Add Gemma-based image recognition scaffold

src/app.py

@@ -17,6 +17,7 @@ from discord_control.actions import setDiscordMute
 from ocr.ocr_actions import runOcrFromScreen
 from translate.translate_actions import saveTranslationText
 from translate.translate_actions import translateTextToJapanese
+from vision.vision_actions import runVisionFromScreen
 from vrc_log.log_actions import collectVrchatLog
 from vrc_log.log_actions import startSelfMonitor
 from vrc_log.log_actions import startVrcLogMonitor
@@ -27,6 +28,7 @@ PARAM_DISCORD_MUTE = "DiscordSend"
 PARAM_LOG_GREP = "vrc_log"
 PARAM_OCR = "ocrEnabled"
 PARAM_TRANSLATE_OCR = "translation"
+PARAM_VISION = "visionEnabled"
 
 
 def create_gateway():
@@ -96,10 +98,20 @@ def create_gateway():
             saved = saveTranslationText(stem, translated)
             gateway.log("INFO", f"translation saved={saved}")
 
+    def on_vision(address, *args):
+        if not args:
+            gateway.log("ERROR", "visionEnabled args empty")
+            return
+
+        if gateway.is_rising_edge(address, args[0]):
+            gateway.log("INFO", f"received {address} args={args}")
+            runVisionFromScreen()
+
     gateway.map(PARAM_DISCORD_MUTE, on_discord_mute)
     gateway.map(PARAM_LOG_GREP, on_log_grep)
     gateway.map(PARAM_OCR, on_ocr)
     gateway.map(PARAM_TRANSLATE_OCR, on_translate_ocr)
+    gateway.map(PARAM_VISION, on_vision)
     gateway.set_default_debug(enabled=False)
     return gateway