Python SDK Examples

• The Python client wraps the rust client.
• It currently does not support the administrator features that the rust client does.
• The PartitionConsumer.stream returns an object which implements the python iterator convention to allow for iterating over the stream in a for-loop.

To see the full docs, visit our pdoc page.

Examples

Producer

Create the topic used to produce and consume records:

$ fluvio topic create python-data

Create a file called python-produce.py:

#!/usr/bin/env python
from datetime import datetime
from fluvio import Fluvio

TOPIC_NAME = "python-data"
PARTITION = 0

if __name__ == "__main__":
   # Connect to cluster
   fluvio = Fluvio.connect()

   # Produce 10 records to topic
   producer = fluvio.topic_producer(TOPIC_NAME)
   for x in range(10):
       producer.send_string("{}: timestamp: {}".format(x, datetime.now()))

   # Flush the last entry
   producer.flush()

Let's run the file:

$ python python-produce.py

Consumer

Create a file called python-consume.py:

#!/usr/bin/env python
from fluvio import Fluvio, Offset

TOPIC_NAME = "python-data"
PARTITION = 0

if __name__ == "__main__":
   # Connect to cluster
   fluvio = Fluvio.connect()

   # Consume last 10 records from topic
   consumer = fluvio.partition_consumer(TOPIC_NAME, PARTITION)
   for idx, record in enumerate( consumer.stream(Offset.from_end(10)) ):
       print("{}".format(record.value_string()))

       if idx >= 9:
           break

Let's run the file:

$ python python-consume.py

Example with a SmartModule

This is a 3 part example:

• Build a SmartModule that converts records to uppercase
• Add a Python script that uses the SmartModule
• Test the Python script

Python Script that uses a SmartModule

Create a file called smartmodule-consumer.py:

#!/usr/bin/env python
import os
from datetime import datetime
from fluvio import Fluvio, Offset, ConsumerConfig

TOPIC_NAME = "hello-python-smartmodule"
PARTITION = 0

# This is an example of a basic Fluvio workflow in Python
#
# 1. Create a topic to store data in via CLI
# 2. Establish a connection to the Fluvio cluster
# 3. Create a producer and send some bytes
# 4. Create a consumer, and stream the data back
if __name__ == "__main__":
   # Currently the Python client does not support creating topics
   # Using the Fluvio CLI
   os.popen("fluvio topic create {}".format(TOPIC_NAME))

   # Connect to cluster
   fluvio = Fluvio.connect()

   # Produce to topic
   producer = fluvio.topic_producer(TOPIC_NAME)
   producer.send_string("Hello World! - Time is: {}".format(datetime.now()))

   # Consume from topic
   # We're just going to get the last record
   consumer = fluvio.partition_consumer(TOPIC_NAME, PARTITION)

   # Create a ConsumerConfig using your "uppercase" smartmodule
   config = ConsumerConfig()
   config.smartmodule(name="uppercase")

   for record in consumer.stream_with_config(Offset.from_end(0), config):
       print("{}".format(record.value_string()))
       break

Test Python Script

Read to test the script:

$ python smartmodule-consumer.py
HELLO WORLD! - TIME IS: 2024-08-25 21:04:51.172045