sane-tag-music: better romanization

This commit is contained in:
2024-07-13 04:50:32 +00:00
parent 1d4df82bde
commit deb355d960

View File

@@ -148,6 +148,7 @@ class MediaType(Enum):
Other = "other"
def maybe_romanize(a: str) -> str|None:
if a == "( ͡° ͜ʖ ͡°)": return "Lenny"
if a == "かめりあ": return "Camellia"
if a == "お握り": return "onigiri"
if a == "存流": return "ARU"
@@ -172,7 +173,12 @@ def romanize(a: str) -> str:
romanized = ""
index = 0
while index < len(a):
for segment_len in range(10, 0, -1):
if a[index].lower() in "abcdefghijklmnopqrstuvwxyz0123456789._-":
# fast path
romanized += a[index]
index += 1
continue
for segment_len in range(16, 0, -1):
r = maybe_romanize(a[index : index+segment_len])
if r is not None:
if len(r) > 1: