JS Code Guardian

VS Code extension for JavaScript/TypeScript (and some embedded JS/TS cases like Vue/Svelte) that helps you catch syntax/keyword typos, offers one-click fixes, and adds useful inline information.

Features

1) Red diagnostics for keyword / syntax typos

The extension scans code and highlights identifiers that look like JavaScript syntax/keyword typos (example: funciton → function).

• Uses fuzzy matching (Damerau–Levenshtein) with context checks to reduce false positives.
• Shows additional “Correct syntax” info in the diagnostic related information when available.

2) Hover help + “Fix with JS Code Guardian”

When you hover over a highlighted issue, you get:

• Message like: Unknown syntax "..." Did you mean "..."?
• A clickable action: [Fix with JS Code Guardian] (implemented via a trusted VS Code command).

3) One-click Quick Fix

From the diagnostic or Quick Fix menu, the extension performs an edit that replaces the wrong range with the computed fix.

4) Dim unused symbols

The extension applies text decorations (opacity) to:

• Unused bindings (variables, imports, functions, classes, parameters)
• Unused object properties in patterns

5) Inline previews for console & popup calls

For calls where the runtime value can be evaluated statically (limited depth), it shows inline hints:

• console.*(...) → inline text preview after the call
• alert(...), confirm(...), prompt(...) → inline popup preview text after the call

Safe built-in methods are evaluated when their inputs are known, including string methods like toUpperCase(), array methods like map(), filter(), join(), slice(), number methods like toFixed(), plus Math.*, Object.keys(), Object.values(), and Array.isArray().

Supported languages

Based on the extension activation and language selectors:

• javascript
• typescript
• javascriptreact
• typescriptreact
• vue
• svelte

How it works (high level)

• Parses source using Babel parser with language-appropriate plugins.
• Uses rules + documentation from: src/dataset/data.json
  • syntax: documentation for known APIs/keywords
  • corrections: typo → fix mappings for keyword/syntax identifiers
• If parsing fails, it falls back to a lightweight tokenization approach to still find likely keyword typos.