config.resume_downloads = true;
    config.file_exists_action = .resume_or_overwrite;

    var client = try downloader.Client.init(gpa.allocator(), config);
    defer client.deinit();

    const bytes = client.download(url, output, progressCallback) catch |err| {
        if (err == error.ResumeNotSupported) {
            std.debug.print("Server doesn't support resume\n", .{});
            return tryFreshDownload(gpa.allocator());
        }
        return err;
    };

    std.debug.print("\nDownloaded {d} bytes this session\n", .{bytes});
}

fn progressCallback(p: downloader.Progress) bool {
    const pct = p.percentage() orelse 0;

    if (p.is_resumed and p.start_offset > 0) {
        std.debug.print("\rResuming: {d:.1}%", .{pct});
    } else {
        std.debug.print("\rDownloading: {d:.1}%", .{pct});
    }

    return true;
}

fn tryFreshDownload(allocator: std.mem.Allocator) !void {
    std.fs.cwd().deleteFile(output) catch {};

    var config = downloader.Config.noResume();
    var client = try downloader.Client.init(allocator, config);
    defer client.deinit();

    _ = try client.download(url, output, progressCallback);
}

How Resume Works

1. Checks for existing partial file
2. Sends HEAD request to verify server support
3. If supported, sends GET with Range header
4. Server responds with 206 Partial Content
5. Download continues from last byte

Requirements

• Server must support Range requests
• Partial file must exist on disk
• File must not have changed on server

Fallback Behavior

When resume fails:

• Server doesn't support: Start fresh
• File modified: Start fresh
• Partial deleted: Start fresh

See Also

• Resume Guide - Detailed explanation
• Configuration - FileExistsAction options