[Refactor] Define MM data parser in processing info instead of processor itself (#33260)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/model_executor/models/gemma3_mm.py

@@ -290,10 +290,8 @@ class Gemma3MultiModalProcessor(BaseMultiModalProcessor[Gemma3ProcessingInfo]):
 
         # HF processor pops the `num_crops` kwarg, which is needed by vLLM
         if (images := mm_data.get("images")) is not None:
-            parsed_images = (
-                self._get_data_parser()
-                .parse_mm_data({"image": images})
-                .get_items("image", ImageProcessorItems)
+            parsed_images = self.data_parser.parse_mm_data({"image": images}).get_items(
+                "image", ImageProcessorItems
             )
             image_sizes = [
                 parsed_images.get_image_size(i) for i in range(len(parsed_images))