Pylightxl

- added comment parsing, see issue [45](https://github.com/PydPiper/pylightxl/issues/41)
- DEPRECATION WARNING: all indexing method that use "formula" as an argument will be replaced
with "output" in future version. Please update your codebase to use "output" instead of "formula".
This was done to simplify indexing the value (``output='v'``), the formula (``output='f'``) or the
comment (``output='c'``).
- added file stream reading for ``readxl`` that now supports ``with block`` for reading. See issue [25](https://github.com/PydPiper/pylightxl/issues/25)

- added handling for datetime parsing

- bug fix: writing to existing file previously would only write to the current working directory, it
now can handle subdirs. In addition inadvertently discovered a bug in python source code ElementTree.iterparse
where ``source`` passed as a string was not closing the file properly. We submitted a issue to python issue tracker.

- updated reading error'ed cells "N/A"
- updated workbook indexing bug from program generated workbooks that did not index from 1

- license update

-----------------
- hot-fix: added python2 support for encoding with cgi instead of html