← back to section

A tester's work runs into volume and repetition. Checking that the totals add up across a thousand orders by hand is impossible — you can sample, and sampling is not a check. Preparing fifty orders in different states before a regression round is an hour of clicking, every time. Comparing two exports of five hundred rows is work where the eye glazes over by the hundredth line.

All of it takes seconds in code, and code does not get tired. Hence the simple answer to "why should a tester program": not to earn the automation title, but to check what cannot be checked by hand.

Why Python

There are many languages and arguing about the best one is pointless — but for entering automation Python wins on three counts.

It reads almost like English. To print a line, Python says print("hello"); Java needs System.out.println("hello") inside a class and a method you also have to declare. While you are learning, extra words get in the way: every one of them needs explaining, and you have nothing to explain it with yet.

Everything for testing is ready-made. API requests, JSON parsing, spreadsheets, running tests — all of it is libraries that install with one command and work out of the box.

It is the language of testing. Your questions have been asked before, and your future colleagues write in the same language — you can show your code and discuss it.

What you will be able to do

Not "know Python" — that is too vague. Concretely:

  • parse a service response and check the fields that matter;
  • compute from an export what the interface never shows;
  • compare two data sets and find what went missing and what appeared extra;
  • write an autotest that checks the API and fails when the service is broken.

The last point is the goal of the section. Everything else is a step towards it.

How to run the first line

There are two ways, and the second one is better at the start.

Install Python. Download the installer for your system from python.org. On Windows tick "Add Python to PATH" in the installer — without it the python command is not found. Check the install:

python3 --version

An answer like Python 3.12.4 means the language is in place. Then create a file first.py, put one line in it and run it:

live example

print("Hello, tester")
Run

Running examples is part of paid access. There the same code runs inside the article: editor, run and check next to the paragraph. Free week →

python3 first.py

Start in the trainer. Nothing to install: the Python trainer is already open, the code is written in the browser and runs on a button. That is easier at the start: you deal with the language, not with installers, paths and encodings.

You will still need the install later — when you get to real autotests in a work project. By then you will understand what exactly you are installing.

How working with code feels

Three things worth accepting right away — they save weeks.

A program runs top to bottom, line by line. No magic: what is written first happens first.

The computer takes you literally. A typo in a variable name is an error. A missing bracket is an error. This is not pedantry: the language does not guess, it executes.

An error is a message, not a disaster. Python says what did not add up and on which line. Reading those messages is half the skill; in the trainer they appear right under your code.

How to go through the section

Every article ends with trainer tasks. They are meant to be solved, not skimmed: reading about code and writing code are different skills, like reading about swimming and swimming.

The order is not accidental: strings are needed to parse responses, dicts to understand JSON, functions to keep a check in one place. You can skip ahead, but it catches up with you at the autotests.

In short

  • A tester needs code not for the job title but to check what hands cannot: volume, repetition, data comparison.
  • Python wins on readability, ready-made testing libraries, and being the language the profession speaks.
  • It is easier to start in the trainer: the code is written and run in the browser, nothing to install.
  • The install from python.org comes later, for autotests in a real project.
  • A program runs top to bottom, takes you literally, and its error messages are hints you need to learn to read.