From b161530cf2edf8ca1bf5ddc0623b607198e0bb27 Mon Sep 17 00:00:00 2001
From: akr <akr@b2dd03c8-39d4-4d8f-98ff-823fe69b080e>
Date: Fri, 15 Feb 2008 18:48:27 +0000
Subject: * encoding.c (rb_enc_nth): moved to string.c.

* string.c (rb_enc_nth): moved from string.c.  use search_nonascii
  for ASCII compatible string.
  (str_nth): wrong optimization removed to fix
  "a".force_encoding("EUC-JP").slice!(0,10) returns
  "a\x00\x00\x00\x00\x00\x00\x00\x00\x00"


git-svn-id: http://svn.ruby-lang.org/repos/ruby/trunk@15492 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
---
 string.c | 53 +++++++++++++++++++++++++++++++++++++++++++++--------
 1 file changed, 45 insertions(+), 8 deletions(-)

(limited to 'string.c')

diff --git a/string.c b/string.c
index 1d94ca29f..345b59b8f 100644
--- a/string.c
+++ b/string.c
@@ -891,20 +891,57 @@ rb_str_s_try_convert(VALUE dummy, VALUE str)
     return rb_check_string_type(str);
 }
 
+char*
+rb_enc_nth(const char *p, const char *e, int nth, rb_encoding *enc)
+{
+    int c;
+
+    if (rb_enc_mbmaxlen(enc) == 1) {
+        p += nth;
+    }
+    else if (rb_enc_mbmaxlen(enc) == rb_enc_mbminlen(enc)) {
+        p += nth * rb_enc_mbmaxlen(enc);
+    }
+    else if (rb_enc_asciicompat(enc)) {
+        const char *p2, *e2;
+        int n;
+
+        while (p < e && 0 < nth) {
+            e2 = p + nth;
+            if (e < e2)
+                return (char *)e;
+            if (ISASCII(*p)) {
+                p2 = search_nonascii(p, e2);
+                if (!p2)
+                    return (char *)e2;
+                nth -= p2 - p;
+                p = p2;
+            }
+            n = rb_enc_mbclen(p, e, enc);
+            p += n;
+            nth--;
+        }
+        if (nth != 0)
+            return (char *)e;
+        return (char *)p;
+    }
+    else {
+        for (c=0; p<e && nth--; c++) {
+            int n = rb_enc_mbclen(p, e, enc);
+
+            p += n;
+        }
+    }
+    if (p > e) p = e;
+    return (char*)p;
+}
+
 static char*
 str_nth(const char *p, const char *e, int nth, rb_encoding *enc, int singlebyte)
 {
     if (singlebyte)
 	p += nth;
     else {
-	if (rb_enc_asciicompat(enc)) {
-	    const char *p2 = search_nonascii(p, e);
-
-	    if (!p2 || p + nth < p2)
-		return (char*)p + nth;
-	    nth -= p2 - p;
-	    p = p2;
-	}
 	p = rb_enc_nth(p, e, nth, enc);
     }
     if (!p) return 0;
-- 
cgit