You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
67 lines
2.3 KiB
67 lines
2.3 KiB
import codecs
|
|
import shutil
|
|
import os
|
|
import tempfile
|
|
from copy import deepcopy
|
|
|
|
import obeliks
|
|
|
|
import classla
|
|
from classla import Document
|
|
from classla.models.common.conll import CoNLLFile
|
|
|
|
from structure_assignment.constants import *
|
|
from structure_assignment.tweak_conllu import tweak as tweak_conllu
|
|
from nova_slovnica.translate_jos import translate as translate_jos
|
|
from nova_slovnica.conllu_to_xml import convert_file as conllu_to_tei
|
|
|
|
def create_nlp(resource_directory):
|
|
NLP_CONFIG_MAP['dir'] = resource_directory + '/classla'
|
|
return classla.Pipeline('sl', **NLP_CONFIG_MAP)
|
|
|
|
class Pipeline:
|
|
|
|
def __init__(self, nlp):
|
|
self.nlp = nlp
|
|
self.tmp_directory = tempfile.mkdtemp()
|
|
self.file_map = {key: self.tmp_directory + '/' + FILE_MAP[key] for key in FILE_MAP.keys()}
|
|
|
|
def import_file(self, file_name, file_key):
|
|
shutil.copyfile(file_name, self.file_map[file_key])
|
|
|
|
def do_tokenise(self):
|
|
input_file_name = self.file_map['strings-list']
|
|
output_file_name = self.file_map['obeliks-tokenised']
|
|
obeliks.run(in_file=input_file_name, out_file=output_file_name, conllu=True)
|
|
|
|
def do_tweak_conllu(self):
|
|
input_file_name = self.file_map['obeliks-tokenised']
|
|
output_file_name = self.file_map['obeliks-tweaked']
|
|
tweak_conllu(input_file_name, output_file_name)
|
|
|
|
def do_parse(self):
|
|
input_file_name = self.file_map['obeliks-tweaked']
|
|
output_file_name = self.file_map['classla-parsed']
|
|
doc = Document(text=None)
|
|
conll_file = CoNLLFile(filename=input_file_name)
|
|
doc.conll_file = conll_file
|
|
result = nlp(doc)
|
|
result.conll_file.write_conll(output_file_name)
|
|
|
|
def do_translate_jos(self):
|
|
input_file_name = self.file_map['classla-parsed']
|
|
dictionary_file_name = resource_directory + '/dict.xml'
|
|
output_file_name = self.file_map['classla-translated']
|
|
translate_jos(input_file_name, dictionary_file_name, output_file_name)
|
|
|
|
def do_conllu_to_tei(self):
|
|
input_file_name = self.file_map['classla-translated']
|
|
output_file_name = self.file_map['tei-initial']
|
|
conllu_to_tei(input_file_name, output_file_name)
|
|
|
|
def export_file(self, file_name, file_key):
|
|
shutil.copyfile(self.file_map[file_key], file_name)
|
|
|
|
def cleanup(self):
|
|
shutil.rmtree(self.tmp_directory, True)
|