Add Extended Text Extraction and Enhanced Filtering Features

README.md

@@ -43,8 +43,8 @@ npx i18n-scan <directory>
 # If installed globally
 i18n-scan <directory>
 
-# Run locally during development
-npm run dev <directory>
+# Run locally during development (note the -- before flags)
+npm run dev <directory> -- [options]
 
 # Run built version locally
 npm start <directory>
@@ -57,16 +57,31 @@ npm start <directory>
 npm run dev ./src
 
 # Scan with custom file extensions
-npm run dev ./src --ext .tsx,.jsx,.ts,.js
+npm run dev ./src -- --ext .tsx,.jsx,.ts,.js
 
 # Scan with custom attributes
-npm run dev ./src --attributes title,alt,placeholder,aria-label
+npm run dev ./src -- --attributes title,alt,placeholder,aria-label
+
+# Filter out non-alphabetic text (symbols, emojis, numbers only)
+npm run dev ./src -- --filter-non-alpha
+
+# Extract from objects, function calls, and schemas
+npm run dev ./src -- --include-literals
+
+# Combine options
+npm run dev ./src -- --filter-non-alpha --truncate 50
+
+# When using npx or installed globally, no -- needed:
+npx i18n-scan ./src --filter-non-alpha --truncate 50
 ```
 
 ### Options
 
 - `-e, --ext <extensions>`: Comma-separated list of file extensions (default: `.tsx,.jsx`)
 - `-a, --attributes <names>`: Comma-separated list of attribute names to extract (default: `title,alt,placeholder`)
+- `-t, --truncate <limit>`: Truncate text to this length, 0 to disable (default: `30`)
+- `-f, --filter-non-alpha`: Filter out text containing only non-alphabetic characters (e.g., `$`, `123`, `🎉`)
+- `-l, --include-literals`: Include string literals from objects, function calls, and validation schemas
 
 ## Example Output
 
@@ -77,6 +92,17 @@ npm run dev ./src --attributes title,alt,placeholder,aria-label
 [src/components/Form.tsx:13] This field is required
 ```
 
+**With `--filter-non-alpha` enabled:**
+
+```
+# Without filter:
+[src/components/ProductCard.tsx:30] $
+[src/components/ProductCard.tsx:33] In Stock
+
+# With filter ($ is filtered out):
+[src/components/ProductCard.tsx:33] In Stock
+```
+
 ## Development
 
 - `npm run dev <directory>`: Run from TypeScript source
@@ -85,6 +111,8 @@ npm run dev ./src --attributes title,alt,placeholder,aria-label
 
 ## What it extracts
 
+### By default (JSX only):
+
 1. **JSX Text Content**: Any text between JSX tags
 
    ```jsx
@@ -99,6 +127,50 @@ npm run dev ./src --attributes title,alt,placeholder,aria-label
    <button title="Close dialog" />        // Extracts: "Close dialog"
    ```
 
+### With `--include-literals` flag:
+
+Extracts additional string literals that are commonly user-facing:
+
+3. **Object Properties**: Strings from objects with keys like `message`, `error`, `label`, `title`, `description`, etc.
+
+   ```typescript
+   const errors = {
+     required: "This field is required", // Extracts
+     invalid: "Invalid email address", // Extracts
+   };
+
+   const config = {
+     className: "btn-primary", // Skips (technical)
+     title: "Welcome back", // Extracts
+     apiUrl: "/api/users", // Skips (technical)
+   };
+   ```
+
+4. **Zod Schema Messages**: Validation messages and descriptions
+
+   ```typescript
+   z.string().email("Invalid email")
+   z.string().min(5, "Too short")
+   z.string().max(100, { message: "Too long"
+   z.string().describe("Your username")
+   ```
+
+5. **Function Calls**: Common user-facing function calls
+
+   ```typescript
+   toast.success("Saved successfully!");
+   toast.error("Failed to save");
+   console.error("Something went wrong");
+   throw new Error("Invalid input");
+   ```
+
+**Smart Filtering**: The tool automatically skips technical strings like:
+
+- Import paths, URLs, API endpoints
+- CSS class names, IDs
+- camelCase identifiers, CONSTANTS
+- Single characters and code patterns
+
 ## Github Repo
 
 https://github.com/hexxt-git/i18n-scan

example/src/test-literals.tsx

@@ -0,0 +1,68 @@
+import { z } from "zod";
+
+// Test 1: Object with user-facing strings
+const errorMessages = {
+  required: "This field is required",
+  invalid: "Invalid input provided",
+  tooShort: "Must be at least 5 characters",
+};
+
+const config = {
+  title: "Welcome to our platform",
+  description: "Start your journey today",
+  buttonText: "Get Started Now",
+};
+
+// Test 2: Zod schemas
+const userSchema = z.object({
+  email: z.string().email("Please enter a valid email address"),
+  password: z
+    .string()
+    .min(8, "Password must be at least 8 characters")
+    .max(100, { message: "Password is too long" }),
+  username: z.string().describe("Your unique username"),
+});
+
+const ageSchema = z.number().min(18, "You must be 18 or older");
+
+// Test 3: Toast/Error calls
+function handleSubmit() {
+  toast.success("Form submitted successfully!");
+  toast.error("Failed to submit form");
+  console.error("Something went wrong");
+  console.warn("This action cannot be undone");
+}
+
+// Test 4: Error throwing
+function validateInput(value: string) {
+  if (!value) {
+    throw new Error("Input cannot be empty");
+  }
+  if (value.length < 3) {
+    throw new Error("Input must be at least 3 characters long");
+  }
+}
+
+// Test 5: Should NOT extract (technical strings)
+const technicalStrings = {
+  className: "flex items-center", // Should skip
+  apiUrl: "/api/users", // Should skip
+  method: "POST", // Should skip
+};
+
+// Test 6: Mixed object (should extract some, skip others)
+const mixedConfig = {
+  endpoint: "/api/submit", // Should skip
+  message: "Your data has been saved", // Should extract
+  className: "button-primary", // Should skip
+  label: "Save Changes", // Should extract
+};
+
+export function TestComponent() {
+  return (
+    <div>
+      <h1>Test Component</h1>
+      <p>Regular JSX text should still work</p>
+    </div>
+  );
+}