Modgrammar

* Issue 11: Change parse_string to assume self-contained input
* Issue 15: Change grammar_whitespace_mode to be 'explicit' by default
* Issue 16: Change SPACE to not match newlines
* Issue 19: Makesure all whitespace/EOL handling is unicode-aware
* Issue 21: Add the ability to produce debug output
* Issue 26: Add the ability to use regexp character class escapes in WORDs
* Issue 31: repr() breaks for abstract classes
* Removed deprecated find_tag and find_tag_all methods
* Added matchtype='complete' option to parser
* Added "encoding" parameter for parse_file()
* Added grammar_noteworthy attribute
* Deprecated WORD(longest=True) in favor of WORD(fullmatch=True)
* Misc doc cleanups

* Issue 20: Repetitions incorrectly allowing null subtokens
* Issue 22: REFs don't have standard attributes (like grammar_name)
* Issue 23: EXCEPT stops after first successful match, even if it doesn't
meet other criteria
* Issue 24: hash() breaks on recursive grammars after grammar_resolve_refs
* Issue 25: WORD("^") does not work as expected
* Fix a leftover reference in the docs to find_tag* methods
* Add a few missing attributes to classdict_map

* Clean up some warning messages from MANIFEST.in rules
* Made ParseError "found text" chooser stop at end-of-line
* Added WS_DEFAULT and WS_NOEOL standard whitespace regexps
* Fixed Issue 7: Changed get/get_all/find/find_all to match either subclass
or tag.
* Fixed Issue 8: Grammars with NOT_FOLLOWED_BY don't correctly return a parse
error before terminating the generator
* Fixed Issue 10: EXCEPT incorrectly matches on prefixes of the result
* Fixed Issue 13: Added "longest=True" option to WORD
* Fixed Issue 14: Add grammar_whitespace_mode setting
* Fixed Issue 17: Added infix_precedence example to docs
* Issue 11: Deprecated parse_string in favor of parse_text
* Issue 16: Deprecated SPACE in favor of WHITESPACE
* Issue 15: Deprecated defining grammars without explicitly specifying
grammar_whitespace_mode
* Issue 18: Deprecated grammar_whitespace = True/False

* Added NOT_FOLLOWED_BY grammar construct
* Added grammar_error_override option
* matchtype="all" no longer auto-advances the buffer position
* New GrammarParser.skip() method
* Fixed Issue 1: matchtype="longest" or "shortest" breaks when possible
matches are the same length.
* Fixed Issue 2: GrammarDefError undefined
* Fixed Issue 3: Misc documentation fixups
* Fixed Issue 4: Whitespace before EOF confuses parse_lines

* Added support for pickling result objects
* Added the ability to use a custom regular expression for whitespace-handling
* Made grammar classes compare equal if they are equivalent.
* Added a bunch more unit tests
* Fixed some issues where grammar_whitespace was not being set correctly on
some constructs.
* Fixed a bunch of misc whitespace parsing bugs

* First packaged version