This is a partial copy of Hunspell 1.1.5, with the following changes:
* '#include "config.h"' removed from src/hunspell/hunspell.hxx
* '#include "config.h"' removed from src/hunspell/license.hunspell
* Two unreferenced local variables removed from src/hunspell/suggestmgr.cxx
* src/hunspell/utf_info.cxx moved to src/hunspell/utf_info.hxx, and #include
  reference in src/hunspell/csutil.cxx changed accordingly
* Change the input params of the constructors to receive a FILE* instead of
  a file path. This is required to use hunspell in the sandbox.
  The patch is in google.patch.

The English dictionary distributed by Firefox has been checked in to the
dictionaries directory. It has several additions over the default
myspell/hunspell dictionary.

* Workaround for non-ASCII characters

Visual Studio on Japanese Windows assumes the source files to be
encoded in Shift_JIS.  The compiler is unhappy with non-ASCII letters
in the source files of Hunspell.  The same problem happens with other
CJK Windows as well.  Here is the workaround for this problem:

Convert 8-bit bytes to hexadecimal escaped forms by

  % perl -i -De 's/([\x80-\xff])/sprintf("\\x%02x", $1)/ge' src/*.cxx


Note that Hunspell upstream is going to fix this problem.  We'll no
longer need the workaround if the problem is fixed in the upstream.