So you have an analyzing morphological transducer and would like to make an
installable tool around it for analyzing running text, and generating wordforms
from morphological tags given by the user.

The main command for this is make-installable.sh. See
./make-installable.sh --help.

For now it will also generate a separate tokenizer, effectively including three
copies of the morphology (one for analyzing, one for generating, one for
tokenizing). Hopefully this will be corrected in the future.

For best results, check the following things:

1) Does the morphology handle capital letters at the beginning of words? If not,
you should use --capcase

2) Does the morphology know about punctuation like .,!? If not, consider using
--punct

Notes about args for existing transducers:

French: --capcase --punct --unweighted
Italian: --capcase --unweighted
Finnish: [none, but consider using tokenizer from
  scripts/tokenization/omorfi-tokenize]
Turkish: --punct --capcase --unweighted
Swedish: --punct --capcase --unweighted
German: --unweighted
English: --capcase
