From 033227b24319f8fdac46ae67dd57e059be923cf3 Mon Sep 17 00:00:00 2001
From: kayomn <kayomn@kayomn.net>
Date: Mon, 17 Oct 2022 15:48:56 +0100
Subject: [PATCH] Replace UTF-8 compares with standard memory compares

---
 src/ona/io.zig    | 37 +++++++++++++++++++++++++++++++++++++
 src/ona/table.zig | 35 +++++++----------------------------
 src/ona/utf8.zig  | 29 -----------------------------
 3 files changed, 44 insertions(+), 57 deletions(-)
 delete mode 100644 src/ona/utf8.zig

diff --git a/src/ona/io.zig b/src/ona/io.zig
index 7c8bee5..ca74716 100644
--- a/src/ona/io.zig
+++ b/src/ona/io.zig
@@ -286,6 +286,43 @@ pub const Writer = struct {
     }
 };
 
+///
+/// Returns `true` if `this_bytes` is the same length and contains the same data as `that_bytes`,
+/// otherwise `false`.
+///
+pub fn equalsBytes(this_bytes: []const u8, that_bytes: []const u8) bool {
+    return std.mem.eql(u8, this_bytes, that_bytes);
+}
+
+test "Equivalence of bytes" {
+    const bytes_sequence = &.{69, 42, 0};
+    const testing = std.testing;
+
+    try testing.expect(equalsBytes(bytes_sequence, bytes_sequence));
+    try testing.expect(!equalsBytes(bytes_sequence, &.{69, 42}));
+}
+
+///
+/// Returns a deterministic hash code compiled from each byte in `bytes`.
+///
+/// **Note** that this operation has `O(n)` time complexity.
+///
+pub fn hashBytes(bytes: []const u8) usize {
+    var hash = @as(usize, 5381);
+
+    for (bytes) |byte| hash = ((hash << 5) + hash) + byte;
+
+    return hash;
+}
+
+test "Hashing bytes" {
+    const bytes_sequence = &.{69, 42, 0};
+    const testing = std.testing;
+
+    try testing.expect(hashBytes(bytes_sequence) == hashBytes(bytes_sequence));
+    try testing.expect(hashBytes(bytes_sequence) != hashBytes(&.{69, 42}));
+}
+
 ///
 /// Writer that silently throws consumed data away and never fails.
 ///
diff --git a/src/ona/table.zig b/src/ona/table.zig
index 040a4f7..b844766 100644
--- a/src/ona/table.zig
+++ b/src/ona/table.zig
@@ -173,35 +173,14 @@ pub fn KeyContext(comptime Key: type) type {
     };
 }
 
-///
-/// Tests if the contents of `this_string` lexically equals the contents of `that_string`.
-///
-fn equalsString(this_string: []const u8, that_string: []const u8) bool {
-    return std.mem.eql(u8, this_string, that_string);
-}
-
-///
-/// Hashes `string` into a hash value of `usize`.
-///
-fn hashString(string: []const u8) usize {
-    var hash = @as(usize, 5381);
-
-    for (string) |byte| hash = ((hash << 5) + hash) + byte;
-
-    return hash;
-}
-
-///
-/// A [KeyContext] for handling `[]const u8` types.
-///
-pub const string_context = KeyContext([]const u8){
-    .hash = hashString,
-    .equals = equalsString,
-};
-
-test "Hashed table manipulation with string context" {
+test "Hashed table manipulation with bytes context" {
     const testing = std.testing;
-    var table = try Hashed([]const u8, u32, string_context).init(testing.allocator);
+    const io = @import("./io.zig");
+
+    var table = try Hashed([]const u8, u32, .{
+        .equals = io.equalsBytes,
+        .hash = io.hashBytes,
+    }).init(testing.allocator);
 
     defer table.deinit();
 
diff --git a/src/ona/utf8.zig b/src/ona/utf8.zig
deleted file mode 100644
index 352a9d6..0000000
--- a/src/ona/utf8.zig
+++ /dev/null
@@ -1,29 +0,0 @@
-const std = @import("std");
-const table = @import("./table.zig");
-
-///
-/// Tests if the contents of `this_utf8_sequence` lexically equals the contents of
-/// `that_utf8_sequence`.
-///
-pub fn equals(this_utf8_sequence: []const u8, that_utf8_sequence: []const u8) bool {
-    return std.mem.eql(u8, this_utf8_sequence, that_utf8_sequence);
-}
-
-///
-/// Returns a deterministic hash for `utf8_sequence`.
-///
-pub fn hash(utf8_sequence: []const u8) usize {
-    var utf8_hash = @as(usize, 5381);
-
-    for (utf8_sequence) |utf8_code| utf8_hash = ((utf8_hash << 5) + utf8_hash) + utf8_code;
-
-    return utf8_hash;
-}
-
-///
-/// A [table.KeyContext] for handling UTF-8 character sequences.
-///
-pub const key_context = table.KeyContext([]const u8){
-    .hash = hash,
-    .equals = equals,
-};