Separate data aggregation and report building

[yeugenius]

Spock-reports has two primary stages that run independently of each other - firstly, data aggregation, with uses Spock global extension mechanism to collect test data and store it in aggregated_report.json and, secondly, actual report building later. Right now they reside is the same package. It would be nice to clearly split these features at least in two different packages and ideally (later) into different projects, like 'spock-data-aggregator' and 'spock-reports'.

The reason why I'm raising this questions is that right now I'm working on project that is going to use data aggregation code extensively without much changes and then automatically publish this data to external service, like JIRA, testcase management system, any sort of webservice etc. This will open possibility to integrate spock-reports with CI/CD pipelines, automatic builds etc, as it would be easy to get test data and publish it where you need.

[renatoathaydes]

Quite a few people seem to be interested in the contents of the aggregated_report.json file for this purpose... I initially added that file in order to be able to aggregate parallel tests reports, but turns out it can be useful for other cases, like merging Gradle multi-module projects for example, besides what you mention about CI.

This shouldn't be a lot of work to achieve, I would consider if someone submitted a PR with that... If I find some time in the future I might try it myself, but I can't promise anything right now.