themis_cli
v2.2.0
Published
Cli for Themis workflow
Downloads
135
Readme
Themis CLI
Made to work with Themis - Programming Judge of the RUG.
Use this cli tool to easily batch download test cases (and only test cases) and test your program against these downloaded test cases. Your program can either be an executable or a .py
file.
Examples
Installation
If you use nvm, you can install themis cli globally using npm:
npm install -g themis_cli
Otherwise, you might need sudo permissions. After installing, you should now be able to use themis everywhere. To check try:
themis
Requirements
This tool requires you to have python3
, requests
, and beautifulsoup4
on your pc. To install the requests
and beautifulsoup4
libraries you can simply run the command:
pip install requests beautifulsoup4
Make sure you have python3
installed before running the above command!
Features
- Batch download test cases
- Run tests locally
- Save user name
- Save preferred year
How to get started
After installation, a script will generate a default configuration for you.
You can change this configuration using themis user <s_number>
and themis year
.
Although not needed, if your are downloading test cases frequently it is nice to save your user name:
themis user <s_number>
You can change the academic year with:
themis year
To check your saved username and year type:
themis info
If you ever lose you configuration file, you can generate a new one with:
themis setup
We can start the test case downloading flow by typing:
themis get
This will run you through options until you are at the location where the tests you want are found.
This will download all the files to the ./tests/
directory.
If the tests directory is not present in your current directory it will create it.
Note, if you want to run themis test [a.out]
on these tests, you should call themis get
from within the directory where your executable will be located.
We can also judge our program locally. To do this, compile your code to an executable (or have a .py
file) and:
themis test [a.out]
In this case a.out
is just an example, any name for the executable would work.
This will test your program against the downloaded tests in the ./tests
directory.
It expects that the directory where you call themis test [a.out]
from contains the a.out
executable and the tests
directory.
It will tell you how many test cases you passed and will store wrong output in the ./tests-diffs
directory.
If no executable is provided in the command (themis test
) then the program will look for a.out
or main
(in that order).
So if your executable is called a.out
or main
you can just run themis test
.
Verbose and no-redirect options
The themis test
command has 2 flags you can use: -v
(--verbose
) and -n
(--no-redirect
). v
prints content of the failed tests' diff files, and n
gives the tests files as an argument to the program (so instead of ./a.out < 1.in
you would get ./a.out 1.in
).
In the end we get a project directory that looks something like:
assignmentDir
│
├── main.c
├── main (executable)
│
├── tests
│ ├── 1.in
│ ├── 1.out
│ ├── 2.in
│ └── 2.out
│
└── tests-diffs
└── 2.diff
Where calling themis get
from assignmentDir would have yielded the tests
directory and themis test
would have yielded the tests-diffs
directory if our program passed test case 1 but not test case 2.
Note that we can use themis test
in this case because our program executable is called main
.