fix(decoder): normalize delimiter names in decode options

[gstarr-cigna]

Summary

toon <file>.toon -d (and any caller passing a named delimiter) decodes tabular arrays into garbage:

$ echo 'products[3]{id,name,price}:
  101,Laptop Pro,1299
  102,Magic Mouse,79
  103,USB-C Cable,19' | toon -d
{
  "products": [
    { "id": "101,Laptop Pro,1299", "name": null, "price": null },
    ...

Root cause

The CLI passes the argparse choice strings 'comma' / 'tab' / 'pipe' as default_delimiter, but DecoderOptions stored the value unchanged and used it directly as the row-splitting character. Rows were split on the literal 5-character string "comma", which never matches, so each row came back as one field.

The encoder already normalizes these names (EncoderOptions, toon/encoder.py); the decoder did not.

Fix

Mirror the encoder's normalization in DecoderOptions (toon/decoder.py): map 'comma'→,, 'tab'→\t, 'pipe'→|. Character forms continue to work unchanged.

Added test_decode_delimiter_name_normalization covering both named and character delimiter forms. Full suite: 137 passed, 1 skipped.