✨[Feat] Add searchable PDF output

[maruson08]

Implements local Searchable PDF output inside /pdf/to-text/ using the existing OCR pass and pdf-lib writer.

• retains Tesseract line layout and PDF.js viewport transforms
• appends an invisible, selectable text layer while preserving original pages
• supports English, Korean, mixed text, selected pages, existing-text skipping, cancellation, and stale-job guards
• bundles a same-origin OFL-licensed Noto Sans KR font with subsetting and pinned integrity checks
• keeps the route inventory at 19 canonical routes and 17 legacy redirects

Validation:

• npm ci --ignore-scripts
• npm run prepare:ocr
• npm run typecheck
• npm test
• npm run build
• node tests/ocr-smoke.test.mjs
• git diff --check
• browser workflow and responsive QA at 1280, 768, and 360 CSS pixels

Closes #102