Update app.py

app.py

@@ -7,21 +7,21 @@ import sys
 from datetime import datetime
 from concurrent.futures import ThreadPoolExecutor, as_completed
 
-# 初始化环境路径
+# Initialize environment paths
 BASE_DIR = os.path.dirname(os.path.abspath(__file__))
 PAPERS_DIR = os.path.join(BASE_DIR, "papers")
 CONFIG_PATH = os.path.join(BASE_DIR, "config.yaml")
 OUTPUT_DIR = os.path.join(BASE_DIR, "mineru_outputs")
-ZIP_OUTPUT_PATH = os.path.join(BASE_DIR, "mineru_results.zip") # 压缩包路径
+ZIP_OUTPUT_PATH = os.path.join(BASE_DIR, "mineru_results.zip") # Zip path
 
 os.makedirs(PAPERS_DIR, exist_ok=True)
 
 def get_debug_info():
-    """读取服务器文件系统状态"""
+    """Read server file system status"""
     now = datetime.now().strftime("%H:%M:%S")
     files = os.listdir(PAPERS_DIR) if os.path.exists(PAPERS_DIR) else "Directory missing"
     
-    # 递归检查输出目录下的内容
+    # Recursively check output directory contents
     output_detail = "Not generated"
     if os.path.exists(OUTPUT_DIR):
         all_output_items = []
@@ -30,55 +30,54 @@ def get_debug_info():
                 all_output_items.append(os.path.join(os.path.relpath(root, OUTPUT_DIR), name))
         output_detail = f"Found {len(all_output_items)} files: {all_output_items[:5]}..." if all_output_items else "Directory exists but is EMPTY"
 
-    return f"[{now}] 📁 papers/ 内容: {files}\n\n[{now}] 📂 mineru_outputs 状态: {output_detail}"
+    return f"[{now}] 📁 papers/ Content: {files}\n\n[{now}] 📂 mineru_outputs Status: {output_detail}"
 
 def save_pdf(file):
-    if file is None: return "❌ 请先选择 PDF", get_debug_info()
+    if file is None: return "❌ Please select a PDF first", get_debug_info()
     try:
         file_path = os.path.join(PAPERS_DIR, os.path.basename(file.name))
         shutil.copy(file.name, file_path)
-        return f"✅ 已保存: {os.path.basename(file.name)}", get_debug_info()
+        return f"✅ Saved: {os.path.basename(file.name)}", get_debug_info()
     except Exception as e:
-        return f"❌ 出错: {str(e)}", get_debug_info()
+        return f"❌ Error: {str(e)}", get_debug_info()
 
 def save_api_settings(api_key, api_base_url=None):
-    if not api_key: return "❌ Key 不能为空", get_debug_info()
+    if not api_key: return "❌ Key cannot be empty", get_debug_info()
     try:
         config = {}
         if os.path.exists(CONFIG_PATH):
             with open(CONFIG_PATH, "r", encoding="utf-8") as f:
                 config = yaml.safe_load(f) or {}
                 
-        # 保存 API Key
+        # Save API Key
         config.setdefault("api_keys", {})["gemini_api_key"] = api_key
         
-        # 如果 api_base_url 不为空（不是 None 且不是空字符串），则保存该值
+        # Save base URL if not empty
         if api_base_url:
             config["api_base_url"] = api_base_url
             
-        # 写入 YAML 文件
+        # Write to YAML
         with open(CONFIG_PATH, "w", encoding="utf-8") as f:
             yaml.dump(config, f, allow_unicode=True)
             
-        # 动态生成成功提示语
-        success_msg = "✅ Key 已保存"
+        success_msg = "✅ Key saved"
         if api_base_url:
-            success_msg += "，Base URL 已更新"
+            success_msg += ", Base URL updated"
             
         return success_msg, get_debug_info()
     except Exception as e:
-        return f"❌ 出错: {str(e)}", get_debug_info()
+        return f"❌ Error: {str(e)}", get_debug_info()
 
 
 def run_mineru_parsing_and_dag_gen():
-    """执行 PDF 解析并捕获完整日志，随后执行DAG生成流程（支持实时流式输出）"""
+    """Execute PDF parsing and DAG generation (supports real-time streaming)"""
     if not os.path.exists(PAPERS_DIR) or not any(f.endswith('.pdf') for f in os.listdir(PAPERS_DIR)):
-        yield "❌ 未发现 PDF 文件", get_debug_info(), "No execution logs."
+        yield "❌ No PDF file found", get_debug_info(), "No execution logs."
         return
 
     full_log = ""
     try:
-        # ================= 第一步：执行 Mineru 解析 =================
+        # ================= Step 1: Mineru Parsing =================
         env = os.environ.copy()
         env["MINERU_FORMULA_ENABLE"] = "false"
         env["MINERU_TABLE_ENABLE"] = "false"
@@ -87,38 +86,37 @@ def run_mineru_parsing_and_dag_gen():
 
         command_mineru = ["mineru", "-p", PAPERS_DIR, "-o", OUTPUT_DIR]
         
-        full_log += "--- Mineru 执行中 ---\n"
-        yield "⏳ 正在执行 Mineru 解析...", get_debug_info(), full_log
+        full_log += "--- Mineru Executing ---\n"
+        yield "⏳ Executing Mineru parsing...", get_debug_info(), full_log
 
-        # 1. 使用 Popen 替代 run，开启实时流
+        # 1. Use Popen for real-time streaming
         process_mineru = subprocess.Popen(
             command_mineru, 
             env=env, 
             stdout=subprocess.PIPE, 
-            stderr=subprocess.STDOUT, # 将 stderr 错误流合并到 stdout 一起输出
+            stderr=subprocess.STDOUT, 
             text=True,
-            bufsize=1 # 开启行缓冲
+            bufsize=1 
         )
 
-        # 2. 逐行读取输出并实时 yield 给 Gradio 界面
+        # 2. Read output line by line and yield to Gradio
         for line in iter(process_mineru.stdout.readline, ''):
             full_log += line
-            yield "⏳ 正在执行 Mineru 解析...", get_debug_info(), full_log
+            yield "⏳ Executing Mineru parsing...", get_debug_info(), full_log
 
         process_mineru.stdout.close()
         returncode_mineru = process_mineru.wait()
 
-        # 如果解析失败，直接 yield 返回
         if returncode_mineru != 0:
-            status = f"❌ Mineru 解析失败 (Exit Code: {returncode_mineru})"
+            status = f"❌ Mineru parsing failed (Exit Code: {returncode_mineru})"
             yield status, get_debug_info(), full_log
             return
 
-        # ================= 第二步：执行 DAG 生成 =================
+        # ================= Step 2: DAG Generation =================
         command_dag = [sys.executable, "gen_dag.py"]
         
-        full_log += "\n--- DAG Gen 执行中 ---\n"
-        yield "⏳ Mineru 解析完成，正在执行 DAG 生成...", get_debug_info(), full_log
+        full_log += "\n--- DAG Gen Executing ---\n"
+        yield "⏳ Mineru parsing complete, executing DAG generation...", get_debug_info(), full_log
         
         process_dag = subprocess.Popen(
             command_dag, 
@@ -130,31 +128,30 @@ def run_mineru_parsing_and_dag_gen():
 
         for line in iter(process_dag.stdout.readline, ''):
             full_log += line
-            yield "⏳ 正在执行 DAG 生成...", get_debug_info(), full_log
+            yield "⏳ Executing DAG generation...", get_debug_info(), full_log
 
         process_dag.stdout.close()
         returncode_dag = process_dag.wait()
         
         if returncode_dag == 0:
-            status = "✅ PDF解析与DAG生成全部完成"
+            status = "✅ PDF parsing & DAG generation fully completed"
         else:
-            status = f"❌ DAG生成失败 (Exit Code: {returncode_dag})"
+            status = f"❌ DAG generation failed (Exit Code: {returncode_dag})"
             
         yield status, get_debug_info(), full_log
 
     except Exception as e:
-        error_log = full_log + f"\n[全局异常] Exception occurred:\n{str(e)}"
-        yield "❌ 运行异常", get_debug_info(), error_log
+        error_log = full_log + f"\n[Global Exception] Exception occurred:\n{str(e)}"
+        yield "❌ Execution Exception", get_debug_info(), error_log
 
 def run_final_generation(task_type="all"):
     """
-    执行对应的生成脚本并压缩结果（支持并行执行）
-    task_type 支持: 'ppt', 'poster', 'pr', 'all'
+    Execute generation scripts and zip results
+    task_type supports: 'ppt', 'poster', 'pr', 'all'
     """
     if not os.path.exists(OUTPUT_DIR):
-        return "❌ 请先执行第二步解析", get_debug_info(), "No output folder found.", None
+        return "❌ Please run the parsing step first", get_debug_info(), "No output folder found.", None
 
-    # 根据传入的 task_type 决定要运行哪些脚本
     scripts_to_run = []
     if task_type == "ppt":
         scripts_to_run = ["gen_ppt.py"]
@@ -165,78 +162,102 @@ def run_final_generation(task_type="all"):
     elif task_type == "all":
         scripts_to_run = ["gen_ppt.py", "gen_poster.py", "gen_pr.py"]
     else:
-        return "❌ 未知任务类型", get_debug_info(), "Invalid task_type.", None
+        return "❌ Unknown task type", get_debug_info(), "Invalid task_type.", None
 
-    full_log = f"🚀 准备启动 {len(scripts_to_run)} 个任务...\n"
+    full_log = f"🚀 Preparing to start {len(scripts_to_run)} tasks...\n"
     success = True
 
-    # 定义单个脚本的执行包装器
     def execute_script(script):
         command = [sys.executable, script]
         result = subprocess.run(
             command,
             capture_output=True,
             text=True,
-            timeout=600 # 每个脚本独立的超时时间
+            timeout=600 
         )
         return script, result
 
     try:
-        # 使用 ThreadPoolExecutor 并行执行脚本
         with ThreadPoolExecutor(max_workers=len(scripts_to_run)) as executor:
-            # 提交所有任务
             future_to_script = {executor.submit(execute_script, s): s for s in scripts_to_run}
             
-            # as_completed 会在某个任务完成时立刻生成结果
             for future in as_completed(future_to_script):
                 script_name = future_to_script[future]
                 try:
-                    # 获取该任务的执行结果
                     _, result = future.result()
                     
-                    full_log += f"\n================ ✅ 执行完成: {script_name} ================\n"
+                    full_log += f"\n================ ✅ Execution Complete: {script_name} ================\n"
                     full_log += f"--- STDOUT ---\n{result.stdout}\n\n--- STDERR ---\n{result.stderr}\n"
                     
-                    # 检查此任务是否失败
                     if result.returncode != 0:
                         success = False
-                        full_log += f"❌ [错误] {script_name} 返回非零退出码 (Exit Code: {result.returncode})\n"
+                        full_log += f"❌ [Error] {script_name} returned non-zero exit code (Exit Code: {result.returncode})\n"
                 
                 except subprocess.TimeoutExpired as e:
                     success = False
-                    full_log += f"\n================ ❌ 任务超时: {script_name} ================\n{str(e)}\n"
+                    full_log += f"\n================ ❌ Task Timeout: {script_name} ================\n{str(e)}\n"
                 except Exception as e:
                     success = False
-                    full_log += f"\n================ ❌ 任务异常: {script_name} ================\n{str(e)}\n"
+                    full_log += f"\n================ ❌ Task Exception: {script_name} ================\n{str(e)}\n"
 
-        # 如果有任何一个脚本执行失败，直接返回，不打包压缩
         if not success:
-            return f"❌ {task_type.upper()} 包含失败任务，请检查日志", get_debug_info(), full_log, None
+            return f"❌ {task_type.upper()} contains failed tasks, please check logs", get_debug_info(), full_log, None
 
-        # 所有脚本都运行成功后，压缩 mineru_outputs 文件夹
+        # Zip the mineru_outputs folder
         zip_base_name = ZIP_OUTPUT_PATH.replace(".zip", "")
         shutil.make_archive(zip_base_name, 'zip', OUTPUT_DIR)
 
-        success_msg = f"✅ {task_type.upper()} 生成并压缩完成"
+        success_msg = f"✅ {task_type.upper()} generated and zipped successfully"
         return success_msg, get_debug_info(), full_log, ZIP_OUTPUT_PATH
 
     except Exception as e:
-        error_log = full_log + f"\n[全局异常] Exception occurred:\n{str(e)}"
-        return "❌ 最终生成发生全局异常", get_debug_info(), error_log, None
+        error_log = full_log + f"\n[Global Exception] Exception occurred:\n{str(e)}"
+        return "❌ Global exception during final generation", get_debug_info(), error_log, None
 
 # ==========================================
-# --- 🚀 全新美化的 UI (Hugging Face 优化版) ---
+# --- 🚀 UI Configuration (Left-Right Layout) ---
 # ==========================================
 
-# 自定义 CSS：让终端日志看起来像真正的 Terminal
 custom_css = """
+/* Prominent Title Animation from Reference */
+#main-title {
+    text-align: center !important;
+    padding: 1rem 0 0.5rem 0;
+}
+#main-title h1 {
+    font-size: 2.6em !important;
+    font-weight: 800 !important;
+    background: linear-gradient(135deg, #4f46e5 0%, #818cf8 50%, #4338ca 100%);
+    background-size: 200% 200%;
+    -webkit-background-clip: text;
+    -webkit-text-fill-color: transparent;
+    background-clip: text;
+    animation: gradient-shift 4s ease infinite;
+    letter-spacing: -0.02em;
+}
+@keyframes gradient-shift {
+    0%, 100% { background-position: 0% 50%; }
+    50% { background-position: 100% 50%; }
+}
+#subtitle {
+    text-align: center !important;
+    margin-bottom: 1.5rem;
+}
+#subtitle p {
+    margin: 0 auto;
+    color: #666;
+    font-size: 1.1rem;
+    font-weight: 500;
+}
+
+/* Terminal Log Style */
 .log-box textarea {
     font-family: 'Courier New', Consolas, monospace !important; 
     font-size: 13px !important; 
     background-color: #1e1e1e !important; 
     color: #4AF626 !important; 
 }
-/* 弱化状态框的边框感，使其更像文字标签 */
+/* Borderless Status Text */
 .status-text textarea {
     background-color: transparent !important;
     border: none !important;
@@ -246,97 +267,93 @@ custom_css = """
 """
 
 with gr.Blocks(theme=gr.themes.Soft(primary_hue="indigo"), css=custom_css) as demo:
-    gr.Markdown("# 📑 PaperX / Mineru 智能解析平台")
-    gr.Markdown("将学术 PDF 一键解析、DAG 结构化并生成多模态产物。")
+    gr.Markdown("# **PaperX / Mineru Parsing Platform**", elem_id="main-title")
+    gr.Markdown("One-click parsing of academic PDFs, DAG structuring, and multi-modal asset generation.", elem_id="subtitle")
     
-    # 1. 隐藏式全局配置
-    with gr.Accordion("⚙️ 1. 全局 API 配置", open=False):
-        with gr.Row():
-            key_input = gr.Textbox(label="API Key", type="password", placeholder="sk-...", scale=1)
-            api_base_url_input = gr.Textbox(label="Base URL (可选)", placeholder="https://api.example.com", scale=1)
-        key_btn = gr.Button("💾 保存 API 配置")
-
-    # 2. 核心上传与解析
-    with gr.Group():
-        gr.Markdown("### 📄 2. 文档解析")
-        # 拖拽自动上传
-        pdf_input = gr.File(label="拖拽或点击上传 PDF", file_types=[".pdf"])
-        
-        # 解析按钮独占一行，成为视觉焦点
-        parse_btn = gr.Button("🚀 ���始执行 Mineru & DAG 抽取", variant="primary", size="lg")
-        
-        # 弱化状态框视觉
-        parse_status = gr.Textbox(
-            show_label=False, 
-            placeholder="等待上传文档...", 
-            lines=1, 
-            interactive=False, 
-            elem_classes="status-text"
-        )
+    with gr.Row():
+        # ================= LEFT COLUMN: SETTINGS & ACTIONS =================
+        with gr.Column(scale=1):
+            
+            # 1. API Configuration
+            with gr.Accordion("⚙️ 1. Global API Configuration", open=False):
+                with gr.Row():
+                    key_input = gr.Textbox(label="API Key", type="password", placeholder="sk-...", scale=1)
+                    api_base_url_input = gr.Textbox(label="Base URL (Optional)", placeholder="https://api.example.com", scale=1)
+                key_btn = gr.Button("💾 Save API Configuration")
+
+            # 2. Document Parsing
+            with gr.Group():
+                gr.Markdown("### 📄 2. Document Parsing")
+                pdf_input = gr.File(label="Drag and drop or click to upload PDF", file_types=[".pdf"])
+                
+                parse_btn = gr.Button("🚀 Start Mineru & DAG Extraction", variant="primary", size="lg")
+                
+                parse_status = gr.Textbox(
+                    show_label=False, 
+                    placeholder="Waiting for document upload...", 
+                    lines=1, 
+                    interactive=False, 
+                    elem_classes="status-text"
+                )
+
+            # 3. Asset Generation
+            with gr.Group():
+                gr.Markdown("### 🎯 3. Asset Generation")
+                gr.Markdown("Generate final formats based on DAG structure:")
+                
+                with gr.Row():
+                    gen_ppt_btn = gr.Button("📊 Generate PPT")
+                    gen_poster_btn = gr.Button("🖼️ Generate Poster")
+                    gen_pr_btn = gr.Button("📰 Generate PR Article")
+                
+                gen_all_btn = gr.Button("✨ Generate All Assets (ALL)", variant="primary")
 
-    # 3. 最终产物生成
-    with gr.Group():
-        gr.Markdown("### 🎯 3. 产物生成")
-        gr.Markdown("基于 DAG 结构生成最终所需格式：")
-        
-        # 层次化布局：上面三个单项，下面一个全部
-        with gr.Row():
-            gen_ppt_btn = gr.Button("📊 单独生成 PPT")
-            gen_poster_btn = gr.Button("🖼️ 单独生成 Poster")
-            gen_pr_btn = gr.Button("📰 单独生成 PR 文章")
-        
-        gen_all_btn = gr.Button("✨ 一键生成全部 (ALL)", variant="primary")
-
-    # 4. 生成结果与下载 (从 Tabs 中独立出来，紧跟工作流)
-    with gr.Group():
-        gr.Markdown("### 📦 4. 生成结果 & 下载")
-        gen_status = gr.Textbox(
-            show_label=False, 
-            placeholder="当前暂无生成任务...", 
-            lines=2, 
-            interactive=False,
-            elem_classes="status-text"
-        )
-        # 默认隐藏下载框，生成成功后再动态展示
-        download_file = gr.File(label="📥 获取最终压缩包", interactive=False, visible=False)
-
-    gr.HTML("<hr style='margin-top: 30px; margin-bottom: 30px;' />") # 添加一条视觉分割线
-
-    # 5. 开发者监控区 (置于最底部，方便后续精简 UI 时直接删除这部分)
-    gr.Markdown("### 🛠️ 开发者后台监控 (仅供调试)")
-    with gr.Tabs():
-        # Tab 1: 实时终端
-        with gr.Tab("📜 终端流 (Terminal)"):
-            cmd_logs = gr.Textbox(
-                label="Stdout / Stderr", 
-                placeholder="等待任务开始...", 
-                lines=15, 
-                interactive=False,
-                elem_classes="log-box" 
-            )
-        
-        # Tab 2: 文件系统快照
-        with gr.Tab("🔍 系统快照 (Debug)"):
-            refresh_btn = gr.Button("🔄 刷新目录树") 
-            debug_view = gr.Textbox(label="Workspace Files", lines=15, interactive=False, value=get_debug_info())
 
-    # ================= 逻辑绑定 =================
+        # ================= RIGHT COLUMN: OUTPUTS & LOGS =================
+        with gr.Column(scale=1):
+            
+            # 4. Results & Downloads
+            with gr.Group():
+                gr.Markdown("### 📦 Generation Results & Download")
+                gen_status = gr.Textbox(
+                    show_label=False, 
+                    placeholder="No generation task currently...", 
+                    lines=2, 
+                    interactive=False,
+                    elem_classes="status-text"
+                )
+                download_file = gr.File(label="📥 Get Final Zip Archive", interactive=False, visible=False)
+
+            # 5. Debugging & Terminal
+            gr.Markdown("### 🛠️ Developer Monitoring (Debug Only)")
+            with gr.Tabs():
+                with gr.Tab("📜 Terminal Stream"):
+                    cmd_logs = gr.Textbox(
+                        label="Stdout / Stderr", 
+                        placeholder="Waiting for task to start...", 
+                        lines=18, 
+                        interactive=False,
+                        elem_classes="log-box" 
+                    )
+                
+                with gr.Tab("🔍 System Snapshot"):
+                    refresh_btn = gr.Button("🔄 Refresh Directory Tree") 
+                    debug_view = gr.Textbox(label="Workspace Files", lines=17, interactive=False, value=get_debug_info())
+
+    # ================= LOGIC BINDINGS =================
     
     key_btn.click(fn=save_api_settings, inputs=[key_input, api_base_url_input], outputs=[parse_status, debug_view])
     
-    # 绑定上传和清除事件，实现自动化
     pdf_input.upload(fn=save_pdf, inputs=pdf_input, outputs=[parse_status, debug_view])
-    pdf_input.clear(fn=lambda: ("ℹ️ 已清空文件", get_debug_info()), outputs=[parse_status, debug_view])
+    pdf_input.clear(fn=lambda: ("ℹ️ File cleared", get_debug_info()), outputs=[parse_status, debug_view])
     
     parse_btn.click(
         fn=run_mineru_parsing_and_dag_gen, 
         outputs=[parse_status, debug_view, cmd_logs]
     )
 
-    # 动态控制下载组件的显示与隐藏
     def trigger_gen(task):
         status, debug, logs, file_path = run_final_generation(task)
-        # 如果 file_path 存在，显示下载组件并赋值；否则保持隐藏
         file_update = gr.update(value=file_path, visible=True) if file_path else gr.update(visible=False)
         return status, debug, logs, file_update