Fix a bug that inconsistency of IndexError vs nil for unknown capture group

Fix GH-139

Author: naitoh

ext/jruby/org/jruby/ext/strscan/RubyStringScanner.java

@@ -751,12 +751,14 @@ public IRubyObject op_aref(ThreadContext context, IRubyObject idx) {
  return context.nil;
  }
 
+ Ruby runtime = context.runtime;
+
  if (idx instanceof RubySymbol || idx instanceof RubyString) {
- if (pattern == null) return context.nil;
+ if (pattern == null) {
+ throw runtime.newRaiseException((RubyClass) getMetaClass().getConstant("IndexError"), "undefined group name reference: " + idx);
+ }
  }
 
- Ruby runtime = context.runtime;
-
  int i = RubyMatchData.backrefNumber(runtime, pattern, regs, idx);
 
  return extractRegion(context, i);

ext/strscan/strscan.c

@@ -1669,17 +1669,15 @@ name_to_backref_number(struct re_registers *regs, VALUE regexp, const char* name
 {
  int num;
 
- num = onig_name_to_backref_number(RREGEXP_PTR(regexp),
-(const unsigned char* )name, (const unsigned char* )name_end, regs);
- if (num >= 1) {
-return num;
- }
- else {
-rb_enc_raise(enc, rb_eIndexError, "undefined group name reference: %.*s",
- rb_long2int(name_end - name), name);
+ if (RTEST(regexp)){
+ num = onig_name_to_backref_number(RREGEXP_PTR(regexp),
+ (const unsigned char* )name, (const unsigned char* )name_end, regs);
+ if (num >= 1) {
+ return num;
+ }
  }
-
- UNREACHABLE;
+ rb_enc_raise(enc, rb_eIndexError, "undefined group name reference: %.*s",
+  rb_long2int(name_end - name), name);
 }
 
 /*
@@ -1768,7 +1766,6 @@ strscan_aref(VALUE self, VALUE idx)
  idx = rb_sym2str(idx);
  /* fall through */
  case T_STRING:
- if (!RTEST(p->regex)) return Qnil;
  RSTRING_GETMEM(idx, name, i);
  i = name_to_backref_number(&(p->regs), p->regex, name, name + i, rb_enc_get(idx));
  break;

test/strscan/test_stringscanner.rb

@@ -458,7 +458,31 @@ def test_matched_string
  def test_AREF
  s = create_string_scanner('stra strb strc')
 
- s.scan(/\w+/)
+ s.scan(/\s+/)
+ assert_nil( s[-2])
+ assert_nil( s[-1])
+ assert_nil( s[0])
+ assert_nil( s[1])
+ assert_nil( s[:c])
+ assert_nil( s['c'])
+
+ s.scan("not match")
+ assert_nil( s[-2])
+ assert_nil( s[-1])
+ assert_nil( s[0])
+ assert_nil( s[1])
+ assert_nil( s[:c])
+ assert_nil( s['c'])
+
+ s.check(/\w+/)
+ assert_nil( s[-2])
+ assert_equal('stra', s[-1])
+ assert_equal('stra', s[0])
+ assert_nil( s[1])
+ assert_raise(IndexError) { s[:c] }
+ assert_raise(IndexError) { s['c'] }
+
+ s.scan("stra")
  assert_nil( s[-2])
  assert_equal('stra', s[-1])
  assert_equal('stra', s[0])
@@ -903,11 +927,11 @@ def test_aref_without_regex
 
  s = create_string_scanner('abc')
  s.get_byte
- assert_nil(s[:c])
- assert_nil(s["c"])
+ assert_raise(IndexError) { s[:c] }
+ assert_raise(IndexError) { s['c'] }
  s.getch
- assert_nil(s[:c])
- assert_nil(s["c"])
+ assert_raise(IndexError) { s[:c] }
+ assert_raise(IndexError) { s['c'] }
  end
 
  def test_size