The latest version, Understand 7.0, introduces Interactive Reports that transform basic code metrics into comprehensive and captivating visual representations. These reports are highly customizable, allowing you to adapt the analysis to meet your specific requirements and uncover the underlying narrative of your coding efforts.
Improving my coding quality is a goal for this year. But what exactly does “improvement” entail? To gauge my progress, I require measurable metrics – and that’s precisely what the Metrics feature in Understand provides. Understand is equipped with an extensive array of metrics, from basic line counts to more complex measures like cyclomatic complexity.
However, mere numbers aren’t enough to keep me motivated. The real challenge lies in how these numbers are presented. I know eating vegetables is healthy, but I won’t consider them unless they’re part of a delicious meal. Similarly, Understand 7.0’s new interactive reports serve up code metrics in a way that is both insightful and engaging. Just like a skilled chef crafts a dish to appeal to the senses, you can customize these reports to suit your taste. This article will delve into how you can use these customizable reports, including creating your own plugins, to visually interpret and comprehend your code, moving beyond mere numbers to unveil the detailed stories your metrics tell.
Inspiration for Ingredients
For this piece, I’m focusing more on how information is presented rather than the information itself. I’ll draw content inspiration from existing views within Understand. Let’s briefly review all the current metrics reports and views available in Understand:
- The Welcome Page features a Line Breakdown Chart, Most Complex Functions Chart, and a Directory Structure by Lines of Code Chart.
- The Project Home page expands on these with additional charts, primarily concerning lines of code and complexity.
Onward from these default pages, the most prominent views include the Metrics Browser and Metrics CSV Export, both accessible from the Metrics menu. There’s also a lesser-known Metrics HTML Export accessible from the Metrics Browser. These views comprehensively display all activated metrics in Project Config and any enabled metric plugins.
Additional, lesser-known metrics views exist within classic reports. Despite showing signs of age, these reports remain popular and continue to be included in the application. The Overview page of the HTML reports displays earlier versions of charts similar to those on the Project Home page. There are also five specific metric reports for my OpenSSL project.
Developing a Line Metrics Report Plugin
Source Lines of Code (SLOC) and related metrics are an excellent foundation for my first plugin. They correlate with time, cost, and potentially, code quality. Here’s my rationale:
- Lines to Time to Money: More lines typically mean more time needed to comprehend, modify, use, or fix the code, and time translates directly to money.
- Lines and Architecture Meaning: Understand can categorize your code into different architectures, such as by license type using an automatic architecture plugin. However, it’s the line metrics that truly reveal how much of my project pertains to each license type.
- Lines and Quality Indicators: Extremely long functions or files are likely hard to understand. The ratio of comments to code also indicates quality.
To present line metrics, I’ll integrate elements from existing displays in Understand:
- Charts: I appreciate the quick, graphical summary provided by the Line Breakdown Chart from the Welcome Page and Project Home Page.
- Tables: I favor the detailed view in the classic reports’ tables, and interactive report tables could be enhanced with filters.
- Hierarchy: The tree structure in the Metrics HTML Export is useful, but it’s bulky and overly detailed. Thus, interactive reports incorporate a breadcrumb system for easier navigation.
The resulting Line Metric Interactive Report plugin can be applied to any architecture (or specific file or the entire project), displaying a line breakdown chart for the current level, a table of child elements, and a breadcrumb navigation system. This new report is available in the Plugin Manager in Understand.
Exploring Complexity Metrics
Cyclomatic complexity is a critical metric for assessing the testability of code, indicating the number of decision points and, hence, the number of test cases required. It also affects understandability and maintainability. The McCabe Cyclomatic Complexity metric, along with standards like HIS Metrics, provides recommended thresholds.
By tweaking my Line Metrics report, I can easily create a complexity report. I’ll replace the line breakdown chart with the “Most Complex Functions” chart and swap line metrics for complexity metrics, including the Cognitive Complexity metric because of
its relevance.
Considering Other Reports
What next? After integrating information from three classic reports and two-thirds of the Welcome Page charts, I could replicate the Metrics CSV Export or Metrics HTML Report. However, instead of overwhelming and cluttered reports, I’ll focus on combining the metrics from the “File Averages” report. This involves both complexity and line metrics, presented in a flat bubble chart for visual contrast against a square table.
