Wraps the 'AntiWord' utility to extract text from Microsoft
Word documents. The utility only supports the old 'doc' format, not the
new xml based 'docx' format. Use the 'xml2' package to read the latter.
Fix gcc8 warning (requested by CRAN)
Windows: shQuote() path to file to make it work for paths with spaces
Capture error messages sent to stderr() by antiword