chore(tests): Create a test to check whether requirements.txt was generated with pip-compile

[hbelmiro]

Chore description

Create a test to check whether requirements.txt was generated with pip-compile. That can be achieved by running pip-compile and comparing the result with the existing requirements.txt.
This test should be run only when a requirements.in or a requirements.txt file is modified.

Labels

/area testing

---

Love this idea? Give it a 👍.

[chahatsagarmain]

@hbelmiro I would like to work on this , Please assign this to me .

[mahdikhashan]

i would like to work on this.

/assign

[mahdikhashan]

@hbelmiro shall i keep the test in the test dir in the root? and may i use python/pytest for it? are there requirements for it to be add to the ci?

[hbelmiro]

@mahdikhashan You can just create a bash script in https://github.com/kubeflow/pipelines/tree/master/.github/resources/scripts and call it from the CI. It can be something like:

# Generate a temporary file for comparison
temp_file=$(mktemp)

# Generate requirements.txt using pip-compile
pip-compile --output-file=- requirements.in > "$temp_file"

# Compare the generated file with the existing requirements.txt
if diff -q "$temp_file" requirements.txt >/dev/null; then
  echo "Success: requirements.txt matches the output of pip-compile."
  rm "$temp_file"
  exit 0
else
  echo "Error: requirements.txt does not match the output of pip-compile."
  echo "Differences:"
  diff "$temp_file" requirements.txt
  rm "$temp_file"
  exit 1
fi

Please check all occurrences of requirements.in/requirements.txt in the test.

find . -name "requirements.in"
./test/sample-test/requirements.in
./test/kfp-functional-test/requirements.in
./backend/metadata_writer/requirements.in
./backend/requirements.in
./backend/src/apiserver/visualization/requirements.in
./sdk/python/requirements.in

[mahdikhashan]

@hbelmiro i need to create a new workflow file for it, right? otherwise, in which workflow shall i add it - probably as a job?

and in my test, i need to make sure the bash script is running in the root of the repo so that it can find all requirements in all nested directories. please correct me if i'm wrong. thanks for your time.