def extraction_pos1(list_srcs_dst, pos_1, pos_2):
pair = []
for line in list_srcs_dst:
dic_morph = defaultdict(lambda: '') # 動詞：動詞に係る助詞
dic_phrase = defaultdict(lambda: '')
for phrase in line:
morph_1 = phrase[1].check_pos_morph(pos_1)
morph_2 = phrase[0].check_pos_morph(pos_2)
phrase_2 = phrase[0].check_pos_phrase(pos_2)
if morph_1 is None:
continue
if morph_2 is not None:
# if morph_2 not in dic_morph[morph_1]:
dic_morph[morph_1] += ' ' + morph_2
dic_phrase[morph_1] += ' ' + phrase_2
for k, v in dic_morph.items():
pair.append([k, v, dic_phrase[k]])
return pair
if __name__ == '__main__':
text = 'neko.txt.cabocha'
sentences = chunk_list(text)
list_ = srcs_dst_list(sentences)
with open('knock46.txt', 'w') as f:
dic = extraction_pos1(list_, '動詞', '助詞')
for pair in dic:
f.write(f'{pair[0]}\t{pair[1]}\t{pair[2]}\n')
# -*- coding: utf-8 -*-
from knock41 import chunk_list
from knock42 import srcs_dst_list
from graphviz import Digraph
import pydot
if __name__ == '__main__':
text = 'neko.txt.cabocha'
sentences = chunk_list(text)
G = Digraph(format='png')
G.attr('node', shape='circle')
list_srcs_dst = srcs_dst_list(sentences)
edge = []
for i, line in enumerate(list_srcs_dst):
if i < 10:
for pair in line:
edge.append((str(pair[0].phrase_surface()), str(pair[1].phrase_surface())))
G.edge(str(pair[0].phrase_surface()), str(pair[1].phrase_surface()))
G.render('binary_tree_graphviz')
g = pydot.graph_from_edges(edge)
g.write_jpeg('binary_tree_pydot.png', prog='dot')
# for i, (dst, srcs) in enumerate(list_srcs_dst.items()):
# if i < 100:
# edge.append((str(dst.phrase_surface()), str(srcs.phrase_surface())))
# G.edge(str(dst.phrase_surface()), str(srcs.phrase_surface()))
# G.render('binaryfield_tree_graphviz')
# g = pydot.graph_from_edges(edge)
# g.write_jpeg('binary_tree_pydot.png', prog='dot')