merge revision(s) r42479,r42490,r42509,r43083,r43084,r43085: [Backport #8756] [Backport #9248]

	* parse.y (rb_enc_symname_type): allow ID_ATTRSET for ID_INSTANCE,
	  ID_GLOBAL, ID_CLASS, ID_JUNK too.  [Bug #8756]

	* parse.y (rb_id_attrset): fix inconsistency with literals, allow
	  ID_ATTRSET and return it itself, but ID_JUNK cannot make ID_ATTRSET.
	  and raise a NameError instead of rb_bug() for invalid argument.

	* parse.y (rb_id_attrset, intern_str): allow junk attrset ID for
	  Struct.

	* parse.y (rb_id_attrset): check if the argument is valid type as an
	  attribute.

	* parse.y (rb_id_attrset): allow other than ID_ATTRSET.

	* parse.y (intern_str): ditto.  try stem ID for ID_INSTANCE,
	  ID_GLOBAL, ID_CLASS, ID_JUNK too.  [Bug #8756]


git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_0_0@44911 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 76 insertions, 35 deletions

diff --git a/parse.y b/parse.y
index 9b5617d000..9f10260eae 100644
--- a/parse.y
+++ b/parse.y
@@ -8678,9 +8678,41 @@ block_dup_check_gen(struct parser_params *parser, NODE *node1, NODE *node2)
     }
 }
 
+static const char id_type_names[][9] = {
+    "LOCAL",
+    "INSTANCE",
+    "",				/* INSTANCE2 */
+    "GLOBAL",
+    "ATTRSET",
+    "CONST",
+    "CLASS",
+    "JUNK",
+};
+
 ID
 rb_id_attrset(ID id)
 {
+    if (!is_notop_id(id)) {
+	switch (id) {
+	  case tAREF: case tASET:
+	    return tASET;	/* only exception */
+	}
+	rb_name_error(id, "cannot make operator ID :%s attrset", rb_id2name(id));
+    }
+    else {
+	int scope = (int)(id & ID_SCOPE_MASK);
+	switch (scope) {
+	  case ID_LOCAL: case ID_INSTANCE: case ID_GLOBAL:
+	  case ID_CONST: case ID_CLASS: case ID_JUNK:
+	    break;
+	  case ID_ATTRSET:
+	    return id;
+	  default:
+	    rb_name_error(id, "cannot make %s ID %+"PRIsVALUE" attrset",
+			  id_type_names[scope], ID2SYM(id));
+
+	}
+    }
     id &= ~ID_SCOPE_MASK;
     id |= ID_ATTRSET;
     return id;
@@ -10062,8 +10094,11 @@ rb_enc_symname_p(const char *name, rb_encoding *enc)
     return rb_enc_symname2_p(name, strlen(name), enc);
 }
 
+#define IDSET_ATTRSET_FOR_SYNTAX ((1U<<ID_LOCAL)|(1U<<ID_CONST))
+#define IDSET_ATTRSET_FOR_INTERN (~(~0U<<ID_SCOPE_MASK) & ~(1U<<ID_ATTRSET))
+
 static int
-rb_enc_symname_type(const char *name, long len, rb_encoding *enc)
+rb_enc_symname_type(const char *name, long len, rb_encoding *enc, unsigned int allowed_atttset)
 {
     const char *m = name;
     const char *e = m + len;
@@ -10148,7 +10183,7 @@ rb_enc_symname_type(const char *name, long len, rb_encoding *enc)
 	    ++m;
 	    break;
 	  case '=':
-	    if (type != ID_CONST && type != ID_LOCAL) return -1;
+	    if (!(allowed_atttset & (1U << type))) return -1;
 	    type = ID_ATTRSET;
 	    ++m;
 	    break;
@@ -10161,15 +10196,15 @@ rb_enc_symname_type(const char *name, long len, rb_encoding *enc)
 int
 rb_enc_symname2_p(const char *name, long len, rb_encoding *enc)
 {
-    return rb_enc_symname_type(name, len, enc) != -1;
+    return rb_enc_symname_type(name, len, enc, IDSET_ATTRSET_FOR_SYNTAX) != -1;
 }
 
 static int
-rb_str_symname_type(VALUE name)
+rb_str_symname_type(VALUE name, unsigned int allowed_atttset)
 {
     const char *ptr = StringValuePtr(name);
     long len = RSTRING_LEN(name);
-    int type = rb_enc_symname_type(ptr, len, rb_enc_get(name));
+    int type = rb_enc_symname_type(ptr, len, rb_enc_get(name), allowed_atttset);
     RB_GC_GUARD(name);
     return type;
 }
@@ -10294,26 +10329,27 @@ intern_str(VALUE str)
 		}
 	    }
 	}
-
-	if (m[last] == '=') {
-	    /* attribute assignment */
-	    if (!rb_enc_symname2_p(name, last, enc))
-		goto junk;
-	    id = rb_intern3(name, last, enc);
-	    if (id > tLAST_OP_ID && !is_attrset_id(id)) {
-		enc = rb_enc_get(rb_id2str(id));
-		id = rb_id_attrset(id);
-		goto id_register;
-	    }
-	    id = ID_ATTRSET;
+	break;
+    }
+    if (name[last] == '=') {
+	/* attribute assignment */
+	if (last > 1 && name[last-1] == '=')
+	    goto junk;
+	id = rb_intern3(name, last, enc);
+	if (id > tLAST_OP_ID && !is_attrset_id(id)) {
+	    enc = rb_enc_get(rb_id2str(id));
+	    id = rb_id_attrset(id);
+	    goto id_register;
 	}
-	else if (rb_enc_isupper(m[0], enc)) {
+	id = ID_ATTRSET;
+    }
+    else if (id == 0) {
+	if (rb_enc_isupper(m[0], enc)) {
 	    id = ID_CONST;
-        }
+	}
 	else {
 	    id = ID_LOCAL;
 	}
-	break;
     }
     if (!rb_enc_isdigit(*m, enc)) {
 	while (m <= name + last && is_identchar(m, e, enc)) {
@@ -10325,7 +10361,7 @@ intern_str(VALUE str)
 	    }
 	}
     }
-    if (m - name < len) id = ID_JUNK;
+    if (id != ID_ATTRSET && m - name < len) id = ID_JUNK;
     if (sym_check_asciionly(str)) symenc = rb_usascii_encoding();
   new_id:
     if (symenc != enc) rb_enc_associate(str, symenc);
@@ -10408,16 +10444,21 @@ rb_id2str(ID id)
     }
 
     if (is_attrset_id(id)) {
-	ID id2 = (id & ~ID_SCOPE_MASK) | ID_LOCAL;
+	ID id_stem = (id & ~ID_SCOPE_MASK);
 	VALUE str;
 
-	while (!(str = rb_id2str(id2))) {
-	    if (!is_local_id(id2)) return 0;
-	    id2 = (id & ~ID_SCOPE_MASK) | ID_CONST;
-	}
+	do {
+	    if (!!(str = rb_id2str(id_stem | ID_LOCAL))) break;
+	    if (!!(str = rb_id2str(id_stem | ID_CONST))) break;
+	    if (!!(str = rb_id2str(id_stem | ID_INSTANCE))) break;
+	    if (!!(str = rb_id2str(id_stem | ID_GLOBAL))) break;
+	    if (!!(str = rb_id2str(id_stem | ID_CLASS))) break;
+	    if (!!(str = rb_id2str(id_stem | ID_JUNK))) break;
+	    return 0;
+	} while (0);
 	str = rb_str_dup(str);
 	rb_str_cat(str, "=", 1);
-	rb_intern_str(str);
+	register_symid_str(id, str);
 	if (st_lookup(global_symbols.id_str, id, &data)) {
             VALUE str = (VALUE)data;
             if (RBASIC(str)->klass == 0)
@@ -10600,43 +10641,43 @@ rb_check_id_cstr(const char *ptr, long len, rb_encoding *enc)
 int
 rb_is_const_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_CONST;
+    return rb_str_symname_type(name, 0) == ID_CONST;
 }
 
 int
 rb_is_class_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_CLASS;
+    return rb_str_symname_type(name, 0) == ID_CLASS;
 }
 
 int
 rb_is_global_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_GLOBAL;
+    return rb_str_symname_type(name, 0) == ID_GLOBAL;
 }
 
 int
 rb_is_instance_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_INSTANCE;
+    return rb_str_symname_type(name, 0) == ID_INSTANCE;
 }
 
 int
 rb_is_attrset_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_ATTRSET;
+    return rb_str_symname_type(name, IDSET_ATTRSET_FOR_INTERN) == ID_ATTRSET;
 }
 
 int
 rb_is_local_name(VALUE name)
 {
-    return rb_str_symname_type(name) == ID_LOCAL;
+    return rb_str_symname_type(name, 0) == ID_LOCAL;
 }
 
 int
 rb_is_method_name(VALUE name)
 {
-    switch (rb_str_symname_type(name)) {
+    switch (rb_str_symname_type(name, 0)) {
       case ID_LOCAL: case ID_ATTRSET: case ID_JUNK:
 	return TRUE;
     }
@@ -10646,7 +10687,7 @@ rb_is_method_name(VALUE name)
 int
 rb_is_junk_name(VALUE name)
 {
-    return rb_str_symname_type(name) == -1;
+    return rb_str_symname_type(name, IDSET_ATTRSET_FOR_SYNTAX) == -1;
 }
 
 #endif /* !RIPPER */