Web Profiler

This project has three main goals:

• profiling the speed of single pages along a prefixed flow, with very detailed timings that will allow to spot infrastructure problems. Pretty much tries to answer to users saying "It's slow"
• regression testing (with very simple checks for now)
• pre-flight checklist. Another take on regression testing

For now everything relies on a very primitive scenario file for which you find an example in the scenarios folder. The format is almost self explanatory. Some documentation will follow.

All the steps are performed serially. This is by design. The client is configured to keep a cookie-jar in order to act as an authenticated user against a regular session. Some very peculiar choiches (especially regarding form POSTing) account for the intricacies of ez-publish legacy. More on this in a later version

docker build -t web-profiler
docker run -v ./scenarios/:/scenarios -e "scenario=/scenarios/scenario_1.json" --rm marco/web-profiler | pino-pretty