def __init__(self, model):
"""Create the UDPipe tool object."""
self.model = model
path = require_file(model)
self.tool = Model.load(path)
if not self.tool:
raise IOError("Cannot load model from file '%s'" % path)
self.error = ProcessingError()
self.conllu_reader = ConlluReader()
self.tokenizer = self.tool.newTokenizer(Model.DEFAULT)
def setUpClass(cls):
cls.doc = Document()
cls.data = os.path.join(os.path.dirname(tb2ud.__file__),
"../test/data/artificials.conllu")
cls._reader = ConlluReader(files=cls.data)
cls._reader.apply_on_document(cls.doc)
# cls.tree = cls.doc.bundles[0].get_tree()
# cls.nodes = cls.tree.descendants
cls.writer = ConlluWriter()
cls._subtreeconverted = False
def from_conllu_string(self, string):
"""Load a document from a conllu-formatted string."""
reader = ConlluReader(filehandle=io.StringIO(string))
reader.apply_on_document(self)
def load_conllu(self, filename=None, **kwargs):
"""Load a document from a conllu-formatted file."""
ConlluReader(files=filename, **kwargs).process_document(self)
def load_conllu(self, filename):
"""Load a document from a conllu-formatted file."""
reader = ConlluReader(files=filename)
reader.process_document(self)
def load_conllu(self, filename=None):
"""Load a document from a conllu-formatted file."""
reader = ConlluReader(files=filename)
reader.apply_on_document(self)