From: Brendan Hansen <brendan.f.hansen@gmail.com>
Date: Wed, 8 Mar 2023 20:38:31 +0000 (-0600)
Subject: added: `core.encoding.utf8` library
X-Git-Url: https://git.brendanfh.com/?a=commitdiff_plain;h=830f0c06513589bfd5cad4df6e267d94904f12fe;p=onyx.git

added: `core.encoding.utf8` library
---

diff --git a/core/encoding/utf8.onyx b/core/encoding/utf8.onyx
new file mode 100644
index 00000000..2bf037ff
--- /dev/null
+++ b/core/encoding/utf8.onyx
@@ -0,0 +1,163 @@
+package core.encoding.utf8
+
+use core {string, array, iter}
+
+rune :: i32
+
+Max_Chars :: 4
+
+append_rune :: (s: &dyn_str, r: rune) -> str {
+    buffer: [Max_Chars] u8;
+    string.append(s, encode_rune(buffer, r));
+    return *s;
+}
+
+advance_rune :: #match #local {}
+
+#overload
+advance_rune :: (s: str, n := 1) -> str {
+    t := s;
+    for n {
+        t = string.advance(t, rune_length_from_first_byte(t[0]));
+    }
+    return t;
+}
+
+#overload
+advance_rune :: (s: &str, n := 1) {
+    for n {
+        string.advance(s, rune_length_from_first_byte(s.data[0]));
+    }
+}
+
+next_rune :: (s: str) -> str {
+    return s.data[0 .. rune_length_from_first_byte(s[0])];
+}
+
+decode_rune :: (s: str) -> (r: rune, size: i32) {
+    len := rune_length_from_first_byte(s[0]);
+    if len == 1 do return (cast(rune) s[0]), 1;
+
+    mask := 0xFF >> (1 + len);
+    r: rune = ((cast(rune) s[0]) & mask);
+
+    for len - 1 {
+        r = r << 6;
+        r = r | ~~(s[it + 1] & 0x3F);
+    }
+
+    return r, len;
+}
+
+decode_last_rune :: (s: str) -> (r: rune, size: i32) {
+    i := s.length - 1;
+    while i >= 0 && (s[i] & 0xC0 == 0x80) {
+        i -= 1;
+    }
+
+    return decode_rune(s[i..s.length]);
+}
+
+encode_rune :: (buffer: [] u8, r: rune) -> str {
+    if r <= 0x7F {
+        buffer[0] = ~~ r;
+        return buffer[0 .. 1];
+    }
+
+    if r <= 0x7FF {
+        buffer[0] = ~~(0xC0 | ((r >> 6) & 0x1F));
+        buffer[1] = ~~(0x80 | (r & 0x3F));
+        return buffer[0 .. 2];
+    }
+
+    if r >= 0xD800 && r <= 0xDFFF {
+        return buffer[0 .. 0];
+    }
+
+    if r <= 0xFFFF {
+        buffer[0] = ~~(0xE0 | ((r >> 12) & 0x0F));
+        buffer[1] = ~~(0x80 | ((r >> 6) & 0x3F));
+        buffer[2] = ~~(0x80 | (r & 0x3F));
+        return buffer[0 .. 3];
+    }
+
+    if r <= 0x10FFFF {
+        buffer[0] = ~~(0xF0 | ((r >> 18) & 0x07));
+        buffer[1] = ~~(0x80 | ((r >> 12) & 0x3F));
+        buffer[2] = ~~(0x80 | ((r >> 6) & 0x3F));
+        buffer[3] = ~~(0x80 | (r & 0x3F));
+        return buffer[0 .. 4];
+    }
+
+    return buffer[0 .. 0];
+}
+
+full_rune :: (buffer: [] u8) -> bool {
+    len := rune_length_from_first_byte(buffer[0]);
+
+    // If there are not enough bytes, then this cannot
+    // be a valid rune.
+    if buffer.length < len do return false;
+
+    return array.every(buffer[1 .. buffer.length], #(it & 0xC0 == 0x80));
+}
+
+rune_count :: (s: str) -> i32 {
+    w := s;
+    count := 0;
+    while !string.empty(w) {
+        advance_rune(&w);
+        count += 1;
+    }
+
+    return count;
+}
+
+rune_length :: (r: rune) -> i32 {
+    if r <= 0x7f     do return 1;
+    if r <= 0x7ff    do return 2;
+    if r <= 0xffff   do return 3;
+    if r <= 0x10ffff do return 4;
+
+    return -1;
+}
+
+rune_length_from_first_byte :: (b: u8) -> i32 {
+    if b & 0x80 == 0x00 do return 1;
+    if b & 0xE0 == 0xC0 do return 2;
+    if b & 0xF0 == 0xE0 do return 3;
+    if b & 0xF8 == 0xF0 do return 4;
+    return -1;
+}
+
+rune_is_start :: (b: u8) -> bool {
+    return rune_length_from_first_byte(b) > 1;
+}
+
+runes :: (s: str) -> Iterator(rune) {
+    return iter.generator(
+        &.{ s = s },
+        ctx => {
+            if string.empty(ctx.s) do return 0, false;
+
+            r, len := decode_rune(ctx.s);
+            string.advance(&ctx.s, len);
+
+            return r, true;
+        }            
+    );
+}
+
+slice :: (s: str, low, high: i32) -> str {
+    advanced := advance_rune(s, low);
+    tmp := advanced;
+
+    bytes := 0;
+    for high - low {
+        _, b := decode_rune(tmp);
+        bytes += b;
+        string.advance(&tmp, b);
+    }
+
+    return advanced.data[0 .. bytes];
+}
diff --git a/core/std.onyx b/core/std.onyx
index fc938665..24ab4075 100644
--- a/core/std.onyx
+++ b/core/std.onyx
@@ -51,6 +51,7 @@ package core
 
 #load "./time/date"
 #load "./encoding/base64"
+#load "./encoding/utf8"
 
 #load "./runtime/common"
 
@@ -119,4 +120,4 @@ package core
 
 #if runtime.runtime == .Js {
     #load "./runtime/platform/js/platform"
-}
\ No newline at end of file
+}
diff --git a/tests/utf8_test b/tests/utf8_test
new file mode 100644
index 00000000..7e8e565c
--- /dev/null
+++ b/tests/utf8_test
@@ -0,0 +1,3 @@
+🂦 🂧 🂨 🂩 🂪 🂫 🂬 🂭 🂮 🂯 🂰 🂱 🂲 🂳 🂴 🂵 🂶 🂷 🂸 🂹 🂺 🂻 🂼 🂽 🂾 🂿 🃀 🃁 🃂 🃃 🃄 🃅 🃆 🃇 🃈 🃉 🃊 🃋 🃌 🃍 🃎 🃏 🃐 🃑 🃒 🃓 🃔 🃕 🃖 🃗 🃘 🃙 🃚 🃛 🃜 🃝 🃞 🃟 
+🂣🂤🂥🂦🂧🂨🂩
+🂦
diff --git a/tests/utf8_test.onyx b/tests/utf8_test.onyx
new file mode 100644
index 00000000..1449b686
--- /dev/null
+++ b/tests/utf8_test.onyx
@@ -0,0 +1,22 @@
+use core
+use core.encoding {utf8}
+
+main :: () {
+    output: dyn_str;
+    for i: 0x1F0A0 .. 0x1F0E0 {
+        utf8.append_rune(&output, i);
+    }
+
+    tmp: str = output;
+    utf8.advance_rune(&tmp, 6);
+
+    for utf8.runes(tmp) {
+        buf: [4] u8;
+        printf("{} ", utf8.encode_rune(buf, it));
+    }
+    print("\n");
+
+    println(utf8.slice(output, 3, 10));
+
+    println(utf8.next_rune(tmp));
+}