Upload 2 files

app.py

@@ -1,13 +1,11 @@
-from fastapi import FastAPI, HTTPException, BackgroundTasks
+from fastapi import FastAPI, HTTPException
 from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
-from typing import List, Dict
+from typing import List
 import os
 import uuid
 import aiohttp
-import asyncio
 import logging
-import tempfile
 import openai
 from pathlib import Path
 import subprocess
@@ -16,9 +14,7 @@ import ssl
 import json
 from fastapi.staticfiles import StaticFiles
 from pydub import AudioSegment
-import shlex
-from ffmpeg import probe as ffmpeg_probe  # 需要安装ffmpeg-python包
-import time
+from PIL import Image
 
 # 配置日志
 logging.basicConfig(level=logging.INFO)
@@ -66,13 +62,15 @@ async def download_image(url, output_path):
                 if response.status == 200:
                     with open(output_path, 'wb') as f:
                         f.write(await response.read())
-                    return output_path
+                    with Image.open(output_path) as img:  # 新增尺寸获取
+                        width, height = img.size
+                    return output_path, width  # 返回尺寸
                 else:
                     logger.error(f"Failed to download image: {response.status}")
-                    return None
+                    return None, 0
     except Exception as e:
         logger.error(f"Error downloading image: {e}")
-        return None
+        return None, 0
 
 # 生成语音
 async def generate_speech(text, voice="alloy", output_path=None):
@@ -112,40 +110,44 @@ PlayResX: 1920
 PlayResY: 1080
 [V4+ Styles]
 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
-Style: Caption,Noto Sans CJK SC,54,&H00FFFFFF,&H000000FF,&H00333333,&H00000000,0,0,0,0,100,100,0,0,1,2,3,2,100,100,50,0
-Style: Speech,Noto Sans CJK SC,48,&H00FFFFFF,&H000000FF,&H00333333,&H00000000,0,0,0,0,100,100,0,0,1,2,3,8,100,100,50,0
+Style: Caption,Noto Sans CJK SC,39,&H00FFFFFF,&H000000FF,&H00333333,&H00000000,0,0,0,0,100,100,0,0,1,2,3,2,10,10,39,0
+Style: Speech,Noto Sans CJK SC,39,&H00FFFFFF,&H000000FF,&H00333333,&H00000000,0,0,0,0,100,100,0,0,1,2,3,8,10,10,39,0
 [Events]
 Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
 """
-def smart_wrap(text, video_width=1920):
-    """智能换行算法"""
-    max_chars = int(video_width // 38)  # 基于典型字体尺寸计算
+def smart_wrap(text, image_width, font_size=48):
+    """动态计算每行字符数"""
+    # 根据实际字体渲染参数计算
+    avg_char_width = font_size * 0.6  # 中文字符平均宽度（像素）
+    max_chars_per_line = max(1, int(image_width / avg_char_width) - 2)  # 保留边距
+    
+    # 实现更精确的断行逻辑
     lines = []
     current_line = []
-    current_len = 0
+    current_width = 0
     
     for char in text:
-        char_len = 2 if ord(char) > 255 else 1  # 全角字符计为2单位
-        if current_len + char_len > max_chars * 2:
+        char_width = font_size if ord(char) > 255 else font_size//2
+        if current_width + char_width > image_width - 100:  # 保留100像素边距
             lines.append(''.join(current_line))
             current_line = [char]
-            current_len = char_len
+            current_width = char_width
         else:
             current_line.append(char)
-            current_len += char_len
+            current_width += char_width
     lines.append(''.join(current_line))
     
     return r'\N'.join(lines)
 
 # 创建caption字幕文件（底部显示）
-def create_caption_subtitle_file(project_dir, captions, panel_start_times, panel_durations):
+def create_caption_subtitle_file(project_dir, captions, panel_start_times, panel_durations, image_widths):
     try:
         subtitle_file = os.path.join(project_dir, "captions.ass")
 
         with open(subtitle_file, "w", encoding="utf-8") as f:
             f.write(ASS_STYLE_HEADER)
-            for i, (caption, start, duration) in enumerate(zip(captions, panel_start_times, panel_durations)):
-                wrapped_text = smart_wrap(caption)
+            for i, (caption, start, duration, width) in enumerate(zip(captions, panel_start_times, panel_durations, image_widths)):
+                wrapped_text = smart_wrap(caption, width)
                 f.write(
                     f"Dialogue: 0,{format_time(start)},{format_time(start + duration)},"
                     f"Caption,,0,0,0,,{wrapped_text}\n"
@@ -157,14 +159,14 @@ def create_caption_subtitle_file(project_dir, captions, panel_start_times, panel
         return None
 
 # 创建speech字幕文件（顶部显示）
-def create_speech_subtitle_file(project_dir, speeches, panel_start_times, panel_durations):
+def create_speech_subtitle_file(project_dir, speeches, panel_start_times, panel_durations, image_widths):
     try:
         subtitle_file = os.path.join(project_dir, "speeches.ass")
         
         with open(subtitle_file, "w", encoding="utf-8") as f:
             f.write(ASS_STYLE_HEADER)
-            for i, (speech, start, duration) in enumerate(zip(speeches, panel_start_times, panel_durations)):
-                wrapped_text = smart_wrap(speech)
+            for i, (speech, start, duration, width) in enumerate(zip(speeches, panel_start_times, panel_durations, image_widths)):
+                wrapped_text = smart_wrap(speech, width)
                 f.write(
                     f"Dialogue: 0,{format_time(start)},{format_time(start + duration)},"
                     f"Speech,,0,0,0,,{wrapped_text}\n"
@@ -271,21 +273,6 @@ async def create_audio_file(project_dir, captions, speeches):
         logger.error(traceback.format_exc())
         return None, {}, [], []
 
-def get_video_dimensions(video_path):
-    try:
-        result = subprocess.run(
-            ["ffprobe", "-v", "error", "-select_streams", "v:0",
-             "-show_entries", "stream=width,height", "-of", "json", video_path],
-            capture_output=True,
-            text=True
-        )
-        data = json.loads(result.stdout)
-        return (int(data['streams'][0]['width']), 
-                int(data['streams'][0]['height']))
-    except Exception as e:
-        logger.warning(f"Video dimension detection failed: {e}")
-        return (1920, 1080)
-
 def process_sub_path(path):
     """深度处理FFmpeg路径转义"""
     # 统一转换为POSIX路径
@@ -329,31 +316,10 @@ def create_video(project_dir, image_paths, caption_subtitle_file, speech_subtitl
         ]
         subprocess.run(cmd1, check=True)
 
-        # 获取视频尺寸（使用改进后的方法）
-        video_width, video_height = get_video_dimensions(temp_video)
-        base_fontsize = max(24, video_width // 50)
         # 构建滤镜链
         combined_filter = (
-            f"subtitles={process_sub_path(caption_subtitle_file)}:"
-            "force_style='"
-            #"FontName=Noto Sans CJK SC,"
-            "Fontsize={},"
-            "Alignment=2,"
-            "MarginV={},"
-            "'".format(
-                int(base_fontsize*0.6), 
-                video_height//100
-            ),
-            f"subtitles={process_sub_path(speech_subtitle_file)}:"
-            "force_style='"
-            #"FontName=Noto Sans CJK SC,"
-            "Fontsize={},"
-            "Alignment=8,"
-            "MarginV={},"
-            "'".format(
-                int(base_fontsize*0.5), 
-                video_height//10
-            )
+            f"subtitles={process_sub_path(caption_subtitle_file)}",
+            f"subtitles={process_sub_path(speech_subtitle_file)}"
         )
         filter_chain = ",".join(combined_filter)
         # 优化ffmpeg命令
@@ -367,10 +333,7 @@ def create_video(project_dir, image_paths, caption_subtitle_file, speech_subtitl
             "-movflags", "+faststart",
             output_video
         ]
-        # 添加执行计时
-        start_time = time.time()
         subprocess.run(cmd_combined, check=True)
-        logger.info(f"Video processed in {time.time()-start_time:.2f}s")
         # 清理临时文件
         os.remove(temp_video)
         return output_video
@@ -409,7 +372,7 @@ def upload_to_local_storage(local_path, relative_path):
         return None
 
 @app.post("/api/generate-video")
-async def generate_video(comic_data: ComicData, background_tasks: BackgroundTasks):
+async def generate_video(comic_data: ComicData):
     # 创建唯一项目ID
     project_id = str(uuid.uuid4())
     # 使用绝对路径创建项目目录
@@ -421,11 +384,15 @@ async def generate_video(comic_data: ComicData, background_tasks: BackgroundTask
     try:
         # 下载图片
         image_paths = []
+        image_widths = []
         for i, panel_url in enumerate(comic_data.panels):
             output_path = os.path.join(project_dir, f"panel_{i}.jpg")
-            result = await download_image(panel_url, output_path)
-            if result:
-                image_paths.append(result)
+            path_result, img_width = await download_image(panel_url, output_path)
+            if path_result:
+                image_paths.append(path_result)
+                image_widths.append(img_width)
+            else:
+                image_widths.append(1920)  # 失败时使用默认宽度
         
         if not image_paths:
             raise HTTPException(status_code=500, detail="Failed to download images")
@@ -443,13 +410,13 @@ async def generate_video(comic_data: ComicData, background_tasks: BackgroundTask
         
         # 创建字幕文件 - 分别为caption和speech创建
         caption_subtitle_file = create_caption_subtitle_file(
-            project_dir, comic_data.captions, panel_start_times, panel_durations
+            project_dir, comic_data.captions, panel_start_times, panel_durations, image_widths
         )
         if not caption_subtitle_file:
             raise HTTPException(status_code=500, detail="Failed to create caption subtitle file")
         
         speech_subtitle_file = create_speech_subtitle_file(
-            project_dir, comic_data.speeches, panel_start_times, panel_durations
+            project_dir, comic_data.speeches, panel_start_times, panel_durations, image_widths
         )
         if not speech_subtitle_file:
             raise HTTPException(status_code=500, detail="Failed to create speech subtitle file")
@@ -480,8 +447,7 @@ async def generate_video(comic_data: ComicData, background_tasks: BackgroundTask
 #            if img_url:
 #                image_urls.append(img_url)
         
-        # 后台任务清理临时文件
-        # background_tasks.add_task(lambda: shutil.rmtree(project_dir, ignore_errors=True))
+        # 清理临时文件
         shutil.rmtree(project_dir, ignore_errors=True)
         
         return {

requirements.txt

@@ -4,5 +4,4 @@ aiohttp>=3.8.4
 openai>=1.2.0
 python-multipart>=0.0.6
 pydantic>=1.10.7
-pydub
-ffmpeg-python
+pydub