AcceptPy is a minimal acceptance test framework. It can be used to automatically and quickly check that a program operates as a developer intends. What a unit test framework is for functions, AcceptPy is for whole programs.
For example we can use AcceptPy to test the operation of the console echo command. To do so we first author a test file (in JSON) which defines a test for the echo command:
{
"test type": "shell command",
"command" : "echo have some pie.",
"expect exit": 0,
"expect stdout": "have some pie.\n"
}
This test checks than executing echo have some pie. prints:
have some pie
to the console and then exits with code 0.
We can run this test by invoking AcceptPy passing the directory that contains the test file:
python3 accept.py /home/user/echo_tests
which produces the following output:
#########################
Welcome to AcceptPy
#########################
searching for tests...
/home/user/echo_test/echo_pie.test
found 1 tests.
#########################
Test 1 of 1
#########################
[running: /home/user/echo_tests/echo_pie.test]
shell command: echo have some pie.
<begin stdout>have some pie.
<end stdout>
[PASSED]
#########################
Results:
#########################
1 of 1 tests passed.
When running the test defined in echo_pie.test AcceptPy:
- Launched a shell subprocess.
- Ran the command
echo have some pie.in the shell. - Checked that the command's exit code was 0.
- Checked that the total of the command's printed stdout was
have some pie.\n
Version 1.2.0 can be downloaded directly from the command line via
wget https://github.com/rmbar/acceptpy/releases/download/v1.3.0/accept.py
or by downloading an archive from https://github.com/rmbar/acceptpy/releases/tag/v1.3.0 and copying /src/main/python/accept.py into a directory of your choice.
A shell command test is defined in a JSON file with the .test extension. It is generally of the form:
{
"test type": "shell command",
"command" : "echo have some pie.",
"expect exit": 0,
"expect stdout": "have some pie.\n",
"expect stdout contains": "pie"
}
When a shell command test file is encountered AcceptPy will spawn a shell subprocess and execute the command string
identified by command. If expect exit is specified AcceptPy will check that the command terminates with the given
exit code. If not, the tool will check that the command terminated with exit code 0. If expect stdout is specified
AcceptPy will check that the text sent to standard out by the command matches the given text. If expect stdout contains is specified
AcceptPy will check that the text sent to standard out by the command contains the given text.
A Python test is defined in a .py file. When AcceptPy encounters such a file it will execute the Python program
in a new subprocess using the command:
python3 file.py
and score the test as a pass if the returned exit code is 0. Any other exit code will result in a failed scoring.
Running multiple tests is accomplished by authoring multiple test files.
For example:
$ echo '{ "test type": "shell command", "command" : "echo dog" }' > echo_dog.test
$ echo '{ "test type": "shell command", "command" : "echo cat" }' > echo_cat.test
$ python3 ~/accept.py ./
(n.b. in these minimal test files AcceptPy will only check that the echo commands terminated with exit code 0.)
... prodcues the output:
#########################
Welcome to AcceptPy
#########################
searching for tests...
./echo_cat.test
./echo_dog.test
found 2 tests.
#########################
Test 1 of 2
#########################
[running: ./echo_cat.test]
shell command: echo cat
<begin stdout>cat
<end stdout>
[PASSED]
#########################
Test 2 of 2
#########################
[running: ./echo_dog.test]
shell command: echo dog
<begin stdout>dog
<end stdout>
[PASSED]
#########################
Results:
#########################
2 of 2 tests passed.
When running AcceptPy the program will recursively search all directories under the given test directory argument
for all files ending in .test or .py.
It is always polite to accept py.