[ruby-dev:33944] Re: [ruby-cvs:22878] Ruby:r15641 (trunk): * string.c (rb_str_chomp_bang): now works on UTF-16.

なかだです。

At Fri, 29 Feb 2008 22:33:28 +0900 (JST),
[e-mail] wrote in [ruby-cvs:22878]:
>   Log:
>     * string.c (rb_str_chomp_bang): now works on UTF-16.
>     
>     * string.c (tr_setup_table): negation should work on non ASCII
>       compatible strings as well.
>     
>     * string.c (rb_str_split_m): awk split should work on non ASCII
>       compatible strings as well.

気づいてない人が多いような気がするので書いておきますが、こういう
用途のためにrb_enc_ascget()というのがあるようです。


Index: string.c
===================================================================
--- string.c	(revision 15645)
+++ string.c	(working copy)
@@ -4447,5 +4447,5 @@ tr_setup_table(VALUE str, char stable[25
     char buf[256];
     struct tr tr;
-    int c;
+    int c, l;
     VALUE table = 0, ptable = 0;
     int i, cflag = 0;
@@ -4454,18 +4454,7 @@ tr_setup_table(VALUE str, char stable[25
     tr.gen = tr.now = tr.max = 0;
     
-    if (RSTRING_LEN(str) > 1) {
-	if (rb_enc_asciicompat(enc)) {
-	    if (RSTRING_PTR(str)[0] == '^') {
-		cflag = 1;
-		tr.p++;
-	    }
-	}
-	else {
-	    c = rb_enc_codepoint(RSTRING_PTR(str), RSTRING_END(str), enc);
-	    if (c == '^') {
-		cflag = 1;
-		tr.p+=rb_enc_codelen(c, enc);
-	    }
-	}
+    if (RSTRING_LEN(str) > 1 && rb_enc_ascget(tr.p, tr.pend, &l, enc) == '^') {
+	cflag = 1;
+	tr.p += l;
     }
     if (first) {
@@ -4859,6 +4848,7 @@ rb_str_split_m(int argc, VALUE *argv, VA
 	    }
 	    else {
-		if (str_strlen(spat, enc2) == 1 &&
-		    rb_enc_codepoint(RSTRING_PTR(spat), RSTRING_END(spat), enc2) == ' ') {
+		int l;
+		if (rb_enc_ascget(RSTRING_PTR(spat), RSTRING_END(spat), &l, enc2) == ' ' &&
+		    RSTRING_LEN(spat) == l) {
 		    awk_split = Qtrue;
 		}
@@ -5299,14 +5289,13 @@ rb_str_chomp_bang(int argc, VALUE *argv,
 	    enc = rb_enc_get(str);
 	    if (rb_enc_mbminlen(enc) > 1) {
-		len = str_strlen(str, enc);
-		pp = rb_enc_nth(p, e, len-1, enc);
+		pp = rb_enc_left_char_head(p, e-rb_enc_mbminlen(enc), enc);
 		if (rb_enc_is_newline(pp, e, enc)) {
 		    e = pp;
-		    len--;
 		}
-		if (len > 0) {
-		    p = rb_enc_nth(p, e, len-1, enc);
-		    if (rb_enc_codepoint(p, e, enc) == '\r') {
-			pp = e = p;
+		pp = e - rb_enc_mbminlen(enc);
+		if (pp >= p) {
+		    pp = rb_enc_left_char_head(p, pp, enc);
+		    if (rb_enc_ascget(pp, e, 0, enc) == '\r') {
+			e = pp;
 		    }
 		}
@@ -5314,5 +5303,5 @@ rb_str_chomp_bang(int argc, VALUE *argv,
 		    return Qnil;
 		}
-		len = pp - RSTRING_PTR(str);
+		len = e - RSTRING_PTR(str);
 		STR_SET_LEN(str, len);
 	    }


-- 
--- 僕の前にBugはない。
--- 僕の後ろにBugはできる。
    中田 伸悦