nix-conf/user/bin/furigana
2025-07-26 16:07:57 -05:00

49 lines
1.7 KiB
Python
Executable file

"""
ぁあぃいぅうぇえぉおかがきぎくぐけげこごさざしじすずせぜそぞただちぢっつづてでとどなにぬねのはばぱひびぴふぶぷへべぺほぼぽまみむめもゃやゅゆょよらりるれろわをんーゎゐゑゕゖゔゝゞ・「」。、
ァアィイゥウェエォオカガキギクグケゲコゴサザシジスズセゼソゾタダチヂッツヅテデトドナニヌネノハバパヒビピフブプヘベペホボポマミムメモャヤュユョヨラリルレロワヲンーヮヰヱヵヶヴヽヾ・「」。、
"""
# vi: filetype=python
import fugashi
import sys
import re
if len(sys.argv) != 2:
print("usage: furigana <word>")
exit(1)
H, K = __doc__.strip().split("\n")
trans = {ord(a): ord(b) for a, b in zip(K, H)}
def align_kana(s: list[str], k: str):
if not s:
return ""
if not k:
return "".join(s)
if s[-1][0] in __doc__:
return align_kana(s[:-1], k[:-len(s[-1])]) + s[-1]
if s[0][0] in __doc__:
return s[0] + align_kana(s[1:], k[len(s[0]):])
if len(s) == 1:
return f" {s[0]}[{k}]"
if any(x[0] in __doc__ for x in s):
i = next(i for i, x in enumerate(s) if x[0] in __doc__)
j = k.find(s[i].translate(trans))
if j != -1:
x = align_kana(s[:i], k[:j])
y = align_kana(s[i+1:], k[j+len(s[i]):])
return x + s[i] + y
return f" {''.join(s)}[{k}]"
output = []
for x in fugashi.Tagger().parseToNodeList(sys.argv[1]):
if x.feature.kana:
s = re.findall("[ぁ-ゔァ-ヴー]+|[^ぁ-ゔァ-ヴー]+", x.surface)
k = x.feature.kana.translate(trans)
output.append(align_kana(s, k))
else:
output.append(x.surface)
print("".join(output).strip())