Merge pull request #250 from presenton/perf/load-icon-finder-once

perf/load icon finder once

servers/fastapi/api/v1/ppt/endpoints/files.py

@@ -5,7 +5,7 @@ from fastapi import APIRouter, Body, File, UploadFile
 
 from constants.documents import UPLOAD_ACCEPTED_FILE_TYPES
 from models.decomposed_file_info import DecomposedFileInfo
-from services import TEMP_FILE_SERVICE
+from services.temp_file_service import TEMP_FILE_SERVICE
 from services.documents_loader import DocumentsLoader
 from utils.randomizers import get_random_uuid
 from utils.validators import validate_files

servers/fastapi/api/v1/ppt/endpoints/icons.py

@@ -1,11 +1,10 @@
 from typing import List
 from fastapi import APIRouter
-from services.icon_finder_service import IconFinderService
+from services.icon_finder_service import ICON_FINDER_SERVICE
 
 ICONS_ROUTER = APIRouter(prefix="/icons", tags=["Icons"])
 
 
 @ICONS_ROUTER.get("/search", response_model=List[str])
 async def search_icons(query: str, limit: int = 20):
-    icon_finder_service = IconFinderService()
-    return await icon_finder_service.search_icons(query, limit)
+    return await ICON_FINDER_SERVICE.search_icons(query, limit)

servers/fastapi/api/v1/ppt/endpoints/outlines.py

@@ -7,7 +7,7 @@ from sqlalchemy.ext.asyncio import AsyncSession
 from models.presentation_outline_model import PresentationOutlineModel
 from models.sql.presentation import PresentationModel
 from models.sse_response import SSECompleteResponse, SSEResponse, SSEStatusResponse
-from services import TEMP_FILE_SERVICE
+from services.temp_file_service import TEMP_FILE_SERVICE
 from services.database import get_async_session
 from services.documents_loader import DocumentsLoader
 from services.score_based_chunker import ScoreBasedChunker

servers/fastapi/api/v1/ppt/endpoints/presentation.py

@@ -2,7 +2,7 @@ import asyncio
 import json
 import os
 import random
-from typing import Annotated, List, Literal, Optional
+from typing import Annotated, List, Optional
 from fastapi import APIRouter, Body, Depends, HTTPException
 from fastapi.responses import StreamingResponse
 from sqlalchemy import delete
@@ -21,7 +21,6 @@ from models.presentation_structure_model import PresentationStructureModel
 from models.presentation_with_slides import PresentationWithSlides
 
 from utils.get_layout_by_name import get_layout_by_name
-from services.icon_finder_service import IconFinderService
 from services.image_generation_service import ImageGenerationService
 from utils.dict_utils import deep_update
 from utils.export_utils import export_presentation
@@ -30,7 +29,7 @@ from models.sql.slide import SlideModel
 from models.sse_response import SSECompleteResponse, SSEResponse
 
 from services.database import get_async_session
-from services import TEMP_FILE_SERVICE
+from services.temp_file_service import TEMP_FILE_SERVICE
 from models.sql.presentation import PresentationModel
 from services.pptx_presentation_creator import PptxPresentationCreator
 from utils.asset_directory_utils import get_exports_directory, get_images_directory
@@ -197,7 +196,6 @@ async def stream_presentation(
         )
 
     image_generation_service = ImageGenerationService(get_images_directory())
-    icon_finder_service = IconFinderService()
 
     async def inner():
         structure = presentation.get_structure()
@@ -234,9 +232,7 @@ async def stream_presentation(
 
             # This will mutate slide
             async_assets_generation_tasks.append(
-                process_slide_and_fetch_assets(
-                    image_generation_service, icon_finder_service, slide
-                )
+                process_slide_and_fetch_assets(image_generation_service, slide)
             )
 
             yield SSEResponse(
@@ -386,7 +382,6 @@ async def generate_presentation_api(
     )
 
     image_generation_service = ImageGenerationService(get_images_directory())
-    icon_finder_service = IconFinderService()
     async_asset_generation_tasks = []
 
     # 7. Generate slide content and save slides
@@ -407,9 +402,7 @@ async def generate_presentation_api(
             content=slide_content,
         )
         async_asset_generation_tasks.append(
-            process_slide_and_fetch_assets(
-                image_generation_service, icon_finder_service, slide
-            )
+            process_slide_and_fetch_assets(image_generation_service, slide)
         )
         slides.append(slide)
         slide_contents.append(slide_content)

servers/fastapi/api/v1/ppt/endpoints/slide.py

@@ -1,4 +1,3 @@
-import importlib
 from typing import Annotated, Optional
 from fastapi import APIRouter, Body, Depends, HTTPException
 from sqlalchemy.ext.asyncio import AsyncSession
@@ -6,7 +5,6 @@ from sqlalchemy.ext.asyncio import AsyncSession
 from models.sql.presentation import PresentationModel
 from models.sql.slide import SlideModel
 from services.database import get_async_session
-from services.icon_finder_service import IconFinderService
 from services.image_generation_service import ImageGenerationService
 from utils.asset_directory_utils import get_images_directory
 from utils.llm_calls.edit_slide import get_edited_slide_content
@@ -42,12 +40,10 @@ async def edit_slide(
     )
 
     image_generation_service = ImageGenerationService(get_images_directory())
-    icon_finder_service = IconFinderService()
 
     # This will mutate edited_slide_content
     new_assets = await process_old_and_new_slides_and_fetch_assets(
         image_generation_service,
-        icon_finder_service,
         slide.content,
         edited_slide_content,
     )

servers/fastapi/services/__init__.py

@@ -1,4 +0,0 @@
-from services.temp_file_service import TempFileService
-
-
-TEMP_FILE_SERVICE = TempFileService()

servers/fastapi/services/icon_finder_service.py

@@ -11,9 +11,9 @@ class IconFinderService:
         self.client = chromadb.PersistentClient(
             path="chroma", settings=Settings(anonymized_telemetry=False)
         )
-        print('Initializing icons collection...')
+        print("Initializing icons collection...")
         self._initialize_icons_collection()
-        print('Icons collection initialized.')
+        print("Icons collection initialized.")
 
     def _initialize_icons_collection(self):
         self.embedding_function = ONNXMiniLM_L6_V2()
@@ -51,3 +51,6 @@ class IconFinderService:
             n_results=k,
         )
         return [f"/static/icons/bold/{each}.png" for each in result["ids"][0]]
+
+
+ICON_FINDER_SERVICE = IconFinderService()

servers/fastapi/services/temp_file_service.py

@@ -65,3 +65,6 @@ class TempFileService:
 
     def cleanup_base_dir(self):
         self.cleanup_temp_dir(self.base_dir)
+
+
+TEMP_FILE_SERVICE = TempFileService()

servers/fastapi/utils/export_utils.py

@@ -8,7 +8,7 @@ from pathvalidate import sanitize_filename
 from models.pptx_models import PptxPresentationModel
 from models.presentation_and_path import PresentationAndPath
 from services.pptx_presentation_creator import PptxPresentationCreator
-from services import TEMP_FILE_SERVICE
+from services.temp_file_service import TEMP_FILE_SERVICE
 from utils.asset_directory_utils import get_exports_directory
 from utils.randomizers import get_random_uuid
 

servers/fastapi/utils/process_slides.py

@@ -3,7 +3,7 @@ from typing import List, Tuple
 from models.image_prompt import ImagePrompt
 from models.sql.image_asset import ImageAsset
 from models.sql.slide import SlideModel
-from services.icon_finder_service import IconFinderService
+from services.icon_finder_service import ICON_FINDER_SERVICE
 from services.image_generation_service import ImageGenerationService
 from utils.asset_directory_utils import get_images_directory
 from utils.dict_utils import get_dict_at_path, get_dict_paths_with_key, set_dict_at_path
@@ -11,7 +11,6 @@ from utils.dict_utils import get_dict_at_path, get_dict_paths_with_key, set_dict
 
 async def process_slide_and_fetch_assets(
     image_generation_service: ImageGenerationService,
-    icon_finder_service: IconFinderService,
     slide: SlideModel,
 ) -> List[ImageAsset]:
 
@@ -33,7 +32,7 @@ async def process_slide_and_fetch_assets(
     for icon_path in icon_paths:
         __icon_query__parent = get_dict_at_path(slide.content, icon_path)
         async_tasks.append(
-            icon_finder_service.search_icons(__icon_query__parent["__icon_query__"])
+            ICON_FINDER_SERVICE.search_icons(__icon_query__parent["__icon_query__"])
         )
 
     results = await asyncio.gather(*async_tasks)
@@ -60,7 +59,6 @@ async def process_slide_and_fetch_assets(
 
 async def process_old_and_new_slides_and_fetch_assets(
     image_generation_service: ImageGenerationService,
-    icon_finder_service: IconFinderService,
     old_slide_content: dict,
     new_slide_content: dict,
 ) -> List[ImageAsset]:
@@ -138,7 +136,7 @@ async def process_old_and_new_slides_and_fetch_assets(
             continue
 
         async_icon_fetch_tasks.append(
-            icon_finder_service.search_icons(new_icon["__icon_query__"])
+            ICON_FINDER_SERVICE.search_icons(new_icon["__icon_query__"])
         )
         new_icons_fetch_status.append(True)