Skip to content

Commit

Permalink
✨ 大幅优化搜索能力, 默认取消使用MyAnimeList搜索, 适配DBD-Raws
Browse files Browse the repository at this point in the history
  • Loading branch information
@KimigaiiWuyi
KimigaiiWuyi committed Nov 17, 2024
1 parent bb8c855 commit 68ee13e
Showing 1 changed file with 116 additions and 44 deletions.
160 changes: 116 additions & 44 deletions src/Bangumi_Auto_Rename.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,8 @@
'访谈',
'Preview',
'预告',
'特典',
'映像',
]
S0_TAG = [
r'OVA',
Expand All @@ -59,6 +61,28 @@
'.rm',
'.rmvb',
]

keywords = [
'01',
'1080P',
'FLAC',
'简繁',
'外挂',
'MKV',
'MP4',
'TV',
'全集',
'HEVC',
'8bit',
'10bit',
'720P',
'2160P',
'4K',
'BD',
'RIP',
'DBD-raws',
]

bracket_patterns = [
r'\[.*?\]',
r'【.*?】',
Expand All @@ -85,7 +109,7 @@
}
season_partten = [
r'S([\d]{1,2})',
r'第([\d一二三四五六七八九零]{1,2})季',
r'第([\d一二三四五六七八九零]{1,2})(季|部分|部)',
r'([\d]{1,2})nd Season',
r'Season ([\d]{1,2})',
]
Expand Down Expand Up @@ -125,6 +149,25 @@
UNDERLINE = "\033[4m"


def clean_title_case_insensitive(title: str):
# 将关键词和标题转换为小写进行匹配
lower_keywords = [kw.lower() for kw in keywords]
j = '|'.join(re.escape(kw) for kw in lower_keywords)
keyword_regex = re.compile(j)

# 遍历所有括号类型
for pattern in bracket_patterns:
# 查找所有匹配的括号内容
matches = re.findall(pattern, title) # 保留原始大小写内容
for match in matches:
# 转为小写进行匹配
if keyword_regex.search(match.lower()):
title = title.replace(match, '') # 删除原始大小写内容

# 返回清理后的标题
return title.strip()[1:-1]


def remove_similar_part(common_parts: List[str], filename: str):
for common_part in common_parts:
if len(common_part) > 3: # 确保只移除长度大于3的部分
Expand Down Expand Up @@ -165,23 +208,6 @@ def find_common_substrings_in_all(
return final_common_substrings


# 移除相似的部分
def remove_similar_parts(
filenames: List[str],
common_parts: List[str],
) -> List[str]:
cleaned_filenames: List[str] = []

for filename in filenames:
for common_part in common_parts:
if len(common_part) > 3: # 确保只移除长度大于3的部分
pattern = re.escape(common_part)
filename = re.sub(pattern, '', filename).strip()
cleaned_filenames.append(filename)

return cleaned_filenames


# 遍历路径并筛选出视频文件
def find_unique_parts_in_videos(
directory: Path,
Expand All @@ -198,9 +224,6 @@ def find_unique_parts_in_videos(
# 找出所有文件的公共相似部分
common_parts = find_common_substrings_in_all(filenames)

# 移除相似部分，保留不同部分
# unique_parts = remove_similar_parts(filenames, common_parts)

return common_parts


Expand Down Expand Up @@ -270,7 +293,7 @@ def chinese_to_number(chinese_numeral):

def extract_season(text: str):
# 匹配 第1季, 第二季 等
match = re.search(r'第([\d一二三四五六七八九零]{1,2})季', text)
match = re.search(r'第([\d一二三四五六七八九零]{1,2})(季|部分|部)', text)
if match:
season_str = match.group(1)
if season_str.isdigit():
Expand All @@ -285,7 +308,7 @@ def extract_season(text: str):
return season_number

for p in season_partten:
if p != r'第[\d一二三四五六七八九零]({1,2})季':
if p != r'第([\d一二三四五六七八九零]{1,2})(季|部分|部)':
match = re.search(p, text)
if match:
return int(match.group(1))
Expand All @@ -306,7 +329,8 @@ def remove_code(s: str) -> str:
return s


def remove_tag(s: str, skip=False):
def remove_tag(title: str, skip=False):
s = title
if skip:
# 创建一个字典来追踪每种括号的匹配次数
counts = {pattern: 0 for pattern in bracket_patterns}
Expand All @@ -328,6 +352,11 @@ def replace_match(pattern, match):
for pattern in bracket_patterns:
s = re.sub(pattern, '', s)

remove_tag_s = s.strip()
print(f'【移除标签】：{remove_tag_s}')
if not remove_tag_s:
s = clean_title_case_insensitive(title)

return s.strip()


Expand Down Expand Up @@ -445,6 +474,7 @@ def get_tv_season_info(tv: Dict) -> List[Dict]:


def process_sub(
itme_path_main_name: str,
item_repeat: Optional[List[str]],
item_path: Path,
work_path: Path,
Expand All @@ -460,6 +490,9 @@ def process_sub(
item_name_l = item_name_remove.lower()
item_suffix = item_path.suffix.lower()

n_item_name_l = item_name.replace(itme_path_main_name, '').lower()
print(f'移去主要内容后的文件名Lower：{n_item_name_l}')

for ignore_dir in IGNORE_DIR:
if ignore_dir in item_path.name:
print(f'忽略文件夹：{item_path.name}')
Expand All @@ -471,7 +504,7 @@ def process_sub(
break
else:
for ex in EXTRA_TAG:
if re.search(rf'{ex.lower()}[\d]{{0,3}}', item_name_l):
if re.search(rf'{ex.lower()}', n_item_name_l):
t = work_path / 'extra'
R[item_path] = t / item_name
break
Expand Down Expand Up @@ -561,24 +594,28 @@ def process_path(path: Path, R: Dict[Path, Path]):
name, tv_info = get_tv_info(rtpath_name)
print(f'剧集名称: {name}')
if IS_ANIME:
search_result = jikan.search(
'anime',
rtpath_name,
page=1,
)
for i in search_result['data']:
if i['type'] == 'Anime':
data = i
break
else:
if not name:
print('【TMDB未搜索到】转为MyAnimeList搜索！')
search_result = jikan.search(
'anime',
rtpath_name,
page=1,
)
for i in search_result['data']:
if i['type'] == 'TV':
if i['type'] == 'Anime':
data = i
break
else:
data = search_result['data'][0]
titles = data['titles']
print(f'【MyAnimeList】【识别结果】: {titles}')
for i in search_result['data']:
if i['type'] == 'TV':
data = i
break
else:
data = search_result['data'][0]
titles = data['titles']
print(f'【MyAnimeList】【识别结果】: {titles}')
else:
titles = None
_WORK_PATH = ANIME_PATH
else:
titles = [{'type': 'Default', 'title': name}]
Expand Down Expand Up @@ -628,15 +665,35 @@ def process_path(path: Path, R: Dict[Path, Path]):
print(f'【提取信息季号】:{int_season}')
int_rtpath_name = extract_season(path.name)
print(f'【提取标题季号】:{int_rtpath_name}')
if int_season == int_rtpath_name:
season_id = int_rtpath_name
break
season_id = int_rtpath_name
break

# 如果不是Season1的情况下，sname处于路径之中，则直接跳过
if not (sname.strip().startswith('Season') and '1' in sname):
if sname in path.name:
break

if not titles:
print('【SEASON】【TMDB未搜索到】转为MyAnimeList搜索！')
search_result = jikan.search(
'anime',
rtpath_name,
page=1,
)
for i in search_result['data']:
if i['type'] == 'Anime':
data = i
break
else:
for i in search_result['data']:
if i['type'] == 'TV':
data = i
break
else:
data = search_result['data'][0]
titles = data['titles']
print(f'【MyAnimeList】【识别结果】: {titles}')

# 或者计算相似度
for title in titles:
similaritys = {}
Expand Down Expand Up @@ -683,13 +740,28 @@ def process_path(path: Path, R: Dict[Path, Path]):

print(path)
repeat = find_unique_parts_in_videos(path)
itme_name = remove_tag(rtpath_name).strip('!')
for item_path in path.iterdir():
if item_path.is_dir():
repeat = find_unique_parts_in_videos(item_path)
for sub_item in item_path.iterdir():
process_sub(repeat, sub_item, work_path, R, season_id)
process_sub(
itme_name,
repeat,
sub_item,
work_path,
R,
season_id,
)
else:
process_sub(repeat, item_path, work_path, R, season_id)
process_sub(
itme_name,
repeat,
item_path,
work_path,
R,
season_id,
)

trans_file(R)
R.clear()
Expand Down

0 comments on commit 68ee13e

Please sign in to comment.