Changes to Spreadsheet::Edit, Spreadsheet::Edit::IO, Spreadsheet::Edit::Log

3.021 - Fix release mistake (wrongly required Win32API::Net on all platforms)
3.020 - Now fully working on Windows (Strawberry perl)

3.015->3.019: Fixing portability issues with installation tests

3.014 - IO::convert_spreadsheet() reworked.

        Auto-detect CSV import column formats when possible to avoid
        corruption when converted to a spreadsheet.  Mainly this is to
        specify "text" format for things which would be spoiled if read
        as a number e.g. zip codes with leading zeroes.  
        Also detects MM/DD/YY vs DD/MM/YY when the DD happens to be > 12.

        Import formats may be provided explicitly with option
           col_formats => [...] 
        The value is a list of codes for columns: "text", 
        "MM/DD/YY", "DD/MM/YY", or undef/empty for Standard 
        (e.g. generic number), or equivalent OO/LO Format Code numbers 
        See https://wiki.documentfoundation.org/Documentation/DevGuide/Spreadsheet_Documents#Filter_Options_for_the_CSV_Filter

        Now csv->ods->csv round trip is usually completely clean.

3.010 - Clean up auto-detect of encoding and quote & separator characters
        in csv files.   It first tries the obvious: UTF-8, '"' and ",", but
        should now work automatically with \t separators.

        Override with 
          input_encoding => "UTF-8",  # for example
          quote_char     => '"',
          sep_char       => ','
        ----
        Support CSV transcoding e.g.
          convert_spreadsheet( inpath => "file_in_utf16.csv", 
                               cvt_to => "csv",
                               output_encoding => "UTF-8" )

        will auto-detect the input encoding (using BOM if present)
        and transcode to a UTF-8 outpath.
        ---
        Separate and generalize the logging functions into
        Spreadsheet::Edit::Log .
        ---
3.009 - Switch tests from Test::More to Test2::V0