Calculate Code Health in Parallel A performance optimization that makes better use of multiple CPU cores. When enabling it, check your JVM settings to ensure you have at least -Xss2m specified as stack size to handle large hotspots.

Performance Implications. Enabling the Full Code Health Scan will increase your total analysis time. CodeScene uses a cache to optimize the calculations, but the first analysis will be slower; with this option enabled, the analysis time increases near linearly with the number of files in your repositories. Sub-sequent analyses will be quicker, thanks to the cache.

Full Code Health Scan Enable this option to scan the health of all code in the project.

Code Health Level for Alerts, Goals, and Quality Gates By default, we don't provide any code health decline alerts or trigger goals for healthy code (e.g. code that degrades from 10.0 => 9.33). You can change that here if you're a perfectionist; putting this value at 10 will warn for all declines. Note that this value serves as a threshold in the PR/CI/CD quality gates as well.

Custom Code Health

Start with our template (JSON file) for defining custom code health rules. CodeScene lets you override specific code health rules via a config file in each Git repository. Download this template to inspect and tweak the rules. The template contains further instructions on how to store and enable the rules. Commit the template as .codescene/code-health-rules.json inside your Git repos. Read more in the documention.

Disable custom code health rules specified in the Git repositories Check this box to disable all overriden rules and run with CodeScene's calibrated weights. This might be a useful exercise to see if there are any overriden issues of interest.

Repository with global code health rules Specify the name of one of the repositories in the project that contains global rules, e.g. codescene-rules-repo. This way you can specify a set of global rules in one Git repository and have those rules applied across all Git repositories in your project.

Note: The delta analysis uses the global rules that were present during the last full analysis. This means that when updating global rules, a full analysis has to be performed before they apply to delta analyses in all project repositories.

Defect Mining Pattern

Specify a regular expression that identifies bug fix commits to get statistics on the defect distribution across your hotspots.

Defect Pattern A regular expression that matches bug fixes in the commit messages. For example, CR{2,}-\d+ matches commits like CR-782 and CR-231, while the expression [Bb]ug|[Ff]ix is a heuristic to match commits containing bug fixes.

Code Comment Patterns

X-Ray

Analyse overloaded methods separately When checked, each overloaded method is treated as a separate unit of analysis in the X-Ray. The default behavior is to combine overloaded methods an present the statistics for all of the overloads as one, single entry.

Include all Revisions When checked, all historic revisions of a file are included in the X-Ray of a Hotspot. This may be quite time consuming, which is why CodeScene limits the X-Ray history to the last 200 revisions unless you enable this option.

Minimum coupling degree Minimum coupling degree info

Early Warning: Rising Hotspots

Range for trend supervision (days) Specifies how far back in time we will look for a rising spot. With more development activity, choose a shorter time span.

Early Warning: Complexity Trends

Cut-off code age for early warning (months) Ignore historic artifacts by only generating early warnings for code that has been modified during the past months as configured here.

Code Age

Use last commit as age reference When checked, the code age is calculated relative to the date of the last commit, otherwise CodeScene uses the current analysis date.

Renamed and Moved Content

Maximum number of files in a commit for rename detection CodeScene detects moved/renamed source code files and folders. However, that detection increases the analysis time. This number specifies the maximum number of files in a commit before we give up on rename detection.