for m in self.morphs:
if m.pos == pos:
return True
return False
def get_pos(self, pos):
res = []
for m in self.morphs:
if m.pos == pos:
res.append(m)
return res
if __name__ == '__main__':
res = []
for chunks in cabocha_into_chunks():
chunks = tuple(map(Chunk_normalized, chunks.values()))
for dc in chunks:
if not dc.has_pos('動詞'):
continue
# 動詞を含む文節において，最左の動詞の基本形を述語とする
v_base = dc.get_pos('動詞')[0].base
srcs = []
for sc_idx in dc.srcs:
# 述語に係る助詞を格とする
ms = chunks[sc_idx].get_pos('助詞')
if ms:
# 文節内に助詞が複数ある場合は最も右のものを選ぶ
srcs.append((ms[-1].base, chunks[sc_idx].norm))
if srcs:
def norm(self):
clause = ''.join(m.surface for m in self.morphs if m.pos != '記号')
return clause
def has_pos(self, pos):
for m in self.morphs:
if m.pos == pos:
return True
return False
if __name__ == '__main__':
def input_k():
return int(input('Enter a number (0: exit) -> '))
for k in iter(input_k, 0):
for chunks in islice(cabocha_into_chunks(), k - 1, k):
chunks = tuple(map(Chunk_normalized, chunks.values()))
G = pydot.Dot(graph_type='digraph')
for i, c in enumerate(chunks): # 同じ単語の可能性もあるので，id で区別
if c.norm:
color = 'red' if c.has_pos('名詞') else 'black'
G.add_node(pydot.Node(i, label=c.norm, color=color))
for i, c in enumerate(chunks):
if c.dst == -1:
continue
if c.norm and chunks[c.dst].norm:
G.add_edge(pydot.Edge(i, c.dst))
G.write_png(f'./out44_{k}.png')
def get_norm(self):
clause = "".join(m.surface for m in self.morphs if m.pos != "記号")
return clause
def has_pos(self, pos):
return any(m.pos == pos for m in self.morphs)
def get_pos(self, pos):
return (m for m in self.morphs if m.pos == pos)
if __name__ == "__main__":
res = []
# for chunks in cabocha_into_chunks():
for chunks in islice(cabocha_into_chunks(), 33, 34): # list[33:34] と同様
chunks = {k: ChunkNormalized(v) for k, v in chunks.items()}
for cv in filter(lambda c: c.has_pos("動詞"), chunks.values()):
# [45-1] 動詞を含む文節において，最左の動詞の基本形を述語とする
v_base = next(cv.get_pos("動詞")).base
srcs = []
for sc_idx in cv.srcs:
# [45-2] 述語に係る助詞を格とする
ms = tuple(chunks[sc_idx].get_pos("助詞"))
if ms:
# 文節内に助詞が複数ある場合は最も右のものを選ぶ
# e.g. 「次のように」
# Morph(surface='の', base='の', pos='助詞', pos1='連体化')
# Morph(surface='に', base='に', pos='助詞', pos1='副詞化')
# if len(ms) != 1:
# print(chunks[sc_idx].norm, ms)
def has_pos(self, pos):
return any(m.pos == pos for m in self.morphs)
def get_pos(self, pos):
return (m for m in self.morphs if m.pos == pos)
def has_sahen_wo(self):
for m1, m2 in zip(self.morphs, self.morphs[1:]):
if [m1.pos1, m2.pos, m2.base] == ["サ変接続", "助詞", "を"]:
return True
return False
if __name__ == "__main__":
res = []
for chunks in islice(cabocha_into_chunks(), 12, 13):
chunks = {k: ChunkNormalized(v) for k, v in chunks.items()}
for cv in filter(lambda c: c.has_pos("動詞"), chunks.values()):
# [45-1] 動詞を含む文節において，最左の動詞の基本形を述語とする
v_base = next(cv.get_pos("動詞")).base
srcs = []
# [47-1] 「サ変接続名詞+を（助詞）」で構成される文節が動詞に係る場合のみを
# 対象とする
failed = True
for sc_idx in reversed(cv.srcs):
if failed and chunks[sc_idx].has_sahen_wo():
# [47-2] 述語は「サ変接続名詞+を+動詞の基本形」とし，
# 文節中に複数の動詞があるときは，最左の動詞を用いる
v_base = chunks[sc_idx].norm + v_base
failed = False
else:
用語を -> 作り出した
"""
import os
import sys
from itertools import islice
from knock41 import cabocha_into_chunks
from knock45 import ChunkNormalized
sys.path.append(os.path.join(os.path.dirname(__file__), "../../"))
from kiyuna.utils.message import message # noqa: E402 isort:skip
if __name__ == "__main__":
res = []
# for chunks in cabocha_into_chunks():
for chunks in islice(cabocha_into_chunks(), 33, 34):
chunks = {k: ChunkNormalized(v) for k, v in chunks.items()}
for c in chunks.values():
if c.dst not in chunks:
# 名詞を含まないのパス
continue
# [48-1] 各文節は（表層形の）形態素列で表現する
tmp = [c.norm]
# [48-2] パスの開始文節から終了文節に至るまで，各文節の表現を"->"で連結する
dst = c.dst
while dst in chunks:
tmp.append(chunks[dst].norm)
dst = chunks[dst].dst
res.append(" -> ".join(tmp) + "\n")
sys.stdout.writelines(res)
message(f"write {len(res)} lines", type="success")