Csvmatch

* Detect character encoding automatically quicker in large files
* Show progress bar (except for Bilenko fuzzy matches)

* Fixes bug whereby columns with the same name from two different files would be coalesced
* Adds Jaro-Winkler fuzzy matching algorithm

* Adds argument to specify fuzzy match threshold (currently only used by Levenshtein)
* Adds argument to show which version of CSV Match is running(!)
* Splits out functionality from command line interface to allow testing
* A suite of tests
* Readme improvements

* Renames `--strip-nonalpha` to `--ignore-nonalpha`
* Adds argument to convert non-Latin characters to their Latin equivalents when comparing
* Adds option to include the matching degree in the output
* Increases the default threshold for a match when using Levenshtein from 40% to 60%
* Move to use a better Metaphone implementation
* Force Unix-style line endings on all platforms

* Adds output header shortcuts
* Fixes Levenshtein failing on Unicode characters
* Adds argument to allow file encodings to be specified
* Improved documentation

* Adds argument to specify which columns are included in the output