Information Technology Reference
In-Depth Information
Automated inspectors alone
will not
detect all problems, but by
automating reviews and reporting before face-to-face code reviews,
teams can concentrate on more interesting and complex analyses of
code. Using automated software inspection tools puts the equivalent
of
many
more eyes on the code. In performing continuous inspections,
teams can see real benefits and timesavings. The inspection reports
bring most routine violations to light, enabling more intelligent, speedy
code reviews, improved decision-making capacity, and the best possi-
ble confidence as to the true health of a software system. Chapter 9
examines how you can utilize feedback from these inspectors to facili-
tate communication and rapid action.
Ta b l e 7-1 summarizes the practices covered in this chapter.
CI Practices Discussed in This Chapter
TABLE 7-1
Practice
Description
Reduce code
complexity
Reduce cyclomatic complexity in your code base by
leveraging automated inspectors such as JavaNCSS
or CCMetrics to identify areas of your code with
higher complexity. Run these inspectors from your
automated build.
Perform design
reviews continuously
Incorporate tools that can help determine packages/
assemblies that are highly dependent on other pack-
ages and may lead to brittle architecture.
Maintain organiza-
tional standards with
code audits
Run tools such as PMD or FxCop that report on cod-
ing standards violations from your automated build.
Reduce duplicate
code
Reduce the amount of duplicate code in a code base
by running tools such as Simian or CPD that pin-
point areas of higher code duplication based on
custom thresholds. Use this information in targeted
refactorings.
Assess code
coverage
Leverage tools such as NCover, Cobertura, or
Clover to identify line and branch coverage test code
percentages. Use this information in determining
areas that can use more tests.
The following questions will help you devise your own continuous
inspection process.
