Batch case OCR

Queue many HPO cases at once. Each case gets a case number and one or more PDF and/or text (.txt) files. When the queue runs, PDFs are read (embedded text + OCR for scans), text files are loaded directly. With Strip PHI on (default), each case produces two downloads: {case#}-clinical-text.txt (OCR/raw) and {case#}-clinical-text-deidentified.txt (stripped for external use). Upload existing .txt files to test PHI stripping without re-OCR.