Python from container

Python from container#

For me it’s common task to build interaption with container which handle python program.

So in this section I want to show you how to build such an interaction.

Python image#

You need to create a Python image. The following cell describes the docker file that’s used to create the image.

%%writefile python_container/dockerfile
FROM python:3.11.4
WORKDIR program
COPY script.py script.py
COPY requirements.txt requirements.txt
RUN pip3 install -r requirements.txt

Overwriting python_container/dockerfile

The next cell describes the requirements.txt we’ll need for this container.

%%writefile python_container/requirements.txt
numpy==1.24.2
psycopg2-binary==2.9.6

Overwriting python_container/requirements.txt

And python script that will be used for this example. This script just pushes some data to the postgres database that is accessible in pg_example_postgres_cont hostname.

%%writefile python_container/script.py
import psycopg2
import random
import string

# establish connection with database
conn = psycopg2.connect(
    dbname = "docker_app_db",
    user = "docker_app",
    password = "docker_app",
    # interesting that host name match
    # with the container name
    # where database deployed
    host = "pg_example_postgres_cont"
)

# add 20 random values to the database
cur = conn.cursor()
for i in range(20):
    text = ''.join(random.choices(string.ascii_lowercase, k=20))
    query = f"INSERT INTO main_table (id, text) VALUES ('{i}', '{text}');"
    cur.execute(query)
cur.close()

print("Adding records is done!")

# commit changes and close the connection
conn.commit()
conn.close()

Overwriting python_container/script.py

Finally, we build the image.

%%bash
docker build -t pg_example_python python_container &> /dev/null

Start containers#

We need to create at least one table in our postgres database. So in the following cell we store the script that will be used to initialise the database.

%%writefile python_container/create_table.sql
CREATE TABLE main_table(
    id TEXT NOT NULL,
    text TEXT NOT NULL
);

Overwriting python_container/create_table.sql

Next, start the python and postgres containers.

You should share the same net between containers, in the following example which is completed by test_project_net.

%%bash
# network
docker network create test_project_net &> /dev/null
# postgres
docker run --rm -d\
    --name pg_example_postgres_cont\
    -e POSTGRES_USER=docker_app\
    -e POSTGRES_PASSWORD=docker_app\
    -e POSTGRES_DB=docker_app_db\
    --net=test_project_net\
    -v ./python_container/create_table.sql:/docker-entrypoint-initdb.d/create_table.sql\
    postgres:15.4 &> /dev/null
# python
docker run --rm -itd\
    --name pg_example_python_cont\
    --net=test_project_net\
    pg_example_python &> /dev/null

Executing python#

In the next cell, we run the Python script in the Python container. The script simply adds some random values to the main_table from the ohter container. The message “Adding records is done!” indicates that the Python program was executed normally.

%%bash
docker exec pg_example_python_cont bash -c "python3 script.py"

Adding records is done!

Check the table#

So now, to make sure we have done everything right, let us select values from the created table.

%%bash
docker exec pg_example_postgres_cont bash -c \
"psql -U docker_app -d docker_app_db -c \"SELECT * FROM main_table;\""

 id |         text         
----+----------------------
 0  | rxldgrgulelclxkokbyi
 1  | heafyvxrlzfjfxdgvdsj
 2  | pkdapikpwobwpkdeyijb
 3  | hlozeyvffyeanoftzylh
 4  | nmkxuqhthxzsiybzpcpe
 5  | ccvfrxjrjjgnwxgbqamz
 6  | fdfokyejpdzrftwavgdb
 7  | sqedmmfbflqslzuanwpx
 8  | rupnmkvnjcevjwzprzyk
 9  | cudsmqmbvmyyoollcygi
 10 | ofctrxmrbputjbgsrpwy
 11 | zkzwfxeflrxwacbgqaox
 12 | yvonchnoianeavxzcjpm
 13 | jwiuyfbwrhwtaqtvtzdk
 14 | gbhrkjyixaurapdualca
 15 | hxrabcnvbxjracvoheen
 16 | fnfmwktakiaxaqumyjce
 17 | poenzivvvotyblpwclrk
 18 | xxhczciogflcdscttvpq
 19 | anfzuhbovjohlqngjbux
(20 rows)

Stop containers#

%%bash
docker stop pg_example_postgres_cont pg_example_python_cont &> /dev/null
docker network rm test_project_net &> /dev/null

Clear python image#

To avoid creating a lot of rubbish in docker images, you should delete the container created for this example.

%%bash
docker rmi pg_example_python &> /dev/null