[codex] Harden OCR runtime for scanned PDFs

.changeset/ocr-runtime-hardening.md

@@ -0,0 +1,8 @@
+---
+'@happyvertical/pdf': patch
+---
+
+Improve Node OCR runtime handling for scanned PDFs by auto-detecting Tesseract
+`tessdata`, surfacing actionable OCR dependency errors, routing explicit
+external OCR providers like `onnx` through the `unpdf` pipeline, and rendering
+OCR pages directly with `pdfjs-dist` to avoid worker-version mismatches.

.gitignore

@@ -27,3 +27,6 @@ pnpm-debug.log*
 .env
 .env.local
 .env.*.local
+
+# OCR runtime cache
+.kreuzberg/

biome.json

@@ -2,9 +2,21 @@
   "$schema": "https://biomejs.dev/schemas/1.9.4/schema.json",
   "files": {
     "include": [
-      "src/**/*.{js,ts,tsx,jsx}",
-      "test/**/*.{js,ts,tsx,jsx}",
-      "*.{js,ts,tsx,jsx}"
+      "src/**/*.js",
+      "src/**/*.jsx",
+      "src/**/*.ts",
+      "src/**/*.tsx",
+      "test/**/*.js",
+      "test/**/*.jsx",
+      "test/**/*.ts",
+      "test/**/*.tsx",
+      "scripts/**/*.js",
+      "scripts/**/*.ts",
+      "*.js",
+      "*.jsx",
+      "*.ts",
+      "*.tsx",
+      "*.json"
     ],
     "ignoreUnknown": true
   },

commitlint.config.js

@@ -13,17 +13,17 @@ export default {
       2,
       'always',
       [
-        'feat',      // New feature
-        'fix',       // Bug fix
-        'docs',      // Documentation changes
-        'style',     // Code style changes (formatting, missing semicolons, etc.)
-        'refactor',  // Code refactoring (neither fixes a bug nor adds a feature)
-        'perf',      // Performance improvements
-        'test',      // Adding or updating tests
-        'build',     // Changes to build system or dependencies
-        'ci',        // Changes to CI configuration files and scripts
-        'chore',     // Other changes that don't modify src or test files
-        'revert',    // Reverts a previous commit
+        'feat', // New feature
+        'fix', // Bug fix
+        'docs', // Documentation changes
+        'style', // Code style changes (formatting, missing semicolons, etc.)
+        'refactor', // Code refactoring (neither fixes a bug nor adds a feature)
+        'perf', // Performance improvements
+        'test', // Adding or updating tests
+        'build', // Changes to build system or dependencies
+        'ci', // Changes to CI configuration files and scripts
+        'chore', // Other changes that don't modify src or test files
+        'revert', // Reverts a previous commit
       ],
     ],
 
@@ -35,17 +35,13 @@ export default {
         'pdf',
         'ocr',
         'unpdf',
-        'deps',      // Dependency updates
-        'release',   // Release-related changes
+        'deps', // Dependency updates
+        'release', // Release-related changes
       ],
     ],
 
     // Subject case - allow sentence-case, lowercase, kebab-case, etc.
-    'subject-case': [
-      2,
-      'never',
-      ['upper-case', 'pascal-case', 'start-case'],
-    ],
+    'subject-case': [2, 'never', ['upper-case', 'pascal-case', 'start-case']],
 
     // Subject full stop - no period at the end
     'subject-full-stop': [2, 'never', '.'],
@@ -73,5 +69,5 @@ export default {
 
     // Body max line length - disable for semantic-release commits
     'body-max-line-length': [0],
-  ],
+  },
 };

lefthook.yml

@@ -5,7 +5,7 @@
 pre-commit:
   commands:
     format:
-      glob: "*.{js,jsx,ts,tsx}"
+      glob: "**/*.{js,jsx,ts,tsx}"
       run: pnpm biome check --write {staged_files}
       stage_fixed: true
 

package.json

@@ -15,11 +15,7 @@
       "pdfjs-dist": "5.4.296"
     }
   },
-  "files": [
-    "dist",
-    "README.md",
-    "LICENSE"
-  ],
+  "files": ["dist", "README.md", "LICENSE"],
   "publishConfig": {
     "registry": "https://npm.pkg.github.com",
     "access": "public"
@@ -45,7 +41,7 @@
     "build": "vite build",
     "build:watch": "vite build --watch",
     "typecheck": "tsc --noEmit",
-    "lint": "biome check",
+    "lint": "biome check .",
     "format": "biome format --write src/",
     "clean": "rm -rf dist *.tsbuildinfo",
     "dev": "npm run build:watch & npm run test:watch",
@@ -56,9 +52,10 @@
     "prepare": "lefthook install"
   },
   "dependencies": {
-    "@happyvertical/ocr": "^0.60.4",
-    "@happyvertical/utils": "^0.71.0",
-    "pdf-to-png-converter": "^3.10.0",
+    "@happyvertical/ocr": "^0.60.33",
+    "@happyvertical/utils": "^0.71.20",
+    "@napi-rs/canvas": "0.1.97",
+    "pdfjs-dist": "5.4.296",
     "unpdf": "^1.4.0"
   },
   "optionalDependencies": {