CSV to Kafka

This example watches local CSV files, converts each row to a JSON object (using the header row as keys), and publishes messages to a Kafka topic. When a CSV file is modified, only the changed rows are re-published.

Prerequisites

• A running Kafka broker (default: localhost:9092)

Configuration

Edit .env to set your Kafka connection:

KAFKA_BOOTSTRAP_SERVERS=localhost:9092
KAFKA_TOPIC=csv-rows

Run

Install deps:

pip install -e .

Run the pipeline in live mode so changes to CSV files are picked up automatically:

cocoindex update -L main.py

Each row is published as a JSON message with key {filename}/{first_column_value}. For example, a row in products.csv with SKU SKU001 gets key products.csv/SKU001.