Add more controls to video/sub synthesis

app/common/config.py

@@ -205,6 +205,11 @@ class Config(QConfig):
         0,
         RangeValidator(-5000, 5000)
     )
+
+    vertical_offset = RangeConfigItem(
+        "Subtitle", "VerticalOffset",
+        0, RangeValidator(-500, 500)
+    )
 
     # ------------------- 软件页面配置 -------------------
     micaEnabled = ConfigItem("MainWindow", "MicaEnabled", False, BoolValidator())

app/common/enums.py

@@ -1,9 +1,13 @@
 # Set the enums to new translated values
 from PyQt5.QtCore import QObject
-from ..core.entities import SubtitleLayoutEnum, InternetTranslateEnum, TodoWhenDoneEnum, Task
+from ..core.entities import SubtitleLayoutEnum, InternetTranslateEnum, TodoWhenDoneEnum, Task, BatchTaskTypeEnum
 
 def Enums_Translate():
     qoEnums = QObject()
+    BatchTaskTypeEnum.TRANSCRIBE._value_ = qoEnums.tr("Create Subtitle from Audio/Video")
+    BatchTaskTypeEnum.SOFT._value_ = qoEnums.tr("Create Soft Subtitle Video")
+    BatchTaskTypeEnum.HARD._value_ = qoEnums.tr("Create Hard Subtitle Video")
+
     SubtitleLayoutEnum.ONLY_ORIGINAL._value_ = qoEnums.tr("Original Only")
     SubtitleLayoutEnum.ONLY_TRANSLATE._value_ = qoEnums.tr("Translated Only")
     SubtitleLayoutEnum.ORIGINAL_ON_TOP._value_ = qoEnums.tr("Original on Top")
@@ -34,8 +38,8 @@ def Enums_Translate():
     Task.Source.FILE_IMPORT._value_ = qoEnums.tr("File Import")
     Task.Source.URL_IMPORT._value_ = qoEnums.tr("URL Import")
 
-    Task.Type.OPTIMIZE._value_ = qoEnums.tr("Optimize")
-    Task.Type.SUBTITLE._value_ = qoEnums.tr("Subtitle")
-    Task.Type.SYNTHESIS._value_ = qoEnums.tr("Synthesis")
-    Task.Type.TRANSCRIBE._value_ = qoEnums.tr("Transcribe")
-    Task.Type.URL._value_ = qoEnums.tr("URL Import")
+    Task.Type.OPTIMIZE._value_ = qoEnums.tr("Optimize + Translate Subtitles")
+    Task.Type.SUBTITLE._value_ = qoEnums.tr("Add Subtitle To Video")
+    Task.Type.SYNTHESIS._value_ = qoEnums.tr("Combine Subtitle with Video")
+    Task.Type.TRANSCRIBE._value_ = qoEnums.tr("Get Subtitle From Video/Audio")
+    Task.Type.URL._value_ = qoEnums.tr("Download Video from URL then Add Subtitle")

app/common/signal_bus.py

@@ -14,6 +14,9 @@ class SignalBus(QObject):
     # 使用网络翻译
     internet_translation_changed = pyqtSignal(bool)
     internet_translation_method_changed = pyqtSignal(bool)
+    # 
+    need_video_changed = pyqtSignal(bool)
+    soft_subtitle_changed = pyqtSignal(bool)
 
     # 新增视频控制相关信号
     video_play = pyqtSignal()  # 播放信号
@@ -23,6 +26,12 @@ class SignalBus(QObject):
     video_segment_play = pyqtSignal(int, int)  # 播放片段信号，参数为开始和结束时间(ms)
     video_subtitle_added = pyqtSignal(str)  # 添加字幕文件信号
 
+    def on_need_video_changed(self, needVideo: bool):
+        self.need_video_changed.emit(needVideo)
+
+    def on_soft_subtitle_changed(self, softSubtitle: bool):
+        self.soft_subtitle_changed.emit(softSubtitle)
+
     def on_subtitle_layout_changed(self, layout: str):
         self.subtitle_layout_changed.emit(layout)
 

app/core/entities.py

@@ -6,6 +6,12 @@
 from typing import Optional
 from PyQt5.QtCore import QObject
 
+class BatchTaskTypeEnum(Enum):
+    """ 批量任务类型 """
+    TRANSCRIBE = "Create Subtitle from Audio/Video"
+    SOFT = "Create Soft Subtitle Video"
+    HARD = "Create Hard Subtitle Video"
+
 class SubtitleLayoutEnum(Enum):
     """ 字幕布局 """
     ONLY_ORIGINAL = "Original Only"
@@ -19,6 +25,15 @@ class InternetTranslateEnum(Enum):
     GOOGLE = "Google Translate"
 
 
+class SupportedImageFormats(Enum):
+    """ 支持的图片格式 """
+    JPG = "jpg"
+    PNG = "png"
+    BMP = "bmp"
+    GIF = "gif"
+    WEBP = "webp"
+
+
 class SupportedAudioFormats(Enum):
     """ 支持的音频格式 """
     AAC = "aac"
@@ -492,11 +507,11 @@ class Source(Enum):
 
     class Type(Enum):
         # 任务类型：transcribe or generate subtitle
-        TRANSCRIBE = "Transcription"
-        SUBTITLE = "Subtitle"
-        OPTIMIZE = "Optimization"
-        SYNTHESIS = "Synthesis"
-        URL = "URL"
+        TRANSCRIBE = "Get Subtitle From Video/Audio"
+        SUBTITLE = "Add Subtitle To Video"
+        OPTIMIZE = "Optimize + Translate Subtitles"
+        SYNTHESIS = "Combine Subtitle with Video"
+        URL = "Download Video from URL then Add Subtitle"
 
     # 任务信息
     id: int = field(default_factory=lambda: randint(0, 100_000_000))
@@ -522,7 +537,6 @@ class Type(Enum):
 
     # 转录（转录模型）
     transcribe_model: Optional[TranscribeModelEnum] = TranscribeModelEnum.JIANYING
-
     transcribe_language: Optional[TranscribeLanguageEnum] = LANGUAGES[TranscribeLanguageEnum.ENGLISH.value]
     use_asr_cache: bool = True
     need_word_time_stamp: bool = False
@@ -559,10 +573,16 @@ class Type(Enum):
     subtitle_layout: Optional[str] = None
     max_word_count_cjk: int = 12
     max_word_count_english: int = 18
-    need_split: bool = True
+    need_split: bool = False
 
     # 视频生成
     need_video: bool = True
     video_save_path: Optional[str] = None
     soft_subtitle: bool = True
     subtitle_style_srt: Optional[str] = None
+    portrait: bool = False
+    portrait_background: Optional[str] = None
+    zoom_video: int = 100
+    zoom_subtitle: int = 100
+    vertical_offset: int = 0
+

app/core/subtitle_processor/optimizer.py

@@ -274,6 +274,8 @@ def translate_single_batch(self, original_subtitle: Dict[int,str], callback = No
             return_text = response.choices[0].message.content
             # logger.info(f"response:{type(return_text)}")
             previous_translation = return_text
+
+            logger.info(f"{key}. Original: {value}\n{key}. Translated: {return_text}")
 
             line = {str(key): return_text}  # Create a dictionary with key and translated text
 

app/core/thread/create_task_thread.py

@@ -21,31 +21,28 @@ class CreateTaskThread(QThread):
     progress = pyqtSignal(int, str)
     error = pyqtSignal(str)
 
-    def __init__(self, file_path, task_type: Task.Type):
+    def __init__(self, file_path, task_type: Task.Type, soft_sub: bool):
         super().__init__()
         self.file_path = file_path
         self.task_type = task_type
+        self.soft_sub = soft_sub
 
     def run(self):
         try:
             if self.task_type == Task.Type.SUBTITLE:
-                self.create_file_task(self.file_path)
+                self.create_file_task(self.file_path, self.soft_sub)
             elif self.task_type == Task.Type.URL:
-                self.create_url_task(self.file_path)
+                self.create_url_task(self.file_path, self.soft_sub)
             elif self.task_type == Task.Type.TRANSCRIBE:
                 self.create_transcription_task(self.file_path)
-            elif self.task_type == Task.Type.OPTIMIZE:
-                self.create_subtitle_optimization_task()
-            elif self.task_type == Task.Type.SYNTHESIS:
-                self.create_video_synthesis_task()
             else:
                 raise ValueError("No matching task type.")
         except Exception as e:
             logger.exception("创建任务失败: %s", str(e))
             self.progress.emit(0, self.tr("创建任务失败"))
             self.error.emit(str(e))
 
-    def create_file_task(self, file_path):
+    def create_file_task(self, file_path, soft_sub: bool):
         logger.info("\n===================")
         logger.info(f"开始创建文件任务：{file_path}")
         # 使用 Path 对象处理路径
@@ -91,7 +88,8 @@ def create_file_task(self, file_path):
         if cfg.subtitle_output_format.value.value == "ass":
             ass_style_name = cfg.subtitle_style_name.value
             ass_style_path = SUBTITLE_STYLE_PATH / f"{ass_style_name}.txt"
-            subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
+            if ass_style_path.exists():
+                subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
         else:
             subtitle_style_srt = None
 
@@ -149,16 +147,17 @@ def create_file_task(self, file_path):
             result_subtitle_save_path=str(result_subtitle_save_path),
             subtitle_layout=cfg.subtitle_layout.value,
             video_save_path=str(video_save_path),
-            soft_subtitle=cfg.soft_subtitle.value,
+            soft_subtitle=soft_sub,
             subtitle_style_srt=subtitle_style_srt,
             need_video=cfg.need_video.value,
+            vertical_offset=cfg.vertical_offset.value,
             type=Task.Type.SUBTITLE,
         )
         self.finished.emit(task)
         self.progress.emit(100, self.tr("创建任务完成"))
         logger.info(f"文件任务创建完成：{task}")
 
-    def create_url_task(self, url, task_type):
+    def create_url_task(self, url, soft_sub: bool):
         logger.info("\n===================")
         logger.info(f"开始创建URL任务：{url}")
         self.progress.emit(5, self.tr("正在获取视频信息"))
@@ -213,7 +212,8 @@ def create_url_task(self, url, task_type):
         if cfg.subtitle_output_format.value.value == "ass" and ass_style_path.exists():
             ass_style_name = cfg.subtitle_style_name.value
             ass_style_path = SUBTITLE_STYLE_PATH / f"{ass_style_name}.txt"
-            subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
+            if ass_style_path.exists():
+                subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
         else:
             subtitle_style_srt = None
 
@@ -266,9 +266,10 @@ def create_url_task(self, url, task_type):
             result_subtitle_save_path=str(result_subtitle_save_path),
             subtitle_layout=cfg.subtitle_layout.value,
             video_save_path=str(video_save_path),
-            soft_subtitle=cfg.soft_subtitle.value,
+            soft_subtitle=soft_sub,
             subtitle_style_srt=subtitle_style_srt,
             need_video=cfg.need_video.value,
+            vertical_offset=cfg.vertical_offset.value,
             task=Task.Type.SUBTITLE,
         )
         self.finished.emit(task)
@@ -307,10 +308,11 @@ def create_transcription_task(self, file_path):
         original_subtitle_save_path = task_work_dir / f"【原始字幕】{file_name}-{cfg.transcribe_model.value.value}-{whisper_type}.srt"
         result_subtitle_save_path = file_full_path.parent / ( cfg.subtitle_file_prefix.value + file_name + cfg.subtitle_file_suffix.value + "." + cfg.subtitle_output_format.value.value )
 
-        if cfg.subtitle_output_format.value.value == "ass" and ass_style_path.exists():
+        if cfg.subtitle_output_format.value.value == "ass":
             ass_style_name = cfg.subtitle_style_name.value
             ass_style_path = SUBTITLE_STYLE_PATH / f"{ass_style_name}.txt"
-            subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
+            if ass_style_path.exists():
+                subtitle_style_srt = ass_style_path.read_text(encoding="utf-8")
         else:
             subtitle_style_srt = None
 
@@ -414,7 +416,7 @@ def create_subtitle_optimization_task(file_path):
         logger.info(f"字幕优化任务创建完成：{task}")
         return task
 
-    def create_video_synthesis_task(subtitle_file, video_file):
+    def create_video_synthesis_task(subtitle_file, video_file, soft_sub: bool):
         logger.info(f"开始创建视频合成任务：{subtitle_file} {video_file}")
         subtitle_file = Path(subtitle_file.strip()).as_posix()
         video_file = Path(video_file.strip()).as_posix()
@@ -430,7 +432,7 @@ def create_video_synthesis_task(subtitle_file, video_file):
             status=Task.Status.GENERATING,
             work_dir=str(task_work_dir),
             file_path=str(Path(video_file)),
-            result_subtitle_save_path=str(Path(subtitle_file)),
+            original_subtitle_save_path=str(Path(subtitle_file)),
             video_save_path=str(video_save_path),
             soft_subtitle=cfg.soft_subtitle.value,
             type=Task.Type.SYNTHESIS,
@@ -500,7 +502,6 @@ def sanitize_filename(name, replacement="_"):
     # 如果文件名为空，返回一个默认名称
     if not sanitized:
         sanitized = "default_filename"
-
     return sanitized
 
 

app/core/thread/subtitle_optimization_thread.py

@@ -129,9 +129,10 @@ def run(self):
             asr_data = from_subtitle_file(str_path)
 
             # 检查是否需要合并重新断句
-            if not asr_data.is_word_timestamp() and need_split and self.task.faster_whisper_one_word:
+            is_word_split = asr_data.is_word_timestamp()
+            if not is_word_split and need_split and self.task.faster_whisper_one_word:
                 asr_data.split_to_word_segments()
-            if asr_data.is_word_timestamp():
+            if is_word_split:
                 self.progress.emit(15, self.tr("字幕断句..."))
                 logger.info("正在字幕断句...")
                 asr_data = merge_segments(asr_data, model=llm_model, 
@@ -141,6 +142,8 @@ def run(self):
                 asr_data.save(save_path=split_path)
                 self.update_all.emit(asr_data.to_json())
 
+
+
             # 制作成请求llm接口的格式 {{"1": "original_subtitle"},...}
             subtitle_json = {str(k): v["original_subtitle"] for k, v in asr_data.to_json().items()}
             self.subtitle_length = len(subtitle_json)

app/core/thread/subtitle_pipeline_thread.py

@@ -65,15 +65,16 @@ def handle_error(error_msg):
 
             # 3. 视频合成
             # self.task.status = Task.Status.GENERATING
-            self.progress.emit(80, self.tr("开始合成视频"))
-            synthesis_thread = VideoSynthesisThread(self.task)
-            synthesis_thread.progress.connect(lambda value, msg: self.progress.emit(int(70 + value * 0.3), msg))
-            synthesis_thread.error.connect(handle_error)
-            synthesis_thread.run()
+            if self.task.need_video:
+                self.progress.emit(80, self.tr("开始合成视频"))
+                synthesis_thread = VideoSynthesisThread(self.task)
+                synthesis_thread.progress.connect(lambda value, msg: self.progress.emit(int(70 + value * 0.3), msg))
+                synthesis_thread.error.connect(handle_error)
+                synthesis_thread.run()
 
-            if self.has_error:
-                logger.info("视频合成过程中发生错误，终止流程")
-                return
+                if self.has_error:
+                    logger.info("视频合成过程中发生错误，终止流程")
+                    return
 
             self.task.status = Task.Status.COMPLETED
             logger.info("处理完成")

app/core/thread/video_synthesis_thread.py

@@ -2,8 +2,8 @@
 from pathlib import Path
 from PyQt5.QtCore import QThread, pyqtSignal
 
-from ..entities import Task
-from ..utils.video_utils import add_subtitles
+from ..entities import Task, VideoInfo
+from ..utils.video_utils import add_subtitles, get_video_info
 from ..utils.logger import setup_logger
 from ...common.config import cfg
 
@@ -37,12 +37,12 @@ def run(self):
             logger.info(f"时间：{datetime.datetime.now()}")
             self.task.status = Task.Status.SYNTHESIZING
             video_file = self.task.file_path
-            if Path(self.task.original_subtitle_save_path).is_file():
+            if self.task.original_subtitle_save_path and Path(self.task.original_subtitle_save_path).is_file():
                 # result sub exist (after optimizing)
-                subtitle_file = self.task.result_subtitle_save_path
-            elif Path(self.task.result_subtitle_save_path).is_file():
-                # No optimzing, original sub only
                 subtitle_file = self.task.original_subtitle_save_path
+            elif self.task.result_subtitle_save_path and Path(self.task.result_subtitle_save_path).is_file():
+                # No optimzing, original sub only
+                subtitle_file = self.task.result_subtitle_save_path
             else:
                 raise RuntimeError("No subtitle file available.")
 
@@ -61,7 +61,30 @@ def run(self):
             logger.info(f"开始合成视频: {video_file}")
             self.progress.emit(10, self.tr("正在合成"))
             self.progress.emit(11, f"Soft subtitle:{soft_subtitle}")
+            if not self.task.video_info:
+                video_info = get_video_info(video_file)
+                w = video_info["width"]
+                h = video_info["height"]
+                duration = int(video_info["duration_seconds"])
+            else:
+                w = self.task.video_info.width
+                h = self.task.video_info.height
+                duration = int(self.task.video_info.duration_seconds)
+
+            if self.task.portrait:
+                width = h
+                height = w
+            else:
+                width = w
+                height = h
+
             add_subtitles(video_file, subtitle_file, video_save_path, soft_subtitle=soft_subtitle,
+                          output_width=width, output_height=height, portrait=self.task.portrait,
+                          vertical_offset=self.task.vertical_offset,
+                          portrait_background=self.task.portrait_background,
+                          duration=duration,
+                          zoom_video=self.task.zoom_video,
+                          zoom_subtitle=self.task.zoom_subtitle,
                           progress_callback=self.progress_callback)
             self.progress.emit(100, self.tr("合成完成"))
             logger.info(f"视频合成完成，保存路径: {video_save_path}")