Does GitLab CI treat Docker entrypoint & cmd differently somehow? - docker

Question

I'm getting some unexpected behaviour when trying to run a GitLab CI job inside a custom Docker container. It runs as expected outside GitLab. I don't even know what it is I'm not understanding... perhaps there's something about how GitLab runs docker containers that I'm missing, or how Docker ENTRYPOINT and CMD work together, or is it even just a mistake in my shell script?
My custom Docker image is intended to run Percy CLI commands. It's based on node:12-slim and basically just installs Puppeteer and Percy. It sets the ENTRYPOINT to a custom shell script:
ENTRYPOINT ["/usr/bin/local/percy-entrypoint.sh"]
but doesn't override the base image CMD.
My entrypoint script essentially just checks for a couple of env vars (and chucks out some debug stuff), then runs a "sitemap" function:
echo "All args:"
echo "$#"
if [ "$1" = "sitemap" ] ; then
percy exec -- node /sitemap.js
else
exec "$#"
fi
I'm running it from a GitLab-CI job as:
snapshot:
stage: test
image: my/percy-buildbox:latest
script:
- sitemap
variables:
PERCY_TOKEN: ${PERCY_TOKEN}
SITEMAP_URL: percy-sitemap.xml
The output generated by this job is:
All args:
sh -c if [ -x /usr/local/bin/bash ]; then
exec /usr/local/bin/bash
elif [ -x /usr/bin/bash ]; then
exec /usr/bin/bash
elif [ -x /bin/bash ]; then
exec /bin/bash
elif [ -x /usr/local/bin/sh ]; then
exec /usr/local/bin/sh
elif [ -x /usr/bin/sh ]; then
exec /usr/bin/sh
elif [ -x /bin/sh ]; then
exec /bin/sh
elif [ -x /busybox/sh ]; then
exec /busybox/sh
else
echo shell not found
exit 1
fi
$ sitemap
/bin/bash: line 134: sitemap: command not found
I don't even... where's all that if stuff coming from? What's it doing? Where am I going wrong?