parse.y: invalid name as mere string

* parse.y (parser_peek_variable_name): treat invalid global, class,
  and instance variable names as mere strings rather than errors.
  [ruby-core:54885] [Bug #8375]

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40635 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,9 @@
+Fri May 10 23:56:34 2013  Nobuyoshi Nakada  <nobu@ruby-lang.org>
+
+	* parse.y (parser_peek_variable_name): treat invalid global, class,
+	  and instance variable names as mere strings rather than errors.
+	  [ruby-core:54885] [Bug #8375]
+
 Fri May 10 20:22:40 2013  Tanaka Akira  <akr@fsij.org>
 
 	* configure.in: Move library checks into "Checks for libraries." part.

parse.y

@@ -6157,6 +6157,70 @@ ripper_flush_string_content(struct parser_params *parser, rb_encoding *enc)
 #define flush_string_content(enc) ((void)(enc))
 #endif
 
+RUBY_FUNC_EXPORTED const unsigned int ruby_global_name_punct_bits[(0x7e - 0x20 + 31) / 32];
+/* this can be shared with ripper, since it's independent from struct
+ * parser_params. */
+#ifndef RIPPER
+#define BIT(c, idx) (((c) / 32 - 1 == idx) ? (1U << ((c) % 32)) : 0)
+#define SPECIAL_PUNCT(idx) ( \
+	BIT('~', idx) | BIT('*', idx) | BIT('$', idx) | BIT('?', idx) | \
+	BIT('!', idx) | BIT('@', idx) | BIT('/', idx) | BIT('\\', idx) | \
+	BIT(';', idx) | BIT(',', idx) | BIT('.', idx) | BIT('=', idx) | \
+	BIT(':', idx) | BIT('<', idx) | BIT('>', idx) | BIT('\"', idx) | \
+	BIT('&', idx) | BIT('`', idx) | BIT('\'', idx) | BIT('+', idx) | \
+	BIT('0', idx))
+const unsigned int ruby_global_name_punct_bits[] = {
+    SPECIAL_PUNCT(0),
+    SPECIAL_PUNCT(1),
+    SPECIAL_PUNCT(2),
+};
+#undef BIT
+#undef SPECIAL_PUNCT
+#endif
+
+static inline int
+is_global_name_punct(const char c)
+{
+    if (c <= 0x20 || 0x7e < c) return 0;
+    return (ruby_global_name_punct_bits[(c - 0x20) / 32] >> (c % 32)) & 1;
+}
+
+static int
+parser_peek_variable_name(struct parser_params *parser)
+{
+    int c;
+    const char *p = lex_p;
+
+    if (p + 1 >= lex_pend) return 0;
+    c = *p++;
+    switch (c) {
+      case '$':
+	if ((c = *p) == '-') {
+	    if (++p >= lex_pend) return 0;
+	    c = *p;
+	}
+	else if (is_global_name_punct(c) || ISDIGIT(c)) {
+	    return tSTRING_DVAR;
+	}
+	break;
+      case '@':
+	if ((c = *p) == '@') {
+	    if (++p >= lex_pend) return 0;
+	    c = *p;
+	}
+	break;
+      case '{':
+	lex_p = p;
+	command_start = TRUE;
+	return tSTRING_DBEG;
+      default:
+	return 0;
+    }
+    if (!ISASCII(c) || c == '_' || ISALPHA(c))
+	return tSTRING_DVAR;
+    return 0;
+}
+
 static int
 parser_parse_string(struct parser_params *parser, NODE *quote)
 {
@@ -6187,16 +6251,10 @@ parser_parse_string(struct parser_params *parser, NODE *quote)
     }
     newtok();
     if ((func & STR_FUNC_EXPAND) && c == '#') {
-	switch (c = nextc()) {
+	int t = parser_peek_variable_name(parser);
-	  case '$':
+	if (t) return t;
-	  case '@':
-	    pushback(c);
-	    return tSTRING_DVAR;
-	  case '{':
-	    command_start = TRUE;
-	    return tSTRING_DBEG;
-	}
 	tokadd('#');
+	c = nextc();
     }
     pushback(c);
     if (tokadd_string(func, term, paren, &quote->nd_nest,
@@ -6403,16 +6461,10 @@ parser_here_document(struct parser_params *parser, NODE *here)
 	/*	int mb = ENC_CODERANGE_7BIT, *mbp = &mb;*/
 	newtok();
 	if (c == '#') {
-	    switch (c = nextc()) {
+	    int t = parser_peek_variable_name(parser);
-	      case '$':
+	    if (t) return t;
-	      case '@':
-		pushback(c);
-		return tSTRING_DVAR;
-	      case '{':
-		command_start = TRUE;
-		return tSTRING_DBEG;
-	    }
 	    tokadd('#');
+	    c = nextc();
 	}
 	do {
 	    pushback(c);
@@ -9998,22 +10050,17 @@ is_special_global_name(const char *m, const char *e, rb_encoding *enc)
     int mb = 0;
 
     if (m >= e) return 0;
-    switch (*m) {
+    if (is_global_name_punct(*m)) {
-      case '~': case '*': case '$': case '?': case '!': case '@':
-      case '/': case '\\': case ';': case ',': case '.': case '=':
-      case ':': case '<': case '>': case '\"':
-      case '&': case '`': case '\'': case '+':
-      case '0':
 	++m;
-	break;
+    }
-      case '-':
+    else if (*m == '-') {
 	if (++m >= e) return 0;
 	if (is_identchar(m, e, enc)) {
 	    if (!ISASCII(*m)) mb = 1;
 	    m += rb_enc_mbclen(m, e, enc);
 	}
-	break;
+    }
-      default:
+    else {
 	if (!rb_enc_isdigit(*m, enc)) return 0;
 	do {
 	    if (!ISASCII(*m)) mb = 1;

test/ruby/test_parse.rb

@@ -361,6 +361,17 @@ class TestParse < Test::Unit::TestCase
     assert_equal("foo 1 bar", "foo #$1 bar")
   end
 
+  def test_dstr_disallowd_variable
+    bug8375 = '[ruby-core:54885] [Bug #8375]'
+    %w[@ @1 @@. @@ @@1 @@. $ $%].each do |src|
+      src = '#'+src+' '
+      str = assert_nothing_raised(SyntaxError, "#{bug8375} #{src.dump}") do
+        break eval('"'+src+'"')
+      end
+      assert_equal(src, str, bug8375)
+    end
+  end
+
   def test_dsym
     assert_nothing_raised { eval(':""') }
   end
@@ -527,13 +538,14 @@ class TestParse < Test::Unit::TestCase
       )
     end
 
-    assert_raise(SyntaxError) do
+    assert_nothing_raised(SyntaxError) do
-      eval %q(
+      x = eval %q(
 <<FOO
 #$
 FOO
       )
     end
+    assert_equal "\#$\n", x
 
     assert_raise(SyntaxError) do
       eval %Q(
@@ -553,14 +565,15 @@ FOO
       )
     end
 
-    assert_raise(SyntaxError) do
+    assert_nothing_raised(SyntaxError) do
-      eval %q(
+      x = eval %q(
 <<FOO
 #$
 foo
 FOO
       )
     end
+    assert_equal "\#$\nfoo\n", x
 
     assert_nothing_raised do
       eval "x = <<""FOO\r\n1\r\nFOO"