refactoring code and setting up api server

api/main.py

@@ -1,22 +1,13 @@
-from fastapi import FastAPI, HTTPException, Depends
+from fastapi import FastAPI, Depends, UploadFile, Response
+from fastapi.responses import FileResponse
 from pydantic import BaseModel
-from typing import List, Annotated
-import models
-from database import engine, SessionLocal
-from sqlalchemy.orm import Session
+import os
+import sys
+from scripts.subtitles import generate_subtitles
+root_path = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
+sys.path.append(root_path)
 
 app = FastAPI()
-models.Base.metadata.create_all(bind=engine)
-
-def get_db():
- db = SessionLocal()
- try:
- yield db
- finally:
- db.close()
-
-
-db_dependency = Annotated[Session, Depends(get_db)]
 
 
 @app.get("/_readyz")
@@ -25,34 +16,21 @@ def read_root():
  return {"status": "ok"}
 
 
-class Needs(BaseModel):
- transcoding: bool
- subtitles: bool
- thumbnails: bool
-
-
-class Project(BaseModel):
- project: str
- videoURL: str
- needs: Needs
-
-
-@app.get("/projects")
-async def read_projects(db:db_dependency):
- result = db.query(models.Projects).all()
- return result
-
-@app.post("/generate-all")
-async def create_project(project:Project, db: db_dependency):
- # Step 1: Insert project into db
- db_project = models.Projects(project=project.project, videoURL=project.videoURL, needTranscoding=project.needs.transcoding, needSubtitles=project.needs.subtitles, needThumbnails=project.needs.thumbnails)
- db.add(db_project)
- db.commit()
- db.refresh(db_project)
+@app.post("/subtitles")
+def create_subtitles(video_name: str, file: UploadFile):
+ if not os.path.exists("videos"):
+ os.makedirs("videos")
 
- # Step 2: Generate transcoding
+ file_path = f"videos/{video_name}.mp4"
+ with open(file_path, "wb") as f:
+ f.write(file.file.read())
 
- # Step 3: Upload to bucket
+ if not file.content_type.startswith("video/"):
+ return {"error": "Invalid file type. Only video files are allowed."}
 
- # Step 4: Update table with bucket
- return project
+ if not os.path.exists("output"):
+ os.makedirs("output")
+ output_path = generate_subtitles(file_path,"vtt","./output",video_name)
+ response = FileResponse(output_path, media_type="text/vtt", filename=f"{video_name}.vtt")
+ response.headers["Content-Disposition"] = f"attachment; filename={video_name}.vtt"
+ return response

api/readme.MD

@@ -4,23 +4,29 @@
 
 ### Requirements:
 - Python3 
-- PostgreSQL DB
 
 ### Installation:
 
 - Create a new Virtual Environment 'env' using `python3 -m venv env`
 - Activate the `env` Environment using `source env/bin/activate`
-- Install libraries using `pip install fastapi sqlalchemy psycopg2-binary uvicorn` or `pip install -r requirements.txt`
-- Create sanchayai DB and replace your connection string in /api/database.py (postgresql:https://postgres:test123@localhost:5432/sanchayai)
+- Install libraries using `pip install -r requirements.txt`
 
 ### Run Project:
 
 - python3 run.py
 - On your browser: http:https://127.0.0.1:5000/docs
 
+### Routes available
+
+- /subtitles (You can try them at http:https://127.0.0.1:5000/docs )
+
+Future routes:
+- /chapters
+- /thumbnails
+
 
 Reference links:
-- https://youtu.be/398DuQbQJq0?si=dAW1-QT3bxgZ45Oq
+- https://youtu.be/398DuQbQJq0?si=dAW1-QT3bxgZ45Oq (can be ignored for now)
 
 
 For any queries, add your doubts in the issues section and attach the `question` label to it.

api/requirements.txt

@@ -11,4 +11,7 @@ sniffio==1.3.0
 SQLAlchemy==2.0.24
 starlette==0.32.0.post1
 typing_extensions==4.9.0
-uvicorn==0.25.0
+uvicorn==0.25.0
+openai-whisper
+tiktoken
+openai

api/scripts/subtitles.py

@@ -0,0 +1,18 @@
+import whisper
+from whisper import utils
+
+
+def generate_subtitles(input_file_path, output_type, output_file_path, output_file_name):
+ model = whisper.load_model("tiny")
+ result = model.transcribe(input_file_path)
+ writer = utils.get_writer(output_type, output_file_path)
+ writer_args = {
+ "highlight_words": False,
+ "max_line_count": None,
+ "max_line_width": None,
+ "max_words_per_line": None,
+ }
+ writer(result, output_file_name, **writer_args)
+ output_path = f"{output_file_path}/{output_file_name}.{output_type}"
+ print("⚡️ Success! ⚡️ Subtitles have been generated @ "+output_path)
+ return output_path