From 20ce4e60957b48b7268c8ed814536904c5252392 Mon Sep 17 00:00:00 2001
From: Sami Vaarala <sami.vaarala@iki.fi>
Date: Fri, 1 Feb 2013 00:20:12 +0200
Subject: [PATCH] move a lonely function to unicode support file

---
 src/duk_unicode.h         |  1 +
 src/duk_unicode_support.c | 23 +++++++++++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/src/duk_unicode.h b/src/duk_unicode.h
index d22954f8..de4c3029 100644
--- a/src/duk_unicode.h
+++ b/src/duk_unicode.h
@@ -48,6 +48,7 @@ int duk_unicode_get_xutf8_length(duk_u32 x);
 size_t duk_unicode_encode_xutf8(duk_u32 x, duk_u8 *out);
 size_t duk_unicode_encode_cesu8(duk_u32 x, duk_u8 *out);
 duk_u32 duk_unicode_xutf8_get_u32(duk_hthread *thr, duk_u8 **ptr, duk_u8 *ptr_start, duk_u8 *ptr_end);
+duk_u32 duk_unicode_unvalidated_utf8_length(duk_u8 *data, duk_u32 blen);
 int duk_unicode_is_whitespace(int x);
 int duk_unicode_is_line_terminator(int x);
 int duk_unicode_is_identifier_start(int x);
diff --git a/src/duk_unicode_support.c b/src/duk_unicode_support.c
index a448207f..d4b1628c 100644
--- a/src/duk_unicode_support.c
+++ b/src/duk_unicode_support.c
@@ -215,6 +215,29 @@ duk_u32 duk_unicode_xutf8_get_u32(duk_hthread *thr, duk_u8 **ptr, duk_u8 *ptr_st
 	return 0;  /* never here */
 }
 
+/* (extended) utf-8 length without codepoint encoding validation, used
+ * for string interning (should probably be inlined).
+ */
+duk_u32 duk_unicode_unvalidated_utf8_length(duk_u8 *data, duk_u32 blen) {
+        duk_u8 *p = data;
+        duk_u8 *p_end = data + blen;
+        duk_u32 clen = 0;
+
+        while (p < p_end) {
+                duk_u8 x = *p++;
+                if (x < 0x80) {
+                        clen++;
+                } else if (x >= 0xc0 ) {
+                        /* 10xxxxxx = continuation chars (0x80...0xbf), above that
+                         * initial bytes.
+                         */
+                        clen++;
+                }
+        }
+
+        return clen;
+}
+
 /*
  *  Unicode range matcher
  *