diff --git a/requirements.txt b/requirements.txt
index bf2e28a8..2e2218fc 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -25,4 +25,5 @@ jieba_fast
jieba
LangSegment>=0.2.0
Faster_Whisper
-wordsegment
\ No newline at end of file
+wordsegment
+srt
\ No newline at end of file
diff --git a/tools/i18n/i18n.py b/tools/i18n/i18n.py
index 00e91bf3..1861a034 100644
--- a/tools/i18n/i18n.py
+++ b/tools/i18n/i18n.py
@@ -3,22 +3,22 @@ import locale
import os
-def load_language_list(language):
- with open(f"./i18n/locale/{language}.json", "r", encoding="utf-8") as f:
+def load_language_list(language, locale_path="./i18n/locale"):
+ with open(os.path.join(locale_path, f"{language}.json"), "r", encoding="utf-8") as f:
language_list = json.load(f)
return language_list
class I18nAuto:
- def __init__(self, language=None):
+ def __init__(self, language=None, locale_path="./i18n/locale"):
if language in ["Auto", None]:
language = locale.getdefaultlocale()[
0
] # getlocale can't identify the system's language ((None, None))
- if not os.path.exists(f"./i18n/locale/{language}.json"):
+ if not os.path.exists(os.path.join(locale_path, f"{language}.json")):
language = "en_US"
self.language = language
- self.language_map = load_language_list(language)
+ self.language_map = load_language_list(language, locale_path)
def __call__(self, key):
return self.language_map.get(key, key)
diff --git a/tools/i18n/locale_diff.py b/tools/i18n/locale_diff.py
index 674f7dd2..e5527be3 100644
--- a/tools/i18n/locale_diff.py
+++ b/tools/i18n/locale_diff.py
@@ -2,11 +2,12 @@ import json
import os
from collections import OrderedDict
+dir_path = "./i18n/locale" # The path to the i18n locale directory, you can change it to your own path
+
# Define the standard file name
-standard_file = "locale/zh_CN.json"
+standard_file = os.path.join(dir_path, "zh_CN.json")
# Find all JSON files in the directory
-dir_path = "locale/"
languages = [
os.path.join(dir_path, f)
for f in os.listdir(dir_path)
@@ -30,7 +31,7 @@ for lang_file in languages:
# Add any missing keys to the language file
for key in diff:
- lang_data[key] = key
+ lang_data[key] = standard_data[key]
# Del any extra keys to the language file
for key in miss:
diff --git a/tools/i18n/scan_i18n.py b/tools/i18n/scan_i18n.py
index f3e52cf4..0db41df0 100644
--- a/tools/i18n/scan_i18n.py
+++ b/tools/i18n/scan_i18n.py
@@ -1,7 +1,20 @@
import ast
-import glob
import json
from collections import OrderedDict
+import os
+
+locale_path = "./i18n/locale" # The path to the i18n locale directory, you can change it to your own path
+scan_list = ["./",
+ "GPT_SoVITS/",
+ "tools/"
+ ] # The path to the directory you want to scan, you can change it to your own path
+scan_subfolders = False # Whether to scan subfolders
+
+
+# 你想要保留的特殊词汇
+special_words_to_keep = {
+
+}
def extract_i18n_strings(node):
@@ -21,20 +34,32 @@ def extract_i18n_strings(node):
return i18n_strings
+strings = []
-# scan the directory for all .py files (recursively)
# for each file, parse the code into an AST
# for each AST, extract the i18n strings
-
-strings = []
-for filename in glob.iglob("**/*.py", recursive=True):
- with open(filename, "r") as f:
+def scan_i18n_strings(filename):
+ with open(filename, "r", encoding="utf-8") as f:
code = f.read()
if "I18nAuto" in code:
tree = ast.parse(code)
i18n_strings = extract_i18n_strings(tree)
print(filename, len(i18n_strings))
strings.extend(i18n_strings)
+
+
+# scan the directory for all .py files (recursively)
+if scan_subfolders:
+ for folder in scan_list:
+ for dirpath, dirnames, filenames in os.walk(folder):
+ for filename in [f for f in filenames if f.endswith(".py")]:
+ scan_i18n_strings(os.path.join(dirpath, filename))
+else:
+ for folder in scan_list:
+ for filename in os.listdir(folder):
+ if filename.endswith(".py"):
+ scan_i18n_strings(os.path.join(folder, filename))
+
code_keys = set(strings)
"""
n_i18n.py
@@ -49,11 +74,13 @@ print()
print("Total unique:", len(code_keys))
-standard_file = "i18n/locale/zh_CN.json"
-with open(standard_file, "r", encoding="utf-8") as f:
- standard_data = json.load(f, object_pairs_hook=OrderedDict)
-standard_keys = set(standard_data.keys())
-
+standard_file = os.path.join(locale_path, "zh_CN.json")
+try:
+ with open(standard_file, "r", encoding="utf-8") as f:
+ standard_data = json.load(f, object_pairs_hook=OrderedDict)
+ standard_keys = set(standard_data.keys())
+except FileNotFoundError:
+ standard_keys = set()
# Define the standard file name
unused_keys = standard_keys - code_keys
print("Unused keys:", len(unused_keys))
@@ -64,12 +91,18 @@ missing_keys = code_keys - standard_keys
print("Missing keys:", len(missing_keys))
for missing_key in missing_keys:
print("\t", missing_key)
+
+
code_keys_dict = OrderedDict()
for s in strings:
- code_keys_dict[s] = s
+ if s in special_words_to_keep:
+ code_keys_dict[s] = special_words_to_keep[s]
+ else:
+ code_keys_dict[s] = s
# write back
+os.makedirs(locale_path, exist_ok=True)
with open(standard_file, "w", encoding="utf-8") as f:
json.dump(code_keys_dict, f, ensure_ascii=False, indent=4, sort_keys=True)
f.write("\n")
diff --git a/tools/srt_slicer/i18n/locale/en_US.json b/tools/srt_slicer/i18n/locale/en_US.json
new file mode 100644
index 00000000..415d36d2
--- /dev/null
+++ b/tools/srt_slicer/i18n/locale/en_US.json
@@ -0,0 +1,60 @@
+{
+ "List 合并小工具": "List Merge Tool",
+ "SRT合并切分插件": "SRT Merge and Split Plugin",
+ "SRT文件": "SRT File",
+ "SRT编辑界面": "SRT Edit Interface",
+ "srt文件内容": "SRT File Content",
+ "上传SRT文件": "Upload SRT File",
+ "上传文件": "Upload Files",
+ "两个文件夹不能相同！！！": "The two folders cannot be the same!!!",
+ "主文件夹": "Main Folder",
+ "作者: ": "Author: ",
+ "使用方法": "How to Use",
+ "保存合并后字幕": "Save Merged Subtitles",
+ "保存子文件夹名称": "Save Subfolder Name",
+ "保存文件夹": "Save Folder",
+ "允许最短长度": "Minimum Allowed Length",
+ "内容预览": "Content Preview",
+ "切分与保存": "Split and Save",
+ "切分完成": "Split Completed",
+ "切分并保存音频、list": "Split and Save Audio, List",
+ "切分预览": "Split Preview",
+ "判定为短间隔时长": "Judged as Short Interval Duration",
+ "到": " to ",
+ "前置保留时间": "Preceding Retention Time",
+ "前置添加静音时间": "Prepend Silence Time",
+ "句末加句号": "Add Period at the End of Sentence",
+ "合并后srt文本": "Merged SRT Text",
+ "合并后的List": "Merged List",
+ "合并字幕": "Merge Subtitles",
+ "合并字幕设置": "Subtitle Merge Settings",
+ "合并文件夹与List": "Merge Folder and List",
+ "后置保留时间": "Following Retention Time",
+ "后置添加静音时间": "Append Silence Time",
+ "扫描文件夹": "Scan Folder",
+ "找不到字幕！！！": "Subtitles Not Found!!!",
+ "找不到音频！！！": "Audio Not Found!!!",
+ "提供SRT文件（可使用剪映或者ASR工具获得）与原始音频文件。": "Provide SRT File (can be obtained via Clip or ASR tools) and Original Audio File.",
+ "提前合并时间间隔很短的字幕": "Merge Subtitles with Short Intervals in Advance",
+ "提示": "Tips",
+ "文件夹路径": "Folder Path",
+ "最大间隔时间": "Maximum Interval Time",
+ "最长允许单句长度": "Maximum Allowed Sentence Length",
+ "根据面板合并短句并过滤你不希望出现的句子。": "Merge short sentences according to the panel and filter out sentences you do not want to appear.",
+ "次文件夹": "Second Folder",
+ "正在切分音频": "Splitting Audio",
+ "正在建设，敬请期待": "Under Construction, Stay Tuned",
+ "注意：该文件夹已存在": "Warning: The folder already exists",
+ "角色名称，留空使用主文件夹的": "Role Name, Leave Blank to Use Main Folder's",
+ "语言": "Language",
+ "读取文件": "Read File",
+ "读取本地文件": "Read Local File",
+ "过滤字幕": "Filter Subtitles",
+ "过滤带有英文的": "Filter Out English",
+ "过滤设置": "Filter Settings",
+ "过滤词语，一行一个": "Filter Words, One Per Line",
+ "这是一个插件，用于依靠SRT文件得到切分与打标好的音频。": "This is a plugin for obtaining split and tagged audio based on SRT files.",
+ "随后保存成切分好的音频与list文件。": "Then save as split audio and list files.",
+ "音频文件": "Audio File",
+ "音频格式": "Audio Format"
+}
diff --git a/tools/srt_slicer/i18n/locale/zh_CN.json b/tools/srt_slicer/i18n/locale/zh_CN.json
new file mode 100644
index 00000000..27567f00
--- /dev/null
+++ b/tools/srt_slicer/i18n/locale/zh_CN.json
@@ -0,0 +1,60 @@
+{
+ "List 合并小工具": "List 合并小工具",
+ "SRT合并切分插件": "SRT合并切分插件",
+ "SRT文件": "SRT文件",
+ "SRT编辑界面": "SRT编辑界面",
+ "srt文件内容": "srt文件内容",
+ "上传SRT文件": "上传SRT文件",
+ "上传文件": "上传文件",
+ "两个文件夹不能相同！！！": "两个文件夹不能相同！！！",
+ "主文件夹": "主文件夹",
+ "作者: ": "作者: ",
+ "使用方法": "使用方法",
+ "保存合并后字幕": "保存合并后字幕",
+ "保存子文件夹名称": "保存子文件夹名称",
+ "保存文件夹": "保存文件夹",
+ "允许最短长度": "允许最短长度",
+ "内容预览": "内容预览",
+ "切分与保存": "切分与保存",
+ "切分完成": "切分完成",
+ "切分并保存音频、list": "切分并保存音频、list",
+ "切分预览": "切分预览",
+ "判定为短间隔时长": "判定为短间隔时长",
+ "到": "到",
+ "前置保留时间": "前置保留时间",
+ "前置添加静音时间": "前置添加静音时间",
+ "句末加句号": "句末加句号",
+ "合并后srt文本": "合并后srt文本",
+ "合并后的List": "合并后的List",
+ "合并字幕": "合并字幕",
+ "合并字幕设置": "合并字幕设置",
+ "合并文件夹与List": "合并文件夹与List",
+ "后置保留时间": "后置保留时间",
+ "后置添加静音时间": "后置添加静音时间",
+ "扫描文件夹": "扫描文件夹",
+ "找不到字幕！！！": "找不到字幕！！！",
+ "找不到音频！！！": "找不到音频！！！",
+ "提供SRT文件（可使用剪映或者ASR工具获得）与原始音频文件。": "提供SRT文件（可使用剪映或者ASR工具获得）与原始音频文件。",
+ "提前合并时间间隔很短的字幕": "提前合并时间间隔很短的字幕",
+ "提示": "提示",
+ "文件夹路径": "文件夹路径",
+ "最大间隔时间": "最大间隔时间",
+ "最长允许单句长度": "最长允许单句长度",
+ "根据面板合并短句并过滤你不希望出现的句子。": "根据面板合并短句并过滤你不希望出现的句子。",
+ "次文件夹": "次文件夹",
+ "正在切分音频": "正在切分音频",
+ "正在建设，敬请期待": "正在建设，敬请期待",
+ "注意：该文件夹已存在": "注意：该文件夹已存在",
+ "角色名称，留空使用主文件夹的": "角色名称，留空使用主文件夹的",
+ "语言": "语言",
+ "读取文件": "读取文件",
+ "读取本地文件": "读取本地文件",
+ "过滤字幕": "过滤字幕",
+ "过滤带有英文的": "过滤带有英文的",
+ "过滤设置": "过滤设置",
+ "过滤词语，一行一个": "过滤词语，一行一个",
+ "这是一个插件，用于依靠SRT文件得到切分与打标好的音频。": "这是一个插件，用于依靠SRT文件得到切分与打标好的音频。",
+ "随后保存成切分好的音频与list文件。": "随后保存成切分好的音频与list文件。",
+ "音频文件": "音频文件",
+ "音频格式": "音频格式"
+}
diff --git a/tools/srt_slicer/srt_utils.py b/tools/srt_slicer/srt_utils.py
new file mode 100644
index 00000000..9c80fc54
--- /dev/null
+++ b/tools/srt_slicer/srt_utils.py
@@ -0,0 +1,139 @@
+import srt
+import shutil
+
+def parse_srt_with_lib(content):
+
+ subtitles = list(srt.parse(content))
+ return subtitles
+
+def generate_srt_with_lib(subtitles):
+ content = srt.compose(subtitles)
+ return content
+
+def merge_subtitles_with_lib(subtitles, short_interval, max_interval, max_text_length=30, add_period=True, merge_zero_interval=True):
+ # 标点符号
+ punctuations = ["。","!", "！", "？", "?", "；", ";", "…"]
+ punctuations_extanded = punctuations
+ punctuations_extanded.extend([ "：", ":", "，", ",", "—",])
+
+ # 直接合并间隔特别短的字幕
+ if merge_zero_interval:
+ eps = short_interval
+ for i in range(len(subtitles) - 1, 0, -1):
+ if subtitles[i-1].content[-1] in punctuations_extanded:
+ continue
+ if abs(subtitles[i].start.total_seconds() - subtitles[i-1].end.total_seconds()) < eps:
+ subtitles[i - 1].end = subtitles[i].end
+ subtitles[i - 1].content += subtitles[i].content
+ subtitles.pop(i)
+
+ merged_subtitles = []
+ current_subtitle = None
+ for subtitle in subtitles:
+ if current_subtitle is None:
+ current_subtitle = subtitle
+ else:
+ current_end = current_subtitle.end.total_seconds()
+ next_start = subtitle.start.total_seconds()
+ if current_subtitle.content[-1] not in punctuations and (next_start - current_end <= max_interval and count_words_multilang(current_subtitle.content + subtitle.content) < max_text_length):
+ current_subtitle.end = subtitle.end
+ comma = '，' if current_subtitle.content[-1] not in punctuations_extanded else ''
+ current_subtitle.content += comma + subtitle.content
+
+ else:
+ if add_period and current_subtitle.content[-1] not in punctuations_extanded:
+ current_subtitle.content += '。'
+ merged_subtitles.append(current_subtitle)
+ current_subtitle = subtitle
+ if current_subtitle is not None:
+ merged_subtitles.append(current_subtitle)
+ # 重新分配id，因为srt.compose需要id连续
+ for i, subtitle in enumerate(merged_subtitles, start=1):
+ subtitle.index = i
+ return merged_subtitles
+
+
+
+def count_words_multilang(text):
+ # 初始化计数器
+ word_count = 0
+ in_word = False
+
+ for char in text:
+ if char.isspace(): # 如果当前字符是空格
+ in_word = False
+ elif char.isascii() and not in_word: # 如果是ASCII字符（英文）并且不在单词内
+ word_count += 1 # 新的英文单词
+ in_word = True
+ elif not char.isascii(): # 如果字符非英文
+ word_count += 1 # 每个非英文字符单独计为一个字
+
+ return word_count
+
+import pydub, os
+
+def slice_audio_with_lib(audio_path, save_folder, format, subtitles, pre_preserve_time, post_preserve_time, pre_silence_time, post_silence_time, language='ZH', character='character'):
+ list_file = os.path.join(save_folder, 'datamapping.list')
+ try:
+ audio = pydub.AudioSegment.from_file(audio_path)
+ except Exception as e:
+ raise e
+ with open(list_file, 'w', encoding="utf-8") as f:
+ for i in range(len(subtitles)):
+ subtitle = subtitles[i]
+ start = subtitle.start.total_seconds() - pre_preserve_time
+ end = subtitle.end.total_seconds() + post_preserve_time
+
+ if i < len(subtitles) - 1:
+ next_subtitle = subtitles[i + 1]
+ end = min(end, 1.0/2*(subtitle.end.total_seconds()+next_subtitle.start.total_seconds()))
+ if i > 0:
+ prev_subtitle = subtitles[i - 1]
+ start = max(start, 1.0/2*(prev_subtitle.end.total_seconds()+subtitle.start.total_seconds()))
+ try:
+ sliced_audio = audio[int(start * 1000):int(end * 1000)]
+ file_name = f'{character}_{i + 1:03d}.{format}'
+ save_path = os.path.join(save_folder, file_name)
+ sliced_audio.export(save_path, format=format)
+ f.write(f"{file_name}|{character}|{language}|{subtitle.content}\n")
+ print(f"Slice {file_name} from {start} to {end}")
+ except Exception as e:
+ raise e
+
+def merge_list_folders(first_list_file, second_list_file, character, first_folder, second_folder):
+ merged_lines = []
+ character1 = ""
+ filenames = set()
+ with open(first_list_file, 'r', encoding="utf-8") as f:
+ first_list = f.readlines()
+ for line in first_list:
+ filename, character1, language, content = line.split('|')
+ filenames.add(filename)
+ if character=="" or character is None:
+ character = character1
+ new_line = f"{filename}|{character}|{language}|{content}"
+ merged_lines.append(new_line)
+ with open(second_list_file, 'r', encoding="utf-8") as f:
+ second_list = f.readlines()
+ for line in second_list:
+ filename, _, language, content = line.split('|')
+ orig_filename = filename
+ num = 1
+ while filename in filenames:
+ filename = f"{filename.rsplit('.', 1)[0]}_{num}.{filename.rsplit('.', 1)[1]}"
+ num += 1
+ try:
+ os.rename(os.path.join(second_folder, orig_filename), os.path.join(first_folder, filename))
+ except Exception as e:
+ raise e
+ new_line = f"{filename}|{character}|{language}|{content}"
+ merged_lines.append(new_line)
+ os.remove(second_list_file)
+ if not os.listdir(second_folder):
+ os.rmdir(second_folder)
+ with open(first_list_file, 'w', encoding="utf-8") as f:
+ f.writelines(merged_lines)
+ return "\n".join(merged_lines)
+
+
+
\ No newline at end of file
diff --git a/tools/srt_slicer/webui.py b/tools/srt_slicer/webui.py
new file mode 100644
index 00000000..2dd6719d
--- /dev/null
+++ b/tools/srt_slicer/webui.py
@@ -0,0 +1,396 @@
+import gradio as gr
+
+import sys
+import os
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
+sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '../..')))
+sys.path.append(os.path.abspath(os.path.dirname(__file__)))
+sys.path.append('.')
+from srt_utils import (
+ merge_subtitles_with_lib,
+ parse_srt_with_lib,
+ generate_srt_with_lib,
+ slice_audio_with_lib,
+ count_words_multilang,
+ merge_list_folders
+)
+
+port = 8991
+is_share = True
+
+if len(sys.argv) > 2:
+ port = int(sys.argv[1])
+ is_share = eval(sys.argv[2])
+
+
+from i18n.i18n import I18nAuto
+
+import os
+
+i18n = I18nAuto(language=None, locale_path="./tools/srt_slicer/i18n/locale")
+
+def merge_srt(input_text, output_text, short_interval=0.1, max_interval=1, max_text_length=30, add_period=True, merge_zero_interval=True):
+ original_subtitles = parse_srt_with_lib(input_text)
+ merged_subtitles = merge_subtitles_with_lib(original_subtitles, short_interval, max_interval, max_text_length, add_period, merge_zero_interval)
+ output_text = generate_srt_with_lib(merged_subtitles)
+ return output_text
+
+
+def slice_audio(
+ input_audio,
+ save_folder,
+ audio_format,
+ output_text,
+ pre_preserve_time,
+ post_preserve_time,
+ pre_silence_time,
+ post_silence_time,
+ language,
+ character,
+
+):
+ if isinstance(input_audio, str) and input_audio != "":
+ pass
+ else:
+ gr.Warning(i18n("找不到音频！！！"))
+ return
+ if output_text == "":
+ gr.Warning(i18n("找不到字幕！！！"))
+ return
+
+ character_folder = os.path.join(save_folder, character)
+ os.makedirs(character_folder, exist_ok=True)
+ subtitles = parse_srt_with_lib(output_text)
+ try:
+ gr.Info(f"{i18n('正在切分音频')} {input_audio} {i18n('到')} {character_folder}")
+ slice_audio_with_lib(
+ input_audio,
+ character_folder,
+ audio_format,
+ subtitles,
+ pre_preserve_time,
+ post_preserve_time,
+ pre_silence_time,
+ post_silence_time,
+ language,
+ character,
+ )
+ gr.Info(f"{i18n('切分完成')} ")
+ except Exception as e:
+ gr.Warning(f"Can't Slice, Error: {e}")
+
+def get_relative_path(path, base):
+ return os.path.relpath(path, base)
+
+def get_srt_and_audio_files(folder):
+ if not os.path.exists(folder):
+ os.makedirs(folder, exist_ok=True)
+ srt_files = []
+ audio_files = []
+ audio_file_formats = ["mp3", "wav", "ogg", "flac"]
+ for root, dirs, files in os.walk(folder):
+ for file in files:
+ if file.lower().endswith(".srt"):
+ srt_files.append(get_relative_path(os.path.join(root, file), folder))
+ for audio_file_format in audio_file_formats:
+ if file.lower().endswith(audio_file_format):
+ audio_files.append(get_relative_path(os.path.join(root, file), folder))
+ srt_file = ""
+ audio_file = ""
+ if len(srt_files) > 0:
+ srt_file = srt_files[0]
+ if len(audio_files) > 0:
+ audio_file = audio_files[0]
+ return gr.Dropdown(srt_files,value=srt_file), gr.Dropdown(audio_files,value=audio_file)
+
+def change_srt_file(folder,srt_file):
+ srt_folder = os.path.dirname(os.path.join(folder, srt_file))
+ basename = os.path.basename(srt_file).rsplit(".", 1)[0]
+ audio_file_formats = ["mp3", "wav", "ogg", "flac"]
+ for file in os.listdir(srt_folder):
+ print(f"basename: {basename}, file: {file}")
+ if basename.lower() in file.lower():
+ for audio_file_format in audio_file_formats:
+ if file.lower().endswith(audio_file_format):
+ return gr.Dropdown(value=get_relative_path(os.path.join(srt_folder, file), folder))
+ return gr.Dropdown(interactive=True)
+
+def filter_srt(input_text, min_length, filter_english, filter_words):
+ subtitles = parse_srt_with_lib(input_text)
+ filtered_subtitles = []
+ for subtitle in subtitles:
+ if count_words_multilang(subtitle.content) >= min_length:
+ flag = False
+ if filter_english:
+ for i in subtitle.content:
+ if i in "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ":
+ flag = True
+ break
+ if not flag and filter_words:
+ filter_words.replace("\r", "\n")
+ for word in filter_words.split("\n"):
+ if word in subtitle.content:
+ flag = True
+ break
+ if not flag:
+ filtered_subtitles.append(subtitle)
+ return generate_srt_with_lib(filtered_subtitles)
+
+def load_srt_from_file(srt_file):
+ try:
+ with open(srt_file, "r", encoding="utf-8") as f:
+ return f.read()
+ except:
+ return ""
+
+def load_audio_from_file(audio_file):
+ try:
+ return gr.Audio(audio_file)
+ except:
+ return gr.Audio(value=None)
+
+def load_from_dropdown(input_folder, srt_files_list, audio_files_list):
+ if isinstance(srt_files_list, str) and isinstance(audio_files_list, str):
+ srt_file= os.path.join(input_folder, srt_files_list)
+ audio_file = os.path.join(input_folder, audio_files_list)
+ return load_srt_from_file(srt_file), load_audio_from_file(audio_file)
+ else:
+ return "", gr.Audio(value=None)
+
+def enable_gr_elements(*args):
+ return [gr.update(interactive=True) for _ in args]
+
+def disable_gr_elements(*args):
+ return [gr.update(interactive=False) for _ in args]
+
+def save_srt_to_file(srt_text, save_folder, character):
+ character_folder = os.path.join(save_folder, character)
+ os.makedirs(character_folder, exist_ok=True)
+ srt_file = os.path.join(character_folder, "merged.srt")
+ with open(srt_file, "w", encoding="utf-8") as f:
+ f.write(srt_text)
+
+def scan_list_folders(folder):
+ if not os.path.exists(folder):
+ os.makedirs(folder, exist_ok=True)
+ list_folders = []
+ for list_folder in os.listdir(folder):
+ if os.path.isdir(os.path.join(folder, list_folder)):
+ list_folders.append(get_relative_path(os.path.join(folder, list_folder), folder))
+ first_list_folder = ""
+ second_list_folder = ""
+ if len(list_folders) > 0:
+ first_list_folder = second_list_folder = list_folders[0]
+ if len(list_folders) > 1:
+ second_list_folder = list_folders[1]
+ return gr.Dropdown(list_folders, value=first_list_folder), gr.Dropdown(list_folders, value=second_list_folder)
+
+def preview_merged_list(first_list_folder, second_list_folder, merge_list_character_name, save_folder):
+ if first_list_folder == "" or second_list_folder == "":
+ return ""
+ if first_list_folder == second_list_folder:
+ gr.Warning(i18n("两个文件夹不能相同！！！"))
+ return ""
+ first_list_folder = os.path.join(save_folder, first_list_folder)
+ second_list_folder = os.path.join(save_folder, second_list_folder)
+ print(f"first_list_folder: {first_list_folder}, second_list_folder: {second_list_folder}")
+ first_list = os.path.join(first_list_folder, [file for file in os.listdir(first_list_folder) if file.lower().endswith(".list")][0])
+ second_list = os.path.join(second_list_folder, [file for file in os.listdir(second_list_folder) if file.lower().endswith(".list")][0])
+ try:
+ return merge_list_folders(first_list, second_list, merge_list_character_name, first_list_folder, second_list_folder)
+ except Exception as e:
+ gr.Warning(f"Can't Merge, Error: {e}")
+ return ""
+
+
+from datetime import datetime
+
+def change_character_name(input_audio):
+ try:
+ input_audio_name = os.path.basename(input_audio).rsplit(".", 1)[0]
+ character = input_audio_name[:20]
+ except:
+ character = datetime.now().strftime("%m%d%H%M")
+ return gr.Textbox(value=character)
+
+def check_character_foldfer(folder, character):
+ character_folder = os.path.join(folder, character)
+ if os.path.exists(character_folder):
+ return gr.Textbox(visible=True)
+ return gr.Textbox(visible=False)
+
+with gr.Blocks() as app:
+ with gr.Row():
+ gr.HTML(f"""{i18n("SRT合并切分插件")}
+ {i18n("这是一个插件，用于依靠SRT文件得到切分与打标好的音频。")}
{i18n("作者: ")}XTer
+ {i18n("使用方法")}
+
+ - {i18n("提供SRT文件（可使用剪映或者ASR工具获得）与原始音频文件。")}
+ - {i18n("根据面板合并短句并过滤你不希望出现的句子。")}
+ - {i18n("随后保存成切分好的音频与list文件。")}
+
""")
+ with gr.Tabs():
+ with gr.Tab(i18n("SRT编辑界面")):
+ with gr.Row():
+ with gr.Column(scale=2) as input_col:
+ with gr.Tabs():
+ with gr.Tab(i18n("读取本地文件")):
+ input_folder = gr.Textbox("input/srt_and_audios", label=i18n("文件夹路径"),interactive=True)
+ scan_button = gr.Button(i18n("扫描文件夹"), variant="secondary",interactive=True)
+ srt_files_list = gr.Dropdown([], label=i18n("SRT文件"),interactive=True)
+ audio_files_list = gr.Dropdown([], label=i18n("音频文件"),interactive=True)
+ srt_read_button = gr.Button(i18n("读取文件"), variant="secondary",interactive=True)
+ with gr.Tab(i18n("上传文件")):
+ input_srt_file = gr.File(label=i18n("上传SRT文件"), type="filepath", file_types=["srt"])
+ upload_audio = gr.Audio(type="filepath",label=i18n("音频文件"))
+ # input_audio_file = gr.File(label=i18n("上传音频文件"), type="audio", file_types=["mp3", "wav", "ogg"])
+ with gr.Tabs():
+ with gr.Tab(i18n("内容预览")):
+ input_audio = gr.Textbox("", label=i18n("音频文件"),interactive=False)
+ input_text = gr.Textbox("", lines=20, max_lines=30, label=i18n("srt文件内容"))
+ input_srt_file.change(load_srt_from_file, [input_srt_file], [input_text])
+ with gr.Column(scale=1) as control_col:
+ with gr.Tabs():
+ with gr.Tab(i18n("合并字幕设置")):
+ merge_zero_interval = gr.Checkbox(label=i18n("提前合并时间间隔很短的字幕"),interactive=True, value=True)
+ short_interval = gr.Slider(value=0.05, minimum=0, maximum=0.5, step=0.005, label=i18n("判定为短间隔时长"),interactive=True,visible=True)
+ max_interval = gr.Slider(value=0.8, minimum=0.1, maximum=10, step=0.1, label=i18n("最大间隔时间"),interactive=True)
+ max_text_length = gr.Slider(value=50,minimum=5,maximum=200,step=1, label=i18n("最长允许单句长度"),interactive=True)
+ add_period = gr.Checkbox(label=i18n("句末加句号"),interactive=True, value=True)
+ merge_button = gr.Button(i18n("合并字幕"), variant="primary")
+
+ with gr.Tab(i18n("过滤设置")):
+ min_length = gr.Slider(value=5, minimum=0, maximum=20, step=1, label=i18n("允许最短长度"),interactive=True)
+ filter_english = gr.Checkbox(label=i18n("过滤带有英文的"),interactive=True)
+ filter_words = gr.Textbox("", label=i18n("过滤词语，一行一个"),lines=5,max_lines=10,interactive=True)
+ filter_button = gr.Button(i18n("过滤字幕"), variant="primary",interactive=False)
+ with gr.Tab(i18n("切分与保存")):
+ with gr.Group():
+ pre_preserve_time = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, label=i18n("前置保留时间"),interactive=True)
+ post_preserve_time = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, label=i18n("后置保留时间"),interactive=True)
+ pre_silence_time = gr.Slider(value=0.05, minimum=0, maximum=1, step=0.01, label=i18n("前置添加静音时间"),interactive=True,visible=False)
+ post_silence_time = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, label=i18n("后置添加静音时间"),interactive=True,visible=False)
+ with gr.Group():
+ language = gr.Dropdown([i18n(i) for i in [ "ZH", "EN", "JA"]], value="ZH", label=i18n("语言"),interactive=True)
+ audio_format = gr.Dropdown(["mp3", "wav", "ogg"], value="wav", label=i18n("音频格式"),interactive=True)
+ with gr.Group():
+ save_folder = gr.Textbox("output/sliced_audio", label=i18n("保存文件夹"),interactive=True)
+ character = gr.Textbox("character", label=i18n("保存子文件夹名称"),interactive=True)
+ character_warning = gr.Textbox(i18n("注意：该文件夹已存在"), label=i18n("提示"),interactive=False,visible=False)
+ save_srt_button = gr.Button(i18n("保存合并后字幕"),variant="secondary",interactive=True)
+ slice_audio_button = gr.Button(i18n("切分并保存音频、list"), variant="primary",interactive=False)
+
+ with gr.Column(scale=2) as output_col:
+ with gr.Tabs():
+ with gr.Tab(i18n("合并后srt文本")):
+ output_text = gr.Textbox("", lines=20, max_lines=30, label="Sliced SRT")
+ with gr.Tab(i18n("切分预览")):
+ gr.Textbox(i18n("正在建设，敬请期待"), label=i18n("提示"),interactive=False)
+ with gr.Tab(i18n("List 合并小工具")):
+ with gr.Row():
+ with gr.Column(scale=2):
+ scan_list_folder = gr.Textbox("output/sliced_audio", label=i18n("文件夹路径"),interactive=True)
+ scan_list_button = gr.Button(i18n("扫描文件夹"), variant="secondary")
+ first_list_folder = gr.Dropdown([], label=i18n("主文件夹"),interactive=True)
+ second_list_folder = gr.Dropdown([], label=i18n("次文件夹"),interactive=True)
+ merge_list_character_name = gr.Textbox("", label=i18n("角色名称，留空使用主文件夹的"),interactive=True)
+ merge_list_button = gr.Button(i18n("合并文件夹与List"), variant="primary")
+ with gr.Column(scale=2):
+ list_preview = gr.Textbox("", lines=20, max_lines=30, label=i18n("合并后的List"))
+
+ scan_list_button.click(scan_list_folders, [scan_list_folder], [first_list_folder, second_list_folder])
+ merge_list_button.click(preview_merged_list, [first_list_folder, second_list_folder, merge_list_character_name, scan_list_folder], [list_preview])
+ save_folder.change(lambda x:gr.Textbox(value=x), [save_folder], [scan_list_folder])
+ scan_list_folder.change(lambda x:gr.Textbox(value=x), [scan_list_folder], [save_folder])
+ scan_button.click(get_srt_and_audio_files, [input_folder], [srt_files_list, audio_files_list])
+ merge_zero_interval.change(lambda x: gr.update(visible=x), [merge_zero_interval],[short_interval])
+ srt_files_list.change(change_srt_file, [input_folder, srt_files_list], [audio_files_list])
+ srt_read_button.click(
+ load_from_dropdown,
+ [input_folder, srt_files_list, audio_files_list],
+ [input_text, input_audio],
+ )
+ input_text.change(
+ disable_gr_elements,
+ [slice_audio_button, filter_button],
+ [slice_audio_button, filter_button],
+ ).then(
+ change_character_name,
+ [input_audio],
+ [character],
+ )
+
+ upload_audio.change(
+ change_character_name,
+ [upload_audio],
+ [character],
+ ).then(
+ lambda x:gr.Textbox(value=x),
+ [upload_audio],
+ [input_audio],
+ )
+
+ merge_button.click(
+ merge_srt,
+ [
+ input_text,
+ output_text,
+ short_interval,
+ max_interval,
+ max_text_length,
+ add_period,
+ merge_zero_interval
+ ],
+ [output_text],
+ ).then(
+ enable_gr_elements,
+ [slice_audio_button, filter_button],
+ [slice_audio_button, filter_button],
+ )
+ slice_audio_button.click(
+ slice_audio,
+ [
+ input_audio,
+ save_folder,
+ audio_format,
+ output_text,
+ pre_preserve_time,
+ post_preserve_time,
+ pre_silence_time,
+ post_silence_time,
+ language,
+ character
+ ],
+
+ )
+ save_srt_button.click(
+ save_srt_to_file,
+ [output_text, save_folder, character],
+
+ )
+ filter_button.click(
+ filter_srt,
+ [output_text, min_length, filter_english, filter_words],
+ [output_text],
+ )
+ character.change(
+ check_character_foldfer,
+ [save_folder, character],
+ [character_warning],
+ )
+if gr.__version__.split(".")[0] == "4":
+ app.launch(
+ server_name="0.0.0.0",
+ inbrowser=True,
+ share=is_share,
+ server_port=port,
+ quiet=True,
+ )
+else:
+ app.queue(concurrency_count=511, max_size=1022).launch(
+ server_name="0.0.0.0",
+ inbrowser=True,
+ share=is_share,
+ server_port=port,
+ quiet=True,
+ )