INPUT:

A table with columns of words, like this:

    LOC1   LOC2   LOC3
     a1     a2     a3
     b1     b2     b3
     c1            c3

There can be as many input columns and rows as you like. Some cells can
be empty (as in this case, there is no c2).

OUTPUT:

A table like this:

    LOC1   LOC2   LOC3   LOC1__LOC2   LOC1__LOC3   LOC2__LOC3
     a1     a2     a3     L(a1,a2)     L(a1,a3)     L(a2,a3)
     b1     b2     b3     L(b1,b2)     L(b1,b3)     L(b2,b3)
     c1            c3     NA           L(c1,c3)     NA      

  L(a1,a2) = Levenshtein difference between strings a1 and a2
  NA = not available

Some of the extra columns may be missing, depending on options you have
set.

================================================================

DETAILS

The input file must be a tab-delimited file with raw, unquoted strings:
- Row items (empty items too) are separated by a single tab. 
- Any quotes or slashes are considered to be just part of the string.

See example file: sample.data

The output file is also a tab-delimited file with raw, unquoted strings.

================================================================

PROGRAM

The program that does all the work is: columns.py
You can call the program directly from the command line, passing
settings as command line options.
Or you can use the GUI: columnsGUI.py

To test the program, use as input file: sample.data
If you want to try features, select the feature definition file:
features-example.txt

================================================================

IMPORTING RESULTS 

When you import the results into another program, such as a spreadsheet,
there are two things to pay attention to.

- You need to set the import format to tab-delimited, for instance in
  Excel, select: Delimited, Delimiters: Tab (and no others).

- You need to import the text columns as raw text (quotes and
  backslashes are just part of the text), for instance in Excel, select:
  Text Qualifier: {none}

There is an example Excel worksheet with two tabs. One tab is an import of
the results on 'sample.data' without features. The other tab is an
import of the results on 'sample.data' with features from
'features-example.txt'.