Morphological attributes are extracted from information present in the Tanl morphed POS tags produced by the POS tagger, and added as a separate attribute to tokens..
This is done by the script morpho-splitter.py present in the PosTagger directory.
It splits clitics and morphology, creating a ConllX file with columns:
id form lemma cpos pos morph
The input should be a file with three columns:
form pos lemma
where pos is a combined POS+Morph tag and lemma contains a representation of clitics, when present, separated by dashes.
morph-splitter.py [options] [file]
-h, --help display this help and exit --usage display script usage