feat(pdf): text extraction, OCR, and analyze endpoints

[simonhagger]

Summary

Adds three new Python sidecar endpoints for document/text processing:

• /extract-text: PDF text extraction via PyMuPDF fitz (per-page)
• /ocr: OCR pipeline via pytesseract + tesseract binary
• /analyze-text: text utility analysis (word count, paragraph count, normalization, language detection)

Adds TypeScript text processing utilities: \wordCount, \paragraphCount, \ runcate,
ormalizeWhitespace, \detectLanguage\ in \libs/util/common.

Changes

• \service.py: /extract-text, /ocr, /analyze-text handlers; proper cv2 import guard
• \python.contract.ts: request/response schemas for all three endpoints
• \channels.ts: channel keys for all three endpoints
• \python-handlers.ts: validated IPC handlers for all three endpoints
• \python-sidecar.ts: sidecar methods for all three endpoints
• \python-api.ts: preload bridge methods
• \desktop-api.ts: DesktopPythonApi interface methods
• \common.ts: wordCount, paragraphCount, truncate, normalizeWhitespace, detectLanguage
• \ est_service.py: 29 Python tests at 93% coverage

Validation

• \pnpm nx run-many -t build test lint\ — all 12 projects pass

• Python sidecar tests — 29 passed, 93% coverage (threshold: 90%)

• \pnpm typecheck\ — clean

• Security review completed

• Threat model updated or N/A explained

• Confirmed no secrets/sensitive data present in committed files