A simple Python unittest drop-in replacement for Python's TextTestRunner.
While I (and others) could learn the new framework dé jour, I'm much more interested in getting things working and just getting things done. I needed to start testing a Flask application, and found that Python's unittest is a tad limited out of the box. Viniciusd's TestRunner.py is a great hack, found the text output to be great but was still hacking HTMLTestRunner (of which it was based) and didn't need to output actual report files still being generated.
I wanted to quickly see what worked, what didn't, and why when it didn't. Therefore I wrote this library, to properly override methods in the TestResult and TextTestRunner classes and output text tables or a simple micro result.
Importing over TextTestRunner is all you need to do.
# from unittest import TextTestRunner
from PrettyTextTestRunner import PrettyTextTestRunner
from unittest import TestLoader, TestSuite
# Import your TestCase files
import test_suite_file1
import test_suite_file2
# Load your TestCases
testcase1 = TestLoader().loadTestsFromTestCase(test_suite_file1.testcase1)
testcase2 = TestLoader().loadTestsFromTestCase(test_suite_file2.testcase2)
# Make a Suite
suite = TestSuite([testcase1, testcase2])
# Run it
runner = PrettyTexyTestRunner()
runner.run(suite)
That's it. It's difficult enough when starting out writting test cases to have to learn a new tool, library, or framework to get things done. Focus on writting better test cases, and look to get easier (and more informative!) results to read out of the box.
Summary
------------------------------------------------------------------------------------------
| Test Suite/Test Case | Count | Pass | Fail | Error | Skip | Unexp | Expect |
| | | | | | | Sucess | Fails |
------------------------------------------------------------------------------------------
| test_case.samples | 3 | 3 | 0 | 0 | 0 | 0 | 0 |
| Totals | 3 | 3 | 0 | 0 | 0 | 0 | 0 |
------------------------------------------------------------------------------------------
test_case.samples
------------------------------------------------------------------------------------------------------
| Test Name | Output | Status |
------------------------------------------------------------------------------------------------------
| test_sample_case | | (0.01s) PASS |
| | | |
| test_sample_case2 | | (0.004s) PASS |
| | | |
| test_sample_case3 | | (0.004s) PASS |
------------------------------------------------------------------------------------------------------
runtime: 0.037s
Don't like that the test reports have border or have a margin between rows? You do you, and make it how you like it, quickly and easily. The method setProps allows you to update the default style to something more condusive for you.
newStyle = {
"titles": ["column title", ... ],
"colWidths": [x, ... ], # In characters
"colJustify": ["right"|"center"|"left", ... ],
"padding": 1, # In characters
"indent": 5, # Indent entire text block by `x` characters
"margins": {"top":0, "bottom": 0, "left": 0, "right": 0},
"border": "solid"|"columnDelimit"
}
runner.setProps(summary=newStyle, testReport=newStyle)
What's that? You want more but less? Can do! Setting the object property mode = "simple" gets you the bare minimum to scratch the itch of every dev-ops stats nerd our there.
runner.mode = "simple"
runner.run(suite)
>>> ...FFFF. pass: 3, failed: 0, errors: 0, runtime: 0.035s
Sure you don't get error messages, or output content from test cases, but maybe you don't need that.