add two LUT-based implementations of isSymbol

src/benchmark/IsSymbol.zig

@@ -0,0 +1,172 @@
+//! This benchmark tests the throughput of grapheme break calculation.
+//! This is a common operation in terminal character printing for terminals
+//! that support grapheme clustering.
+const IsSymbol = @This();
+
+const std = @import("std");
+const builtin = @import("builtin");
+const assert = std.debug.assert;
+const Allocator = std.mem.Allocator;
+const Benchmark = @import("Benchmark.zig");
+const options = @import("options.zig");
+const UTF8Decoder = @import("../terminal/UTF8Decoder.zig");
+const symbols1 = @import("../unicode/symbols1.zig");
+const symbols2 = @import("../unicode/symbols2.zig");
+
+const log = std.log.scoped(.@"is-symbol-bench");
+
+opts: Options,
+
+/// The file, opened in the setup function.
+data_f: ?std.fs.File = null,
+
+pub const Options = struct {
+    /// Which test to run.
+    mode: Mode = .ziglyph,
+
+    /// The data to read as a filepath. If this is "-" then
+    /// we will read stdin. If this is unset, then we will
+    /// do nothing (benchmark is a noop). It'd be more unixy to
+    /// use stdin by default but I find that a hanging CLI command
+    /// with no interaction is a bit annoying.
+    data: ?[]const u8 = null,
+};
+
+pub const Mode = enum {
+    /// "Naive" ziglyph implementation.
+    ziglyph,
+
+    /// Ghostty's table-based approach.
+    table1,
+    table2,
+};
+
+/// Create a new terminal stream handler for the given arguments.
+pub fn create(
+    alloc: Allocator,
+    opts: Options,
+) !*IsSymbol {
+    const ptr = try alloc.create(IsSymbol);
+    errdefer alloc.destroy(ptr);
+    ptr.* = .{ .opts = opts };
+    return ptr;
+}
+
+pub fn destroy(self: *IsSymbol, alloc: Allocator) void {
+    alloc.destroy(self);
+}
+
+pub fn benchmark(self: *IsSymbol) Benchmark {
+    return .init(self, .{
+        .stepFn = switch (self.opts.mode) {
+            .ziglyph => stepZiglyph,
+            .table1 => stepTable1,
+            .table2 => stepTable1,
+        },
+        .setupFn = setup,
+        .teardownFn = teardown,
+    });
+}
+
+fn setup(ptr: *anyopaque) Benchmark.Error!void {
+    const self: *IsSymbol = @ptrCast(@alignCast(ptr));
+
+    // Open our data file to prepare for reading. We can do more
+    // validation here eventually.
+    assert(self.data_f == null);
+    self.data_f = options.dataFile(self.opts.data) catch |err| {
+        log.warn("error opening data file err={}", .{err});
+        return error.BenchmarkFailed;
+    };
+}
+
+fn teardown(ptr: *anyopaque) void {
+    const self: *IsSymbol = @ptrCast(@alignCast(ptr));
+    if (self.data_f) |f| {
+        f.close();
+        self.data_f = null;
+    }
+}
+
+fn stepZiglyph(ptr: *anyopaque) Benchmark.Error!void {
+    const self: *IsSymbol = @ptrCast(@alignCast(ptr));
+
+    const f = self.data_f orelse return;
+    var r = std.io.bufferedReader(f.reader());
+    var d: UTF8Decoder = .{};
+    var buf: [4096]u8 = undefined;
+    while (true) {
+        const n = r.read(&buf) catch |err| {
+            log.warn("error reading data file err={}", .{err});
+            return error.BenchmarkFailed;
+        };
+        if (n == 0) break; // EOF reached
+
+        for (buf[0..n]) |c| {
+            const cp_, const consumed = d.next(c);
+            assert(consumed);
+            if (cp_) |cp| {
+                std.mem.doNotOptimizeAway(symbols1.isSymbol(cp));
+            }
+        }
+    }
+}
+
+fn stepTable1(ptr: *anyopaque) Benchmark.Error!void {
+    const self: *IsSymbol = @ptrCast(@alignCast(ptr));
+
+    const f = self.data_f orelse return;
+    var r = std.io.bufferedReader(f.reader());
+    var d: UTF8Decoder = .{};
+    var buf: [4096]u8 = undefined;
+    while (true) {
+        const n = r.read(&buf) catch |err| {
+            log.warn("error reading data file err={}", .{err});
+            return error.BenchmarkFailed;
+        };
+        if (n == 0) break; // EOF reached
+
+        for (buf[0..n]) |c| {
+            const cp_, const consumed = d.next(c);
+            assert(consumed);
+            if (cp_) |cp| {
+                std.mem.doNotOptimizeAway(symbols1.table.get(cp));
+            }
+        }
+    }
+}
+
+fn stepTable2(ptr: *anyopaque) Benchmark.Error!void {
+    const self: *IsSymbol = @ptrCast(@alignCast(ptr));
+
+    const f = self.data_f orelse return;
+    var r = std.io.bufferedReader(f.reader());
+    var d: UTF8Decoder = .{};
+    var buf: [4096]u8 = undefined;
+    while (true) {
+        const n = r.read(&buf) catch |err| {
+            log.warn("error reading data file err={}", .{err});
+            return error.BenchmarkFailed;
+        };
+        if (n == 0) break; // EOF reached
+
+        for (buf[0..n]) |c| {
+            const cp_, const consumed = d.next(c);
+            assert(consumed);
+            if (cp_) |cp| {
+                std.mem.doNotOptimizeAway(symbols2.table.get(cp));
+            }
+        }
+    }
+}
+
+test IsSymbol {
+    const testing = std.testing;
+    const alloc = testing.allocator;
+
+    const impl: *IsSymbol = try .create(alloc, .{});
+    defer impl.destroy(alloc);
+
+    const bench = impl.benchmark();
+    _ = try bench.run(.once);
+}