• fix bug with integer conversion (http://dev.swissr.org/issues/113)

• PROBLEM: values outside the integer range (i.e. 12345678901) didn't give an NA (and a warning message) but have been converted to a wrong value
• WHERE:
• free/pro: reading an Excel sheet and explicitely defining: 'colClasses = "integer"' or 'type = "integer". It's assumed that only very few people do this.
• pro: when (auto) detecting a data.frame column type for numeric values the type become an 'integer' in some situations. This probably affects more people than above
•   SOLUTION:
• an error will be raised for too large/small values (same as in read.table)
• numeric data.frame column types default to 'double' (as in the free version)

• adapt RUnit tests to new behaviour   
• some small cosmetic changes

Credits to Gyorgy Ottucsak who found the problem in xlsReadWritePro and send us a report (incl. testfile). Thanks a lot!