Public-Mirror/LunaTranslator
1
0
Fork 0
mirror of https://github.com/HIllya51/LunaTranslator.git synced 2025-01-01 10:04:12 +08:00
Code Issues Projects Releases Wiki Activity

weight

Browse Source
This commit is contained in:
test123456654321 2024-10-26 12:50:27 +08:00
parent c3949794ba
commit 469b057707
1 changed files with 44 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

45
src/LunaTranslator/myutils/languageguesser.py
View File

@ -1,4 +1,5 @@
from collections import Counter from collections import defaultdict
import re
def inrange(n, s, e): def inrange(n, s, e):
@ -37,7 +38,8 @@ def guess(string: str):
(0xA960, 0xA97F), (0xA960, 0xA97F),
(0xD7B0, 0xD7FF), (0xD7B0, 0xD7FF),
), ),
"ja": lambda c: inranges( "ja": {
lambda c: inranges(
ord(c), ord(c),
(0x3040, 0x309F), (0x3040, 0x309F),
(0x30A0, 0x30FF), (0x30A0, 0x30FF),
@ -46,11 +48,18 @@ def guess(string: str):
(0x3100, 0x312F), (0x3100, 0x312F),
(0x31A0, 0x31BF), (0x31A0, 0x31BF),
(0x3000, 0x303F), (0x3000, 0x303F),
), ): 10,
"zh": lambda c: inranges( lambda c: inranges(
ord(c), ord(c),
(0x4E00, 0x9FA5), (0x4E00, 0x9FA5),
), ): 3,
},
"zh": {
lambda c: inranges(
ord(c),
(0x4E00, 0x9FA5),
): 5
},
"ar": lambda c: inranges( "ar": lambda c: inranges(
ord(c), ord(c),
(0x0600, 0x06FF), (0x0600, 0x06FF),
@ -63,17 +72,27 @@ def guess(string: str):
(0x0660, 0x0669), (0x0660, 0x0669),
(0x06F0, 0x06F9), (0x06F0, 0x06F9),
), ),
"en": lambda c: inranges( "en": {
lambda c: inranges(
ord(c), ord(c),
(0x0000, 0x00FF), (0x0000, 0x00FF),
), ): 0.2
},
} }
cnt = [] string = re.sub(r"ZX\wZ", "", string)
cnt = defaultdict(int)
for c in string.strip(): for c in string.strip():
for lang in checkers: for lang, ck in checkers.items():
if checkers[lang](c): if isinstance(ck, dict):
cnt.append(lang) for f, w in ck.items():
print(lang, f, c, f(c))
if f(c):
cnt[lang] += w
else:
if ck(c):
cnt[lang] += 1
if not cnt: if not cnt:
return "en" return "en"
common = Counter(cnt).most_common()
return common[0][0] max_key = max(cnt, key=cnt.get)
return max_key