Fix segmentation fault of StringScanner#charpos when String#byteslice returns non string value [Bug #17756] (#20)

kachick · web-flow · commit 92961cde2b42 · 2021-03-31T14:56:28.000+09:00
diff --git a/ext/strscan/strscan.c b/ext/strscan/strscan.c
@@ -445,13 +445,10 @@ static VALUE
 strscan_get_charpos(VALUE self)
 {
     struct strscanner *p;
-    VALUE substr;
 
     GET_SCANNER(self, p);
 
-    substr = rb_funcall(p->str, id_byteslice, 2, INT2FIX(0), LONG2NUM(p->curr));
-
-    return rb_str_length(substr);
+    return LONG2NUM(rb_enc_strlen(S_PBEG(p), CURPTR(p), rb_enc_get(p->str)));
 }
 
 /*
diff --git a/test/strscan/test_stringscanner.rb b/test/strscan/test_stringscanner.rb
@@ -206,6 +206,23 @@ def test_pos_unicode
     assert_equal 11, s.charpos
   end
 
+  def test_charpos_not_use_string_methods
+    string = +'abcädeföghi'
+    scanner = create_string_scanner(string)
+
+    class << string
+      EnvUtil.suppress_warning do
+        undef_method(*instance_methods)
+      end
+    end
+
+    assert_equal 0, scanner.charpos
+    assert_equal "abcä", scanner.scan_until(/ä/)
+    assert_equal 4, scanner.charpos
+    assert_equal "defö", scanner.scan_until(/ö/)
+    assert_equal 8, scanner.charpos
+  end
+
   def test_concat
     s = create_string_scanner('a'.dup)
     s.scan(/a/)

Original file line number	Diff line number	Diff line change
`@@ -445,13 +445,10 @@ static VALUE`
`445`	`445`	`strscan_get_charpos(VALUE self)`
`446`	`446`	`{`
`447`	`447`	`struct strscanner *p;`
`448`		`- VALUE substr;`
`449`	`448`
`450`	`449`	`GET_SCANNER(self, p);`
`451`	`450`
`452`		`- substr = rb_funcall(p->str, id_byteslice, 2, INT2FIX(0), LONG2NUM(p->curr));`
`453`		`-`
`454`		`- return rb_str_length(substr);`
	`451`	`+ return LONG2NUM(rb_enc_strlen(S_PBEG(p), CURPTR(p), rb_enc_get(p->str)));`
`455`	`452`	`}`
`456`	`453`
`457`	`454`	`/*`