From a8ceed4b92cbc8f5f40fb6429ecd48ca30f3c02d Mon Sep 17 00:00:00 2001
From: Rui Ueyama <ruiu@cs.stanford.edu>
Date: Sun, 5 Jul 2020 20:49:58 +0900
Subject: [PATCH] Add UTF-32 string literal

---
 test/unicode.c | 11 +++++++++++
 tokenize.c     | 29 +++++++++++++++++++++++++++++
 2 files changed, 40 insertions(+)

diff --git a/test/unicode.c b/test/unicode.c
index 4f14a97..1bd99a3 100644
--- a/test/unicode.c
+++ b/test/unicode.c
@@ -41,6 +41,17 @@ int main() {
   ASSERT(u'b', u"βb"[1]);
   ASSERT(0, u"βb"[2]);
 
+  ASSERT(4, sizeof(U""));
+  ASSERT(20, sizeof(U"\xffzzz"));
+  ASSERT(0, memcmp(U"", "\0\0\0\0", 4));
+  ASSERT(0, memcmp(U"abc", "a\0\0\0b\0\0\0c\0\0\0\0\0\0\0", 16));
+  ASSERT(0, memcmp(U"日本語", "\345e\0\0,g\0\0\236\212\0\0\0\0\0\0", 16));
+  ASSERT(0, memcmp(U"🍣", "c\363\001\0\0\0\0\0", 8));
+  ASSERT(u'β', U"βb"[0]);
+  ASSERT(u'b', U"βb"[1]);
+  ASSERT(0, U"βb"[2]);
+  ASSERT(1, U"\xffffffff"[0] >> 31);
+
   printf("OK\n");
   return 0;
 }
diff --git a/tokenize.c b/tokenize.c
index 647f9a4..d4bdd32 100644
--- a/tokenize.c
+++ b/tokenize.c
@@ -257,6 +257,28 @@ static Token *read_utf16_string_literal(Token *cur, char *start) {
   return tok;
 }
 
+// Read a UTF-8-encoded string literal and transcode it in UTF-32.
+//
+// UTF-32 is a fixed-width encoding for Unicode. Each code point is
+// encoded in 4 bytes.
+static Token *read_utf32_string_literal(Token *cur, char *start, Type *ty) {
+  char *end = string_literal_end(start + 1);
+  uint32_t *buf = calloc(4, end - start);
+  int len = 0;
+
+  for (char *p = start + 1; p < end;) {
+    if (*p == '\\')
+      buf[len++] = read_escaped_char(&p, p + 1);
+    else
+      buf[len++] = decode_utf8(&p, p);
+  }
+
+  Token *tok = new_token(TK_STR, cur, start, end - start + 1);
+  tok->ty = array_of(ty, len + 1);
+  tok->str = (char *)buf;
+  return tok;
+}
+
 static Token *read_char_literal(Token *cur, char *start, Type *ty) {
   char *p = start + 1;
   if (*p == '\0')
@@ -497,6 +519,13 @@ Token *tokenize(File *file) {
       continue;
     }
 
+    // UTF-32 string literal
+    if (startswith(p, "U\"")) {
+      cur = read_utf32_string_literal(cur, p + 1, ty_uint);
+      p += cur->len + 1;
+      continue;
+    }
+
     // Character literal
     if (*p == '\'') {
       cur = read_char_literal(cur, p, ty_int);
-- 
GitLab