Run Your First Query¶

This lesson connects to CUBRID, sends one SQL statement and reads the one row it returns. It is the lesson that follows Your First CUBRID Connection, which prepares the database and Python.

Before you start¶

You need what the preparation lesson set up:

a CUBRID database that answers: the Docker lab printed CUBRID is ready (Before you start), or you have the five values for your own database;
pycubrid installed in a virtual environment (Step 1).

Work in the same folder as Step 1. In a new terminal, activate the virtual environment again with source .venv/bin/activate. Each step is a file that you run from the terminal, so nothing carries over from an earlier Python session.

Write the program¶

Create a file named first_query.py in that folder:

import pycubrid

# Connection settings for the disposable local lab. With your own database,
# replace them with the values from your administrator.
conn = pycubrid.connect(
    host="localhost",  # the machine that runs the CUBRID broker
    port=33000,  # the broker port (33000 is the CUBRID default)
    database="testdb",  # the database name
    user="dba",
    password="",
)

cur = conn.cursor()  # a cursor sends SQL and reads results
cur.execute("SELECT 1 + 1")  # execute sends one SQL statement
print(cur.fetchone())  # fetchone reads one result row

cur.close()
conn.close()

Run it¶

python first_query.py

It prints one row with one value:

(2,)

If you see an error instead, find its last line in If the first connection fails.

What the program did¶

pycubrid.connect() opens a connection, your session with the database. It needs the five values: host, port, database, user and password.
conn.cursor() creates a cursor. A cursor works inside a connection: it sends SQL statements and reads their results.
cur.execute() sends one SQL statement to the server.
cur.fetchone() reads one row of the result. A row is a tuple, so the single value 2 prints as (2,).
close() ends the cursor and then the connection. Close what you opened, so the server does not keep an idle session for you.

By default connect() sets no time limit of its own on reaching the server. If a firewall silently drops the connection, the call waits until the operating system gives up. With a database on another machine, add for example connect_timeout=10 to the connect() call; the Connection Guide describes it and the other timeouts.

Next step¶

The first query only read a computed value. Next, save a row and read it back: create a small table, save one row and read it from a new connection.