Code Inspection, Lint & Refactoring Tools

Chapter 5 drew a clear line: the Profiler measures a genuinely running app, while static analysis examines code at rest. This chapter is that static-analysis side — Android Lint's own warnings, and the refactoring tools that go well beyond a plain find-and-replace.

Lint: Catching Problems Without Running the App

Android Lint is a static analyzer specific to Android projects, checking for real, common issues: hardcoded strings that should live in strings.xml for proper localization, unused resources, deprecated API usage, potential null-pointer risks, and accessibility warnings on views missing a content description. Warnings carry severity levels — Error, Warning, Info, Typo — and appear two ways: inline squiggly underlines while typing, or a full project scan via Analyze > Inspect Code.

Reading and Acting on a Lint Warning

Many lint warnings come with an automated, one-click fix — clicking the yellow lightbulb icon next to a warning often applies a correct fix directly, not just a description of the problem. When a warning is a genuinely known, accepted false positive rather than a real issue, suppressing it deliberately — via @SuppressLint in code, or a lint baseline for the whole project — records that decision explicitly, rather than simply ignoring the warning silently and losing that reasoning later.

Refactoring Tools Beyond Find-and-Replace

Rename updates every real reference to a variable, method, or class across the entire project safely — not just text that happens to match the same string. Extract Method/Extract Variable pulls a chunk of existing code into its own named method or variable automatically, generating the correct surrounding code. Safe Delete checks whether something is actually unused anywhere in the project before deleting it, warning explicitly if it's still referenced somewhere unexpected rather than deleting blindly.

Why "Rename" Is Safer Than Find-and-Replace

A plain text find-and-replace can't tell the difference between a variable named count and an unrelated string literal or comment that happens to contain the word "count" — it operates purely on matching text. Rename instead operates on the IDE's own actual understanding of the code's structure — which specific declaration a given identifier refers to — making it a genuinely different, safer operation, not just a smarter find-and-replace.

Hands-On Exercises