Add bencoding decoding library

6 files changed, 291 insertions, 0 deletions

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..e73c965
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+zig-cache/
+zig-out/
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..d0f1eeb
--- /dev/null
+++ b/README.md
@@ -0,0 +1,7 @@
+# zbt 
+
+Bittorrent client using Zig and Capy for UI.
+
+Somewhat inspired by codecrafters bittorrent course.
+
+Built as a learning exercise.
\ No newline at end of file
diff --git a/build.zig b/build.zig
new file mode 100644
index 0000000..96ba17c
--- /dev/null
+++ b/build.zig
@@ -0,0 +1,35 @@
+const std = @import("std");
+
+pub fn build(b: *std.Build) void {
+    const target = b.standardTargetOptions(.{});
+    const optimize = b.standardOptimizeOption(.{});
+
+    const exe = b.addExecutable(.{
+        .name = "zbt",
+        .root_source_file = .{ .path = "src/main.zig" },
+        .target = target,
+        .optimize = optimize,
+    });
+
+    b.installArtifact(exe);
+    const run_cmd = b.addRunArtifact(exe);
+    run_cmd.step.dependOn(b.getInstallStep());
+
+    if (b.args) |args| {
+        run_cmd.addArgs(args);
+    }
+
+    const run_step = b.step("run", "Run the app");
+    run_step.dependOn(&run_cmd.step);
+
+    const unit_tests = b.addTest(.{
+        .root_source_file = .{ .path = "src/main.zig" },
+        .target = target,
+        .optimize = optimize,
+    });
+    //b.installArtifact(unit_tests); // Useful if you want to debug the test binary with lldb or something
+
+    const run_unit_tests = b.addRunArtifact(unit_tests);
+    const test_step = b.step("test", "Run unit tests");
+    test_step.dependOn(&run_unit_tests.step);
+}
diff --git a/sample.torrent b/sample.torrent
new file mode 100644
index 0000000..44d1411
--- /dev/null
+++ b/sample.torrent
@@ -0,0 +1 @@
+d8:announce55:http://bittorrent-test-tracker.codecrafters.io/announce10:created by13:mktorrent 1.14:infod6:lengthi92063e4:name10:sample.txt12:piece lengthi32768e6:pieces60:�v�z*����kg&���-n"u��vfVsn���R��5�
�z����	r'�����ee
\ No newline at end of file
diff --git a/src/bencode.zig b/src/bencode.zig
new file mode 100644
index 0000000..b78ee9f
--- /dev/null
+++ b/src/bencode.zig
@@ -0,0 +1,239 @@
+//! Bencoding
+//! See specification here https://wiki.theory.org/BitTorrentSpecification#Bencoding
+
+const std = @import("std");
+
+pub const Error = error.Malformatted || std.io.AnyReader.Error;
+
+// All content is owned by the BValue and must be freed with deinit.
+pub const BValue = union(enum) {
+    string: []const u8,
+    int: i64,
+    list: std.ArrayList(BValue),
+    dict: std.StringArrayHashMap(BValue),
+
+    pub fn deinit(self: *BValue, a: std.mem.Allocator) void {
+        switch (self.*) {
+            .string => |s| {
+                a.free(s);
+            },
+            .list => |*l| {
+                for (l.items) |*i| {
+                    i.deinit(a);
+                }
+                l.deinit();
+            },
+            .dict => |*d| {
+                var it = d.iterator();
+                while (it.next()) |entry| {
+                    a.free(entry.key_ptr.*);
+                    entry.value_ptr.*.deinit(a);
+                }
+                d.deinit();
+            },
+            .int => {},
+        }
+    }
+};
+
+pub fn bdecodeBuf(a: std.mem.Allocator, buf: []const u8) !BValue {
+    var fbs = std.io.fixedBufferStream(buf);
+    return try bdecode(a, fbs.reader());
+}
+
+pub fn bdecode(a: std.mem.Allocator, base_reader: anytype) anyerror!BValue {
+    var reader = PeekStream.init(base_reader.any());
+    return bdecodeInner(a, &reader, 0);
+}
+
+const PeekStream = std.io.PeekStream(.{ .Static = 1 }, std.io.AnyReader);
+
+// Note: uses defined types only to avoid trying to recursively evaulate this function
+// at compile time, otherwise we run into https://github.com/ziglang/zig/issues/13724
+fn bdecodeInner(a: std.mem.Allocator, peekStream: *PeekStream, depth: u32) !BValue {
+    if (depth > 100) {
+        // TODO diagnostic...
+        return error.Malformatted;
+    }
+    var reader = peekStream.reader();
+    var byte = try reader.readByte();
+    if (std.ascii.isDigit(byte)) {
+        try peekStream.putBackByte(byte);
+        return .{ .string = try readString(a, peekStream) };
+    } else {
+        switch (byte) {
+            'i' => {
+                const max_len = comptime std.fmt.comptimePrint("{}", .{std.math.minInt(i64)}).len;
+                var s = reader.readUntilDelimiterAlloc(a, 'e', max_len) catch return error.Malformatted;
+                defer a.free(s);
+                const i = std.fmt.parseInt(i64, s, 10) catch return error.Malformatted;
+                return .{ .int = i };
+            },
+            'l' => {
+                var list = std.ArrayList(BValue).init(a);
+                errdefer {
+                    for (list.items) |*i| {
+                        i.deinit(a);
+                    }
+                    list.deinit();
+                }
+                while (true) {
+                    const b2 = try reader.readByte();
+                    if (b2 == 'e') break;
+                    try peekStream.putBackByte(b2);
+                    var val = try bdecodeInner(a, peekStream, depth + 1);
+                    errdefer val.deinit(a);
+                    try list.append(val);
+                }
+                return .{ .list = list };
+            },
+            'd' => {
+                var dict = std.StringArrayHashMap(BValue).init(a);
+                errdefer {
+                    var it = dict.iterator();
+                    while (it.next()) |entry| {
+                        a.free(entry.key_ptr.*);
+                        entry.value_ptr.*.deinit(a);
+                    }
+                    dict.deinit();
+                }
+                lp: while (true) {
+                    const b2 = try reader.readByte();
+                    if (b2 == 'e') break :lp;
+                    try peekStream.putBackByte(b2);
+                    var key = try readString(a, peekStream);
+                    errdefer a.free(key);
+                    var val = try bdecode(a, reader);
+                    errdefer val.deinit(a);
+                    try dict.put(key, val);
+                }
+                return .{ .dict = dict };
+            },
+            else => return error.Malformatted, // TODO diagnostics
+        }
+    }
+}
+
+// Result is owned by the caller and must be freed
+fn readString(a: std.mem.Allocator, peekStream: *PeekStream) ![]const u8 {
+    var reader = peekStream.reader();
+    const max_len = comptime std.fmt.comptimePrint("{}", .{std.math.maxInt(usize)}).len;
+    const str_len_s = reader.readUntilDelimiterAlloc(a, ':', max_len) catch {
+        return error.Malformatted;
+    };
+    defer a.free(str_len_s);
+    var strlen = std.fmt.parseInt(usize, str_len_s, 10) catch return error.Malformatted;
+    var string = try a.alloc(u8, strlen);
+    errdefer a.free(string);
+    reader.readNoEof(string) catch return error.Malformatted;
+    return string;
+}
+
+test "bdecode empty" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.EndOfStream, bdecodeBuf(a, ""));
+}
+
+test "bdecode too short" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "1"));
+}
+
+test "bdecode plain number" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "12"));
+}
+
+test "bdecode garbage" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "xz1234"));
+}
+
+test "bdecode number" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "i123e");
+    defer bval.deinit(a);
+    try std.testing.expectEqualDeep(BValue{ .int = 123 }, bval);
+}
+
+test "bdecode number negative" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "i-123e");
+    defer bval.deinit(a);
+    try std.testing.expectEqualDeep(BValue{ .int = -123 }, bval);
+}
+
+test "bdecode number empty" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "ie"));
+}
+
+test "bdecode number just sign" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "i-e"));
+}
+
+test "bdecode number no end" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "i123123671283"));
+}
+
+test "bdecode number out of range" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "i9223372036854775808e"));
+}
+
+test "bdecode string" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "5:hello");
+    defer bval.deinit(a);
+    try std.testing.expectEqualDeep(BValue{ .string = "hello" }, bval);
+}
+test "bdecode string too short" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "5:hell"));
+}
+
+test "bdecode list" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "l5:hello5:worlde");
+    defer bval.deinit(a);
+    try std.testing.expectEqual(@as(usize, 2), bval.list.items.len);
+    try std.testing.expectEqualStrings("hello", bval.list.items[0].string);
+    try std.testing.expectEqualStrings("world", bval.list.items[1].string);
+}
+
+test "invalid list" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.EndOfStream, bdecodeBuf(a, "l5:hello5:world")); // missing end
+}
+
+test "dict" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "d5:hello5:worlde");
+    defer bval.deinit(a);
+    var v = bval.dict.getPtr("hello") orelse return error.TestExpectedNotNull;
+    try std.testing.expectEqualStrings("world", v.string);
+}
+
+test "invalid dict no value" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "d5:hello5:world2:hie"));
+}
+
+test "invalid dict wrong key type" {
+    var a = std.testing.allocator;
+    try std.testing.expectError(error.Malformatted, bdecodeBuf(a, "di32e5:helloe"));
+}
+
+test "nested structure" {
+    var a = std.testing.allocator;
+    var bval = try bdecodeBuf(a, "d5:hello5:world2:hili123ei456el4:nesteee");
+    defer bval.deinit(a);
+    var v = bval.dict.getPtr("hello") orelse return error.TestExpectedNotNull;
+    try std.testing.expectEqualStrings("world", v.string);
+    var v2 = bval.dict.getPtr("hi") orelse return error.TestExpectedNotNull;
+    try std.testing.expectEqualDeep(v2.*.list.items[0], BValue{ .int = 123 });
+    try std.testing.expectEqualDeep(v2.*.list.items[1], BValue{ .int = 456 });
+    try std.testing.expectEqualStrings("nest", v2.*.list.items[2].list.items[0].string);
+}
diff --git a/src/main.zig b/src/main.zig
new file mode 100644
index 0000000..fa8a1b6
--- /dev/null
+++ b/src/main.zig
@@ -0,0 +1,7 @@
+const std = @import("std");
+
+pub fn main() !void {}
+
+test {
+    _ = @import("bencode.zig");
+}