2 - Getting a Codecov account and uploading coverage
Create a new branch step2
git checkout main
git pull
git checkout -b 'step2'
Before we continue, let’s create an account and set up our repository on Codecov.
Create a Codecov account
Create an account on Codecov by clicking on our signup page and following the prompts.
Set up the Team Bot. If you are using your personal account, you can create a codecov.yml
file in the root directory.
codecov:
bot: {{ YOUR_USERNAME }} # e.g. bot: thomasrockhu-codecov
ignore:
- "api/app.py"
Now that we have set up our Codecov account, let’s upload coverage reports.
Create CI pipeline
Create a configuration file for your CI. The CI workflows below will checkout the code, install requirements, run tests, and upload the coverage reports to Codecov.
Bitbucket Pipelines
You'll need to add a Codecov upload token as an environment variable in BitBucket. You can find one in the Configuration tab for your repository in the Codecov UI, under General.
First, enable Bitbucket Pipelines for this repository by going to Repository Settings
-> Settings
(under Pipelines). Click Enable Pipelines
.
Next, add the Codecov token under Repository Settings
-> Repository Variables
(under Pipelines). Under name, enter CODECOV_TOKEN
. Copy the token from above as the value.
pipelines:
default:
- step:
image: python:3.10.4
name: api
script:
- pip install -r api/requirements.txt
- pytest --cov api/
- curl -Os https://cli.codecov.io/latest/linux/codecov
- chmod +x codecov
- ./codecov upload-process -t $CODECOV_TOKEN
Commit the changes and run the CI workflow
Let’s commit our code and open a pull request. Run
git add .
git commit -m 'step2: upload coverage reports to Codecov'
git push origin step2
Next, go to the demo repository on Bitbucket and open a pull request (PR).
When opening pull requests, be sure to select your own repository as the base branch.
You will see status checks on the PR from the CI. You can click on them to see the progress of the CI/CD. Once complete, you will receive status checks and a comment from Codecov.
codecov/patch
and codecov/project
status checks
One of Codecov’s core uses is getting code coverage information directly in a developer’s workflow. The status checks shown above are part of that.
- The
project
check is used to maintain a coverage percentage across an entire codebase - The
patch
check is used for code changed in a PR.
A more mature repository may have a strict project
percentage, while a project new to code coverage may only expect new code to be well-tested (patch
coverage). We will be setting some of these values in this section, but you can read more about status checks here.
If you go to the project on Codecov, you will notice that the dashboard is blank. This is because no coverage information has been uploaded onto the default branch. We can fix this by merging the pull request.
Merge the pull request on Bitbucket and wait for the CI/CD to complete.
When opening pull requests, be sure to select your own repository as the base branch.
Updated 4 months ago