Created
June 25, 2026 10:06
-
-
Save nubskr/26cd0cb7b6507b16a4f77b156995c366 to your computer and use it in GitHub Desktop.
Code reference to web crawler video
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| const std = @import("std"); | |
| const Io = std.Io; | |
| const stats_path = "crawl.stats"; | |
| const Job = struct { url: []const u8, depth: usize }; | |
| const State = struct { | |
| io: Io, | |
| gpa: std.mem.Allocator, | |
| arena: std.mem.Allocator, | |
| client: *std.http.Client, | |
| queue: Io.Queue(Job), | |
| max_pages: usize, | |
| max_depth: usize, | |
| started: std.atomic.Value(usize) = .init(0), | |
| pending: std.atomic.Value(usize) = .init(0), | |
| visited: std.atomic.Value(usize) = .init(0), | |
| }; | |
| pub fn main(init: std.process.Init) !void { | |
| const gpa = init.gpa; | |
| const arena = init.arena.allocator(); | |
| const args = try init.minimal.args.toSlice(arena); | |
| if (args.len < 2) { | |
| std.debug.print("usage: {s} <url> [max_pages] [workers] [max_depth]\n", .{args[0]}); | |
| return; | |
| } | |
| const max_pages = if (args.len > 2) @max(try std.fmt.parseInt(usize, args[2], 10), 1) else 50; | |
| const workers = if (args.len > 3) @max(try std.fmt.parseInt(usize, args[3], 10), 1) else 8; | |
| const max_depth = if (args.len > 4) try std.fmt.parseInt(usize, args[4], 10) else 3; | |
| var client: std.http.Client = .{ .allocator = gpa, .io = init.io }; | |
| defer client.deinit(); | |
| const queue_buffer = try gpa.alloc(Job, max_pages); | |
| defer gpa.free(queue_buffer); | |
| var state: State = .{ | |
| .io = init.io, | |
| .gpa = gpa, | |
| .arena = arena, | |
| .client = &client, | |
| .queue = .init(queue_buffer), | |
| .max_pages = max_pages, | |
| .max_depth = max_depth, | |
| }; | |
| enqueue(&state, try arena.dupe(u8, args[1]), 0); | |
| var group: Io.Group = .init; | |
| defer group.cancel(init.io); | |
| for (0..@min(workers, max_pages)) |_| try group.concurrent(init.io, worker, .{&state}); | |
| try group.await(init.io); | |
| } | |
| fn worker(s: *State) Io.Cancelable!void { | |
| while (true) { | |
| const job = s.queue.getOne(s.io) catch |err| switch (err) { | |
| error.Closed => return, | |
| error.Canceled => return error.Canceled, | |
| }; | |
| defer done(s, job.url, job.depth); | |
| const html = fetch(s.gpa, s.client, job.url) catch |err| { | |
| std.log.warn("{s}: {t}", .{ job.url, err }); | |
| continue; | |
| }; | |
| defer s.gpa.free(html); | |
| _ = s.visited.fetchAdd(1, .monotonic); | |
| writeStats(s, "visited", job.url, job.depth, html.len) catch {}; | |
| if (job.depth >= s.max_depth) continue; | |
| var pos: usize = 0; | |
| while (nextHref(html, &pos)) |href| { | |
| const url = resolve(s.arena, job.url, href) catch continue; | |
| enqueue(s, url, job.depth + 1); | |
| } | |
| } | |
| } | |
| fn enqueue(s: *State, url: []const u8, depth: usize) void { | |
| while (true) { | |
| const started = s.started.load(.monotonic); | |
| if (started >= s.max_pages) return; | |
| if (s.started.cmpxchgWeak(started, started + 1, .monotonic, .monotonic) == null) break; | |
| } | |
| _ = s.pending.fetchAdd(1, .monotonic); | |
| s.queue.putOneUncancelable(s.io, .{ .url = url, .depth = depth }) catch unreachable; | |
| writeStats(s, "found", url, depth, 0) catch {}; | |
| } | |
| fn done(s: *State, url: []const u8, depth: usize) void { | |
| if (s.pending.fetchSub(1, .monotonic) == 1) { | |
| s.queue.close(s.io); | |
| } | |
| writeStats(s, "done", url, depth, 0) catch {}; | |
| } | |
| fn writeStats(s: *State, event: []const u8, url: []const u8, depth: usize, bytes: usize) !void { | |
| var buf: [4096]u8 = undefined; | |
| const text = try std.fmt.bufPrint(&buf, | |
| \\pages found: {d}/{d} | |
| \\pages visited: {d} | |
| \\in flight: {d} | |
| \\max depth: {d} | |
| \\last event: {s} | |
| \\last depth: {d} | |
| \\last bytes: {d} | |
| \\last url: {s} | |
| \\ | |
| , .{ | |
| s.started.load(.monotonic), | |
| s.max_pages, | |
| s.visited.load(.monotonic), | |
| s.pending.load(.monotonic), | |
| s.max_depth, | |
| event, | |
| depth, | |
| bytes, | |
| url, | |
| }); | |
| try Io.Dir.cwd().writeFile(s.io, .{ .sub_path = stats_path, .data = text }); | |
| } | |
| fn fetch(gpa: std.mem.Allocator, client: *std.http.Client, url: []const u8) ![]u8 { | |
| var body: Io.Writer.Allocating = .init(gpa); | |
| defer body.deinit(); | |
| const result = try client.fetch(.{ .location = .{ .url = url }, .response_writer = &body.writer }); | |
| std.log.info("{d} {s}", .{ @intFromEnum(result.status), url }); | |
| if (result.status.class() != .success) return error.BadStatus; | |
| return body.toOwnedSlice(); | |
| } | |
| fn resolve(gpa: std.mem.Allocator, base_url: []const u8, href: []const u8) ![]u8 { | |
| const clean = std.mem.trim(u8, href, &std.ascii.whitespace); | |
| if (clean.len == 0 or clean[0] == '#') return error.Skip; | |
| var buf: [4096]u8 = undefined; | |
| if (clean.len > buf.len) return error.Skip; | |
| @memcpy(buf[0..clean.len], clean); | |
| var rest: []u8 = &buf; | |
| const uri = try (try std.Uri.parse(base_url)).resolveInPlace(clean.len, &rest); | |
| const http = | |
| std.ascii.eqlIgnoreCase(uri.scheme, "http") or | |
| std.ascii.eqlIgnoreCase(uri.scheme, "https"); | |
| if (!http) return error.Skip; | |
| return std.fmt.allocPrint(gpa, "{f}", .{&uri}); | |
| } | |
| fn nextHref(html: []const u8, pos: *usize) ?[]const u8 { | |
| const hit = std.mem.cut(u8, html[pos.*..], "href=\"") orelse return null; | |
| const after = hit[1]; | |
| const href, const rest = std.mem.cutScalar(u8, after, '"') orelse return null; | |
| pos.* = html.len - rest.len; | |
| return href; | |
| } | |
| test "href demo parser" { | |
| const html = "<a href=\"/a\"></a><a href=\"/b\"></a>"; | |
| var pos: usize = 0; | |
| try std.testing.expectEqualStrings("/a", nextHref(html, &pos).?); | |
| try std.testing.expectEqualStrings("/b", nextHref(html, &pos).?); | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment