NSE Live Tick Data Collector

I want a Python-based service that connects to the NSE equity feed as soon as the market opens (9:15 AM IST) and keeps running until closing (3:30 PM IST). Every equity-tick that comes through during that window needs to be captured and committed—nothing filtered out or deleted in the process.

For each tick the tool must log:
• stock symbol
• precise timestamp
• LTP and LTQ
• cumulative traded volume
• total buy quantity and total sell quantity
• the best five bid prices with their quantities
• the best five ask prices with their quantities
• an identifier for the stock universe being tracked

I will do all rule-based filtering afterward, so please ensure the raw stream for every stock is preserved exactly as received.

Storage needs to be an SQL database; feel free to design the schema that best suits fast inserts and later querying. I would like the code to rely on SQLAlchemy for ORM / connection handling so it stays portable across engines (PostgreSQL is what I use in production).

Provide clean, well-commented source in a Git-friendly structure plus a brief README that explains:
• how to set up any NSE credentials or API keys
• how to launch the collector, including restart behaviour if the connection drops
• the database schema or migration file you created
• a quick way to replay or inspect a day’s worth of captured data

A working proof—e.g., a short recorded session or sample DB dump—will serve as acceptance that the collector records every required field uninterrupted for the full session.