This is not a bug. Simply, input X
ML is not well-formed.
& should be escaped with &
http://www.marketwatch.com/tools/quotes/news.asp?symb=MSFT&doctype=200
to
http://www.marketwatch.com/tools/quotes/news.asp?symb=MSFT&doctype=200.
This is a bug due to GUI file encodings. Now it is possible to choose encoding from settings dialog and default encoding is UTF-8 which supports such characters. Fix is available from Web-Harvest 2.0 coming soon.