From 0fe255a7c1ea7f8c324bfa58cdfb2c77f06bd823 Mon Sep 17 00:00:00 2001
From: Bruce Hill <bruce@bruce-hill.com>
Date: Sun, 3 Mar 2024 17:49:40 -0500
Subject: Fix up unicode escapes

---
 compile.c       |  2 +-
 parse.c         | 14 ++++++++++++--
 test/strings.tm |  3 +++
 3 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/compile.c b/compile.c
index 20e16ad1..3ef93325 100644
--- a/compile.c
+++ b/compile.c
@@ -452,7 +452,7 @@ CORD compile(env_t *env, ast_t *ast)
                 if (isprint(c))
                     code = CORD_cat_char(code, c);
                 else
-                    CORD_sprintf(&code, "%r\\x%02X", code, c);
+                    CORD_sprintf(&code, "%r\\x%02X", code, (uint8_t)c);
                 break;
             }
             }
diff --git a/parse.c b/parse.c
index 19a3a288..b67ea418 100644
--- a/parse.c
+++ b/parse.c
@@ -140,6 +140,15 @@ const char *unescape(const char **out) {
     if (unescapes[(int)escape[1]]) {
         *endpos = escape + 2;
         return heap_str(unescapes[(int)escape[1]]);
+    } else if (escape[1] == 'U' && escape[2]) {
+        char *endptr = NULL;
+        long codepoint = strtol(escape+2, &endptr, 16);
+        uint32_t ustr[2] = {codepoint, 0};
+        size_t bufsize = 8;
+        uint8_t buf[bufsize];
+        (void)u32_to_u8(ustr, bufsize, buf, &bufsize);
+        *endpos = endptr;
+        return heap_strn((char*)buf, bufsize);
     } else if (escape[1] == 'x' && escape[2] && escape[3]) {
         char *endptr = (char*)&escape[3+1];
         char c = (char)strtol(escape+2, &endptr, 16);
@@ -940,8 +949,9 @@ PARSER(parse_string) {
     if (*pos == '\\') {
         CORD cord = CORD_EMPTY;
         do {
-            char c = unescape(&pos)[0];
-            cord = CORD_cat_char(cord, c);
+            const char *c = unescape(&pos);
+            cord = CORD_cat(cord, c);
+            // cord = CORD_cat_char(cord, c);
         } while (*pos == '\\');
         return NewAST(ctx->file, start, pos, StringLiteral, .cord=cord);
     }
diff --git a/test/strings.tm b/test/strings.tm
index 6ef139ac..a7631824 100644
--- a/test/strings.tm
+++ b/test/strings.tm
@@ -5,3 +5,6 @@
 = "hello amélie!"
 >> str:lower():title()
 = "Hello Amélie!"
+
+>> \UE9
+= "é"
-- 
cgit v1.2.3