action

.gitea/workflows/deploy.yaml

@@ -0,0 +1,37 @@
+name: Deploy WebSocket Server
+
+on:
+  push:
+    branches:
+      - main  # 或者是 master，请根据您的分支名称修改
+
+jobs:
+  deploy:
+    runs-on: ubuntu  # 或者您的 Gitea runner 标签
+    steps:
+      - name: SSH Remote Commands
+        uses: appleboy/ssh-action@v1.0.0
+        with:
+          host: ${{ secrets.SERVER_HOST }}
+          username: ${{ secrets.SERVER_USERNAME }}
+          password: ${{ secrets.SERVER_PASSWORD }}
+          port: 22
+          script: |
+            # 进入项目目录
+            cd /root/V2_micropython/
+            
+            # 拉取最新代码
+            echo "Pulling latest code..."
+            git pull
+            
+            # 进入 websocket_server 目录
+            cd websocket_server
+            
+            # 重启服务
+            echo "Restarting Docker Compose services..."
+            docker compose down
+            docker compose up -d
+            
+            # 检查部署状态
+            echo "Checking deployment status..."
+            docker compose ps

display.py

@@ -211,7 +211,37 @@ class Display:
         self.tft.line(x, y + 5, x + 3, y + 8, st7789.GREEN)
         self.tft.line(x + 3, y + 8, x + 10, y, st7789.GREEN)
 
-    def render_confirm_screen(self, asr_text=""):
+    def measure_text(self, text):
+        """计算文本宽度"""
+        width = 0
+        for char in text:
+            if ord(char) > 127:
+                width += 16
+            else:
+                width += 8
+        return width
+
+    def draw_centered_text(self, text, x, y, w, h, color, bg=None):
+        """在指定区域居中显示文本"""
+        if not self.tft: return
+        text_width = self.measure_text(text)
+        start_x = x + (w - text_width) // 2
+        start_y = y + (h - 16) // 2
+        
+        # 确保不超出边界
+        start_x = max(x, start_x)
+        
+        if bg is not None:
+             self.tft.fill_rect(x, y, w, h, bg)
+        self.text(text, start_x, start_y, color)
+
+    def draw_button(self, text, x, y, w, h, bg_color, text_color=st7789.WHITE):
+        """绘制带居中文字的按钮"""
+        if not self.tft: return
+        self.tft.fill_rect(x, y, w, h, bg_color)
+        self.draw_centered_text(text, x, y, w, h, text_color)
+
+    def render_confirm_screen(self, asr_text="", waiting=False):
         """渲染确认界面"""
         if not self.tft:
             return
@@ -220,12 +250,14 @@ class Display:
         
         # Header
         self.tft.fill_rect(0, 0, 240, 30, st7789.CYAN)
-        self.text("说完了吗？", 75, 8, st7789.BLACK)
+        self.draw_centered_text("说完了吗？", 0, 0, 240, 30, st7789.BLACK)
         
         # Content box
         self.tft.fill_rect(10, 50, 220, 90, 0x4208) # DARKGREY
         
-        if asr_text:
+        if waiting:
+            self.draw_centered_text("正在识别...", 10, 50, 220, 90, st7789.YELLOW)
+        elif asr_text:
             # 自动换行逻辑
             max_width = 200
             lines = []
@@ -254,18 +286,119 @@ class Display:
             
             for i, line in enumerate(lines):
                 # 计算水平居中
-                line_width = 0
+                line_width = self.measure_text(line)
-                for c in line:
-                    line_width += 16 if ord(c) > 127 else 8
-                
                 center_x = 20 + (200 - line_width) // 2
                 self.text(line, center_x, start_y + i * 20, st7789.WHITE, wait=False)
         else:
-            self.text("未识别到文字", 70, 85, st7789.WHITE)
+            self.draw_centered_text("未识别到文字", 10, 50, 220, 90, st7789.WHITE)
         
         # Buttons
-        self.tft.fill_rect(20, 160, 90, 30, st7789.GREEN)
+        self.draw_button("短按确认", 20, 160, 90, 30, st7789.GREEN, st7789.BLACK)
-        self.text("短按确认", 30, 168, st7789.BLACK)
+        self.draw_button("长按重录", 130, 160, 90, 30, st7789.RED, st7789.WHITE)
 
-        self.tft.fill_rect(130, 160, 90, 30, st7789.RED)
+    def render_recording_screen(self, asr_text="", audio_level=0, is_recording=False):
-        self.text("长按重录", 140, 168, st7789.WHITE)
+        """渲染录音界面"""
+        if not self.tft:
+            return
+        
+        self.tft.fill(st7789.BLACK)
+        
+        self.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
+        self.draw_centered_text("语音识别", 0, 0, 240, 30, st7789.BLACK)
+        
+        self.draw_mic_icon(105, 50)
+        
+        if audio_level > 0:
+            bar_width = min(int(audio_level * 2), 200)
+            self.tft.fill_rect(20, 100, bar_width, 10, st7789.GREEN)
+        
+        if asr_text:
+            self.text(asr_text[:20], 20, 130, st7789.WHITE, wait=False)
+        
+        if is_recording:
+            self.draw_button("松开停止", 60, 200, 120, 25, st7789.RED, st7789.WHITE)
+        else:
+            self.draw_button("长按录音", 60, 200, 120, 25, st7789.BLUE, st7789.WHITE)
+
+    def render_result_screen(self, status="", prompt="", image_received=False):
+        """渲染结果界面"""
+        if not self.tft:
+            return
+        
+        if status == "OPTIMIZING":
+            self.tft.fill(st7789.BLACK)
+            self.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
+            self.draw_centered_text("AI 生成中", 0, 0, 240, 30, st7789.BLACK)
+            
+            self.draw_centered_text("正在思考...", 0, 60, 240, 20, st7789.CYAN)
+            self.draw_centered_text("优化提示词中", 0, 80, 240, 20, st7789.CYAN)
+            self.draw_progress_bar(40, 110, 160, 6, 0.3, st7789.CYAN)
+            
+        elif status == "RENDERING":
+            self.tft.fill(st7789.BLACK)
+            self.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
+            self.draw_centered_text("AI 生成中", 0, 0, 240, 30, st7789.BLACK)
+            
+            self.draw_centered_text("正在绘画...", 0, 60, 240, 20, st7789.YELLOW)
+            self.draw_centered_text("AI作画中", 0, 80, 240, 20, st7789.YELLOW)
+            self.draw_progress_bar(40, 110, 160, 6, 0.7, st7789.YELLOW)
+            
+        elif status == "COMPLETE" or image_received:
+            # Don't clear screen, image is already there
+            self.tft.fill_rect(230, 230, 10, 10, st7789.GREEN)
+            
+        elif status == "ERROR":
+            self.tft.fill(st7789.BLACK)
+            self.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
+            self.draw_centered_text("AI 生成中", 0, 0, 240, 30, st7789.BLACK)
+            self.draw_centered_text("生成失败", 0, 50, 240, 20, st7789.RED)
+            
+        if prompt and not image_received:
+            self.tft.fill_rect(10, 140, 220, 50, 0x2124) # Dark Grey
+            self.text("提示词:", 15, 145, st7789.CYAN)
+            self.text(prompt[:25] + "..." if len(prompt) > 25 else prompt, 15, 165, st7789.WHITE)
+        
+        if not image_received:
+            self.draw_button("长按返回", 60, 210, 120, 25, st7789.BLUE, st7789.WHITE)
+
+    def draw_loading_spinner(self, x, y, angle, color=st7789.WHITE):
+        """绘制旋转加载图标"""
+        if not self.tft:
+            return
+        
+        import math
+        rad = math.radians(angle)
+        
+        center_x = x + 10
+        center_y = y + 10
+        radius = 8
+        
+        for i in range(8):
+            theta = math.radians(i * 45) + rad
+            px = int(center_x + radius * math.cos(theta))
+            py = int(center_y + radius * math.sin(theta))
+            
+            self.tft.pixel(px, py, color)
+
+    def draw_progress_bar(self, x, y, width, height, progress, color=st7789.CYAN):
+        """绘制进度条"""
+        if not self.tft:
+            return
+        
+        self.tft.fill_rect(x, y, width, height, 0x4208) # DARKGREY
+        if progress > 0:
+            bar_width = int(width * min(progress, 1.0))
+            self.tft.fill_rect(x, y, bar_width, height, color)
+
+    def draw_mic_icon(self, x, y, active=True):
+        """绘制麦克风图标"""
+        if not self.tft:
+            return
+        
+        color = st7789.GREEN if active else 0x4208 # DARKGREY
+        
+        self.tft.fill_rect(x + 5, y, 10, 5, color)
+        self.tft.fill_rect(x + 3, y + 5, 14, 10, color)
+        self.tft.fill_rect(x + 8, y + 15, 4, 8, color)
+        self.tft.fill_rect(x + 6, y + 23, 8, 2, color)
+        self.tft.fill_rect(x + 8, y + 25, 4, 3, color)

main.py

@@ -79,7 +79,7 @@ def connect_wifi(display=None, max_retries=5):
                 # 简单的加载动画
                 if display and display.tft:
                     if time.ticks_ms() % 200 < 50: # 节流刷新
-                        draw_loading_spinner(display, 120, 150, spinner_angle, st7789.CYAN)
+                        display.draw_loading_spinner(120, 150, spinner_angle, st7789.CYAN)
                         spinner_angle = (spinner_angle + 45) % 360
             
             if wlan.isconnected():
@@ -108,164 +108,6 @@ def connect_wifi(display=None, max_retries=5):
     return False
 
 
-def draw_mic_icon(display, x, y, active=True):
-    """绘制麦克风图标"""
-    if not display or not display.tft:
-        return
-    
-    color = st7789.GREEN if active else DARKGREY
-    
-    display.tft.fill_rect(x + 5, y, 10, 5, color)
-    display.tft.fill_rect(x + 3, y + 5, 14, 10, color)
-    display.tft.fill_rect(x + 8, y + 15, 4, 8, color)
-    display.tft.fill_rect(x + 6, y + 23, 8, 2, color)
-    display.tft.fill_rect(x + 8, y + 25, 4, 3, color)
-
-
-def draw_loading_spinner(display, x, y, angle, color=st7789.WHITE):
-    """绘制旋转加载图标"""
-    if not display or not display.tft:
-        return
-    
-    import math
-    rad = math.radians(angle)
-    
-    # Clear previous (simple erase)
-    # This is tricky without a buffer, so we just draw over. 
-    # For better performance we should remember previous pos.
-    
-    center_x = x + 10
-    center_y = y + 10
-    radius = 8
-    
-    for i in range(8):
-        theta = math.radians(i * 45) + rad
-        px = int(center_x + radius * math.cos(theta))
-        py = int(center_y + radius * math.sin(theta))
-        
-        # Brightness based on angle (simulated by color or size)
-        # Here we just draw dots
-        display.tft.pixel(px, py, color)
-
-def draw_check_icon(display, x, y):
-    """绘制勾选图标"""
-    if not display or not display.tft:
-        return
-    
-    display.tft.line(x, y + 5, x + 3, y + 8, st7789.GREEN)
-    display.tft.line(x + 3, y + 8, x + 10, y, st7789.GREEN)
-
-
-def draw_progress_bar(display, x, y, width, height, progress, color=st7789.CYAN):
-    """绘制进度条"""
-    if not display or not display.tft:
-        return
-    
-    display.tft.fill_rect(x, y, width, height, DARKGREY)
-    if progress > 0:
-        bar_width = int(width * min(progress, 1.0))
-        display.tft.fill_rect(x, y, bar_width, height, color)
-
-
-def render_recording_screen(display, asr_text="", audio_level=0, is_recording=False):
-    """渲染录音界面"""
-    if not display or not display.tft:
-        return
-    
-    display.tft.fill(st7789.BLACK)
-    
-    display.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
-    display.text("语音识别", 80, 8, st7789.BLACK)
-    
-    draw_mic_icon(display, 105, 50, True)
-    
-    if audio_level > 0:
-        bar_width = min(int(audio_level * 2), 200)
-        display.tft.fill_rect(20, 100, bar_width, 10, st7789.GREEN)
-    
-    if asr_text:
-        display.text(asr_text[:20], 20, 130, st7789.WHITE, wait=False)
-    
-    display.tft.fill_rect(60, 200, 120, 25, st7789.RED)
-    if is_recording:
-        display.text("松开停止", 85, 205, st7789.WHITE)
-    else:
-        display.text("长按录音", 85, 205, st7789.WHITE)
-
-
-def render_confirm_screen(display, asr_text=""):
-    """渲染确认界面"""
-    if not display or not display.tft:
-        return
-    
-    display.tft.fill(st7789.BLACK)
-    
-    display.tft.fill_rect(0, 0, 240, 30, st7789.CYAN)
-    display.text("说完了吗？", 75, 8, st7789.BLACK)
-    
-    display.tft.fill_rect(10, 50, 220, 80, DARKGREY)
-    display.text(asr_text if asr_text else "未识别到文字", 20, 75, st7789.WHITE)
-    
-    display.tft.fill_rect(20, 150, 80, 30, st7789.GREEN)
-    display.text("短按确认", 30, 158, st7789.BLACK)
-    
-    display.tft.fill_rect(140, 150, 80, 30, st7789.RED)
-    display.text("长按重录", 155, 158, st7789.WHITE)
-
-
-def render_result_screen(display, status="", prompt="", image_received=False):
-    """渲染结果界面"""
-    if not display or not display.tft:
-        return
-    
-    if status == "OPTIMIZING":
-        display.tft.fill(st7789.BLACK)
-        display.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
-        display.text("AI 生成中", 80, 8, st7789.BLACK)
-        
-        display.text("正在思考...", 80, 60, st7789.CYAN)
-        display.text("优化提示词中", 70, 80, st7789.CYAN)
-        draw_progress_bar(display, 40, 110, 160, 6, 0.3, st7789.CYAN)
-        # Spinner will be drawn by main loop
-        
-    elif status == "RENDERING":
-        display.tft.fill(st7789.BLACK)
-        display.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
-        display.text("AI 生成中", 80, 8, st7789.BLACK)
-        
-        display.text("正在绘画...", 80, 60, st7789.YELLOW)
-        display.text("AI作画中", 85, 80, st7789.YELLOW)
-        draw_progress_bar(display, 40, 110, 160, 6, 0.7, st7789.YELLOW)
-        # Spinner will be drawn by main loop
-        
-    elif status == "COMPLETE" or image_received:
-        # Don't clear screen, image is already there
-        
-        # Draw a small indicator to show it's done, but don't cover the image
-        # Maybe a small green dot in the corner?
-        display.tft.fill_rect(230, 230, 10, 10, st7789.GREEN)
-        
-    elif status == "ERROR":
-        display.tft.fill(st7789.BLACK)
-        display.tft.fill_rect(0, 0, 240, 30, st7789.WHITE)
-        display.text("AI 生成中", 80, 8, st7789.BLACK)
-        display.text("生成失败", 80, 50, st7789.RED)
-        
-    if prompt and not image_received:
-        display.tft.fill_rect(10, 140, 220, 50, 0x2124) # Dark Grey
-        display.text("提示词:", 15, 145, st7789.CYAN)
-        display.text(prompt[:25] + "..." if len(prompt) > 25 else prompt, 15, 165, st7789.WHITE)
-    
-    # Only show back button if not showing full image, or maybe show it transparently?
-    # For now, let's not cover the image with the button hint
-    if not image_received:
-        display.tft.fill_rect(60, 210, 120, 25, st7789.BLUE)
-        display.text("长按返回", 90, 215, st7789.WHITE)
-
-
-
-
-
 
 
 
@@ -368,9 +210,6 @@ def process_message(msg, display, image_state, image_data_list):
 def print_asr(text, display=None):
     """打印ASR结果"""
     print(f"ASR: {text}")
-    if display and display.tft:
-        display.fill_rect(0, 40, 240, 160, st7789.BLACK)
-        display.text(text, 0, 40, st7789.WHITE, wait=False)
 
 
 def get_boot_button_action(boot_btn):
@@ -459,6 +298,7 @@ def main():
     current_status = ""
     image_generation_done = False
     confirm_waiting = False
+    recording_stop_time = 0
     
     def connect_ws(force=False):
         nonlocal ws
@@ -502,13 +342,13 @@ def main():
         # WiFi 和 WS 都连接成功后，进入录音界面
         ui_screen = UI_SCREEN_RECORDING
         if display.tft:
-            render_recording_screen(display, "", 0, False)
+            display.render_recording_screen("", 0, False)
     else:
         print("Running in offline mode")
         # 即使离线也进入录音界面（虽然不能用）
         ui_screen = UI_SCREEN_RECORDING
         if display.tft:
-            render_recording_screen(display, "离线模式", 0, False)
+            display.render_recording_screen("离线模式", 0, False)
     
     read_buf = bytearray(4096)
     last_audio_level = 0
@@ -533,18 +373,25 @@ def main():
                 if time.ticks_diff(now, last_spinner_time) > 100:
                     if display.tft:
                         # Clear previous spinner (draw in BLACK)
-                        draw_loading_spinner(display, 110, 80, spinner_angle, st7789.BLACK)
+                        display.draw_loading_spinner(110, 80, spinner_angle, st7789.BLACK)
                         
                         spinner_angle = (spinner_angle + 45) % 360
                         
                         # Draw new spinner
                         color = st7789.CYAN if current_status == "OPTIMIZING" else st7789.YELLOW
-                        draw_loading_spinner(display, 110, 80, spinner_angle, color)
+                        display.draw_loading_spinner(110, 80, spinner_angle, color)
                     
                     last_spinner_time = now
 
             btn_action = get_boot_button_action(boot_btn)
             
+            # ASR timeout check
+            if ui_screen == UI_SCREEN_CONFIRM and confirm_waiting:
+                if time.ticks_diff(time.ticks_ms(), recording_stop_time) > 2000:
+                    confirm_waiting = False
+                    if display.tft:
+                        display.render_confirm_screen("", waiting=False)
+            
             # Hold to Record Logic (Press to Start, Release to Stop)
             if ui_screen == UI_SCREEN_RECORDING:
                 if boot_btn.value() == 0 and not is_recording:
@@ -556,7 +403,7 @@ def main():
                     current_status = ""
                     image_generation_done = False
                     if display.tft:
-                        render_recording_screen(display, "", 0, True)
+                        display.render_recording_screen("", 0, True)
                     if ws is None or not ws.is_connected():
                         connect_ws()
                     if ws and ws.is_connected():
@@ -574,8 +421,13 @@ def main():
                     is_recording = False
                     ui_screen = UI_SCREEN_CONFIRM
                     image_generation_done = False
+                    
+                    # 启动等待计时
+                    confirm_waiting = True
+                    recording_stop_time = time.ticks_ms()
+                    
                     if display.tft:
-                        render_confirm_screen(display, current_asr_text)
+                        display.render_confirm_screen(current_asr_text, waiting=True)
                     # Consume action to prevent triggering other events
                     btn_action = 0
 
@@ -591,7 +443,7 @@ def main():
                     ui_screen = UI_SCREEN_RESULT
                     image_generation_done = False
                     if display.tft:
-                        render_result_screen(display, "OPTIMIZING", current_asr_text, False)
+                        display.render_result_screen("OPTIMIZING", current_asr_text, False)
                     time.sleep(0.5)
             
             elif btn_action == 2:
@@ -603,7 +455,7 @@ def main():
                     is_recording = False
                     image_generation_done = False
                     if display.tft:
-                        render_recording_screen(display, "", 0, False)
+                        display.render_recording_screen("", 0, False)
                     time.sleep(0.5)
 
             elif btn_action == 3:
@@ -637,12 +489,13 @@ def main():
                                     if event_data[0] == "asr":
                                         current_asr_text = event_data[1]
                                         print(f"Received ASR: {current_asr_text}")
+                                        confirm_waiting = False
                                         
                                         # 收到 ASR 结果，跳转到 CONFIRM 界面
                                         if ui_screen == UI_SCREEN_RECORDING or ui_screen == UI_SCREEN_CONFIRM:
                                             ui_screen = UI_SCREEN_CONFIRM
                                             if display.tft:
-                                                render_confirm_screen(display, current_asr_text)
+                                                display.render_confirm_screen(current_asr_text, waiting=False)
                                     
                                     elif event_data[0] == "font_update":
                                         # 如果还在录音界面等待，刷新一下（虽然可能已经跳到 CONFIRM 了）
@@ -652,21 +505,21 @@ def main():
                                         current_status = event_data[1]
                                         status_text = event_data[2] if len(event_data) > 2 else ""
                                         if display.tft and ui_screen == UI_SCREEN_RESULT:
-                                            render_result_screen(display, current_status, current_prompt, image_generation_done)
+                                            display.render_result_screen(current_status, current_prompt, image_generation_done)
                                     
                                     elif event_data[0] == "prompt":
                                         current_prompt = event_data[1]
                                         if display.tft and ui_screen == UI_SCREEN_RESULT:
-                                            render_result_screen(display, current_status, current_prompt, image_generation_done)
+                                            display.render_result_screen(current_status, current_prompt, image_generation_done)
                                     
                                     elif event_data[0] == "image_done":
                                         image_generation_done = True
                                         if display.tft and ui_screen == UI_SCREEN_RESULT:
-                                            render_result_screen(display, "COMPLETE", current_prompt, True)
+                                            display.render_result_screen("COMPLETE", current_prompt, True)
                                     
                                     elif event_data[0] == "error":
                                         if display.tft and ui_screen == UI_SCREEN_RESULT:
-                                            render_result_screen(display, "ERROR", current_prompt, False)
+                                            display.render_result_screen("ERROR", current_prompt, False)
                     except Exception as e:
                         print(f"WS Recv Error: {e}")