2000-03-08

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@635 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,11 @@
+Wed Mar  8 02:08:43 2000  Yukihiro Matsumoto  <matz@netlab.co.jp>
+
+	* parse.y: escape expansion too early.
+
+	* string.c (rb_f_scan): Kernel#scan added.
+
+	* regex.c (re_compile_pattern): support \cX et al.
+
 Tue Mar  7 01:44:27 2000  Yukihiro Matsumoto  <matz@netlab.co.jp>
 
 	* io.c (set_stdin): simplified procedure, allows $stdin = DATA;
@@ -6,7 +14,7 @@ Tue Mar  7 01:44:27 2000  Yukihiro Matsumoto  <matz@netlab.co.jp>
 	* io.c (set_outfile): ditto.
 
 	* re.c (Init_Regexp): new method Regexp#last_match added; it's a
-	  replacement for $~.
+	  alternative for $~.
 
 	* configure.in (DEFAULT_KCODE): KCODE_NONE should be the default.
 

ToDo

@@ -4,9 +4,9 @@ Language Spec.
 - compile time string concatenation, "hello" "world" => "helloworld"
 - rescue modifier; a rescue b => begin a rescue; b end
 - %w(a\ b\ c abc) => ["a b c", "abc"]
-* class variable (prefix @@?) ??
+- objectify symbols
+- class variable (prefix @@) - still need work for singletons
 * operator !! for rescue. ???
-* objectify symbols
 * objectify characters
 * ../... outside condition invokes operator method too.
 * ... inside condition turns off just before right condition.???
@@ -58,6 +58,8 @@ Standard Libraries
 - regexp: (?>..), \G
 - Struct::new([name,]member,...)
 - IO#reopen accepts path as well
+- Kernel#scan
+- call initialize for builtin class too (not yet: Regexp, IO, etc)
 * String#scanf(?)
 * Object#fmt(?)
 * Integer#{bin,oct,hex,heX}
@@ -69,7 +71,6 @@ Standard Libraries
 * optional stepsize argument for succ()
 * performance tune for String's non-bang methods.
 * Ruby module -- Ruby::Version, Ruby::Interpreter
-* call initialize for builtin class too
 
 Extension Libraries
 

eval.c

@@ -340,7 +340,7 @@ rb_disable_super(klass, name)
     }
     else {
 	rb_clear_cache_by_id(mid);
-	rb_add_method(ruby_class, mid, 0, NOEX_UNDEF);
+	rb_add_method(klass, mid, 0, NOEX_UNDEF);
     }
 }
 
@@ -4302,7 +4302,7 @@ rb_call_super(argc, argv)
     VALUE result;
 
     if (ruby_frame->last_class == 0) {	
-	rb_raise(rb_eNameError, "superclass method `%s' disabled",
+	rb_raise(rb_eNameError, "superclass method `%s' must be enabled by rb_enable_super()",
 		 rb_id2name(ruby_frame->last_func));
     }
 

object.c

@@ -497,7 +497,13 @@ static VALUE
 sym_to_s(sym)
     VALUE sym;
 {
-    return rb_str_new2(rb_id2name(SYM2ID(sym)));
+    char *name, *buf;
+
+    name = rb_id2name(SYM2ID(sym));
+    buf = ALLOCA_N(char, strlen(name)+2);
+    sprintf(buf, ":%s", name);
+
+    return rb_str_new2(buf);
 }
 
 static VALUE
@@ -1100,7 +1106,7 @@ Init_Object()
     rb_define_global_const("NIL", Qnil);
 
     rb_cSymbol = rb_define_class("Symbol", rb_cObject);
-    rb_undef_method(CLASS_OF(rb_cNilClass), "new");
+    rb_undef_method(CLASS_OF(rb_cSymbol), "new");
     rb_define_method(rb_cSymbol, "type", sym_type, 0);
     rb_define_method(rb_cSymbol, "to_i", sym_to_i, 0);
     rb_define_method(rb_cSymbol, "to_s", sym_to_s, 0);

parse.y

@@ -2067,20 +2067,10 @@ read_escape()
 
       case 'x':	/* hex constant */
 	{
-	    char buf[2];
+	    int numlen;
-	    int i;
 
-	    for (i=0; i<2; i++) {
+	    c = scan_hex(lex_p, 2, &numlen);
-		int cc = nextc();
+	    lex_p += numlen;
-
-		if (cc == -1) goto eof;
-		buf[i] = cc;
-		if (!ISXDIGIT(buf[i])) {
-		    pushback(buf[i]);
-		    break;
-		}
-	    }
-	    c = scan_hex(buf, i, &i);
 	}
 	return c;
 
@@ -3595,9 +3585,16 @@ str_extend(list, term)
 		tokadd(c);
 		goto loop_again;
 	      case '\\':
-		c = read_escape();
+		c = nextc();
-		tokadd(c);
+		if (c == -1) return (NODE*)-1;
-		goto loop_again;
+		if (c == term) {
+		    tokadd(c);
+		}
+		else {
+		    tokadd('\\');
+		    tokadd(c);
+		}
+		break;
 	      case '{':
 		if (brace != -1) nest++;
 	      case '\"':

regex.c

@@ -1067,7 +1067,7 @@ calculate_must_string(start, end)
   return must;
 }
 
-static int
+static unsigned int
 read_backslash(c)
      int c;
 {
@@ -1099,6 +1099,47 @@ read_backslash(c)
   return c;
 }
 
+static unsigned int
+read_special(p, pend, pp)
+     const char *p, *pend, **pp;
+{
+  int c;
+
+  PATFETCH_RAW(c);
+  switch (c) {
+  case 'M':
+    PATFETCH_RAW(c);
+    if (c != '-') return -1;
+    PATFETCH_RAW(c);
+    *pp = p;
+    if (c == '\\') {
+      return read_special(p, pend, pp) | 0x80;
+    }
+    else if (c == -1) return ~0;
+    else {
+      return ((c & 0xff) | 0x80);
+    }
+
+  case 'C':
+    PATFETCH_RAW(c);
+    if (c != '-') return ~0;
+  case 'c':
+    PATFETCH_RAW(c);
+    *pp = p;
+    if (c == '\\') {
+      c = read_special(p, pend, pp);
+    }
+    else if (c == '?') return 0177;
+    else if (c == -1) return ~0;
+    return c & 0x9f;
+  default:
+    return read_backslash(c);
+  }
+
+ end_of_pattern:
+  return ~0;
+}
+
 /* re_compile_pattern takes a regular-expression string
    and converts it into a buffer full of byte commands for matching.
 
@@ -1470,6 +1511,16 @@ re_compile_pattern(pattern, size, bufp)
 	    had_num_literal = 1;
 	    break;
 
+	  case 'M':
+	  case 'C':
+	  case 'c':
+	    p0 = --p;
+	    c = read_special(p, pend, &p0);
+	    if (c > 255) goto invalid_escape;
+	    p = p0;
+	    had_num_literal = 1;
+	    break;
+
 	  default:
 	    c = read_backslash(c);
 	    if (ismbchar(c)) {
@@ -2173,6 +2224,16 @@ re_compile_pattern(pattern, size, bufp)
 	BUFPUSH(c1);
 	break;
 
+      case 'M':
+      case 'C':
+      case 'c':
+	p0 = --p;
+	c = read_special(p, pend, &p0);
+	if (c > 255) goto invalid_escape;
+	p = p0;
+	had_num_literal = 1;
+	goto numeric_char;
+
       default:
 	c = read_backslash(c);
 	goto normal_char;
@@ -2335,6 +2396,9 @@ re_compile_pattern(pattern, size, bufp)
 
  nested_meta:
   FREE_AND_RETURN(stackb, "nested *?+ in regexp");
+
+ invalid_escape:
+  FREE_AND_RETURN(stackb, "Invalid escape character syntax");
 }
 
 void

string.c

@@ -2483,6 +2483,13 @@ rb_str_scan(str, pat)
     return str;
 }
 
+static VALUE
+rb_f_scan(self, pat)
+    VALUE self, pat;
+{
+    return rb_str_scan(uscore_get(), pat);
+}
+
 static VALUE
 rb_str_hex(str)
     VALUE str;
@@ -2755,6 +2762,7 @@ Init_String()
     rb_define_global_function("chomp!", rb_f_chomp_bang, -1);
 
     rb_define_global_function("split", rb_f_split, -1);
+    rb_define_global_function("scan", rb_f_scan, 1);
 
     rb_define_method(rb_cString, "slice", rb_str_aref_m, -1);
     rb_define_method(rb_cString, "slice!", rb_str_slice_bang, -1);

version.h

@@ -1,4 +1,4 @@
 #define RUBY_VERSION "1.5.3"
-#define RUBY_RELEASE_DATE "2000-03-07"
+#define RUBY_RELEASE_DATE "2000-03-08"
 #define RUBY_VERSION_CODE 153
-#define RUBY_RELEASE_CODE 20000307
+#define RUBY_RELEASE_CODE 20000308