From: "NARUSE, Yui" <naruse@...>
Date: 2010-01-11T15:53:55+09:00
Subject: [ruby-dev:40029] Re: [Feature #2571]  	文字列のハミング距離

成瀬です。

(2010/01/11 15:35), Kenta Murata wrote:
> ところで、拡張ライブラリにする際に、str_strlen を公開 API にしてもらいたいのですが、
> それは可能でしょうか？ rb_str_strlen になるのかな？

名前がちょっと悩む所なんですがとりあえず、

diff --git a/include/ruby/intern.h b/include/ruby/intern.h
index 14e23b8..5500a9e 100644
--- a/include/ruby/intern.h
+++ b/include/ruby/intern.h
@@ -662,6 +662,7 @@ VALUE rb_str_associated(VALUE);
 void rb_str_setter(VALUE, ID, VALUE*);
 VALUE rb_str_intern(VALUE);
 VALUE rb_sym_to_s(VALUE);
+log rb_str_strlen(VALUE, rb_encoding*);
 VALUE rb_str_length(VALUE);
 long rb_str_offset(VALUE, long);
 size_t rb_str_capacity(VALUE);
diff --git a/string.c b/string.c
index 2dcc330..6a7d625 100644
--- a/string.c
+++ b/string.c
@@ -1033,8 +1033,8 @@ count_utf8_lead_bytes_with_word(const VALUE *s)
 }
 #endif
 
-static long
-str_strlen(VALUE str, rb_encoding *enc)
+long
+rb_str_strlen(VALUE str, rb_encoding *enc)
 {
     const char *p, *e;
     long n;
@@ -1092,7 +1092,7 @@ rb_str_length(VALUE str)
 {
     long len;
 
-    len = str_strlen(str, STR_ENC_GET(str));
+    len = rb_str_strlen(str, STR_ENC_GET(str));
     return LONG2NUM(len);
 }
 
@@ -1573,11 +1573,11 @@ rb_str_substr(VALUE str, long beg, long len)
 	    goto sub;
 	}
 	else {
-	    beg += str_strlen(str, enc);
+	    beg += rb_str_strlen(str, enc);
 	    if (beg < 0) return Qnil;
 	}
     }
-    else if (beg > 0 && beg > str_strlen(str, enc)) {
+    else if (beg > 0 && beg > rb_str_strlen(str, enc)) {
 	return Qnil;
     }
     if (len == 0) {
@@ -2266,8 +2266,8 @@ rb_str_index(VALUE str, VALUE sub, long offset)
     if (is_broken_string(sub)) {
 	return -1;
     }
-    len = str_strlen(str, enc);
-    slen = str_strlen(sub, enc);
+    len = rb_str_strlen(str, enc);
+    slen = rb_str_strlen(sub, enc);
     if (offset < 0) {
 	offset += len;
 	if (offset < 0) return -1;
@@ -2329,7 +2329,7 @@ rb_str_index_m(int argc, VALUE *argv, VALUE str)
 	pos = 0;
     }
     if (pos < 0) {
-	pos += str_strlen(str, STR_ENC_GET(str));
+	pos += rb_str_strlen(str, STR_ENC_GET(str));
 	if (pos < 0) {
 	    if (TYPE(sub) == T_REGEXP) {
 		rb_backref_set(Qnil);
@@ -2340,7 +2340,7 @@ rb_str_index_m(int argc, VALUE *argv, VALUE str)
 
     switch (TYPE(sub)) {
       case T_REGEXP:
-	if (pos > str_strlen(str, STR_ENC_GET(str)))
+	if (pos > rb_str_strlen(str, STR_ENC_GET(str)))
 	    return Qnil;
 	pos = str_offset(RSTRING_PTR(str), RSTRING_END(str), pos,
 			 rb_enc_check(str, sub), single_byte_optimizable(str));
@@ -2382,8 +2382,8 @@ rb_str_rindex(VALUE str, VALUE sub, long pos)
     if (is_broken_string(sub)) {
 	return -1;
     }
-    len = str_strlen(str, enc);
-    slen = str_strlen(sub, enc);
+    len = rb_str_strlen(str, enc);
+    slen = rb_str_strlen(sub, enc);
     /* substring longer than string */
     if (len < slen) return -1;
     if (len - pos < slen) {
@@ -2433,7 +2433,7 @@ rb_str_rindex_m(int argc, VALUE *argv, VALUE str)
     VALUE sub;
     VALUE vpos;
     rb_encoding *enc = STR_ENC_GET(str);
-    long pos, len = str_strlen(str, enc);
+    long pos, len = rb_str_strlen(str, enc);
 
     if (rb_scan_args(argc, argv, "11", &sub, &vpos) == 2) {
 	pos = NUM2LONG(vpos);
@@ -2981,7 +2981,7 @@ rb_str_aref(VALUE str, VALUE indx)
 	    long beg, len;
 	    VALUE tmp;
 
-	    len = str_strlen(str, STR_ENC_GET(str));
+	    len = rb_str_strlen(str, STR_ENC_GET(str));
 	    switch (rb_range_beg_len(indx, &beg, &len, len, 0)) {
 	      case Qfalse:
 		break;
@@ -3136,7 +3136,7 @@ rb_str_splice(VALUE str, long beg, long len, VALUE val)
 
     StringValue(val);
     enc = rb_enc_check(str, val);
-    slen = str_strlen(str, enc);
+    slen = rb_str_strlen(str, enc);
 
     if (slen < beg) {
       out_of_range:
@@ -3232,14 +3232,14 @@ rb_str_aset(VALUE str, VALUE indx, VALUE val)
 	    rb_raise(rb_eIndexError, "string not matched");
 	}
 	beg = rb_str_sublen(str, beg);
-	rb_str_splice(str, beg, str_strlen(indx, 0), val);
+	rb_str_splice(str, beg, rb_str_strlen(indx, 0), val);
 	return val;
 
       default:
 	/* check if indx is Range */
 	{
 	    long beg, len;
-	    if (rb_range_beg_len(indx, &beg, &len, str_strlen(str, 0), 2)) {
+	    if (rb_range_beg_len(indx, &beg, &len, rb_str_strlen(str, 0), 2)) {
 		rb_str_splice(str, beg, len, val);
 		return val;
 	    }
@@ -6605,13 +6605,13 @@ rb_str_justify(int argc, VALUE *argv, VALUE str, char jflag)
 	enc = rb_enc_check(str, pad);
 	f = RSTRING_PTR(pad);
 	flen = RSTRING_LEN(pad);
-	fclen = str_strlen(pad, enc);
+	fclen = rb_str_strlen(pad, enc);
 	singlebyte = single_byte_optimizable(pad);
 	if (flen == 0 || fclen == 0) {
 	    rb_raise(rb_eArgError, "zero width padding");
 	}
     }
-    len = str_strlen(str, enc);
+    len = rb_str_strlen(str, enc);
     if (width < 0 || len >= width) return rb_str_dup(str);
     n = width - len;
     llen = (jflag == 'l') ? 0 : ((jflag == 'r') ? n : n/2);
@@ -6829,7 +6829,7 @@ rb_str_rpartition(VALUE str, VALUE sep)
     }
     return rb_ary_new3(3, rb_str_substr(str, 0, pos),
 		          sep,
-		          rb_str_substr(str,pos+str_strlen(sep,STR_ENC_GET(sep)),RSTRING_LEN(str)));
+		          rb_str_substr(str,pos+rb_str_strlen(sep,STR_ENC_GET(sep)),RSTRING_LEN(str)));
 }
 
 /*

-- 
NARUSE, Yui  <naruse@airemix.jp>