process-supernotelisted

# Process Supernote ## Steps ### 1. Find new or modified Supernote files - List `.note` files in `/Users/alexhedtke/My Drive/Supernote/Note/` (including subdirectories) - Read `/Users/alexhedtke/Documents/Exobrain harness/processing-log.json` - Compare file modification times against the log to find new or updated files ### 2. Extract pages from each .note file Run the parser: ```bash python3 "/Users/alexhedtke/Documents/Exobrain harness/transcript-processing/supernote-parser.py" "<note_file>" "/tmp/supernote_pages/<note_name>" ``` This outputs individual PNG page images, a JSON summary, and **per-page SHA-256 hashes** in `page_hashes`. ### 3. Diff pages using hashes (skip already-processed pages) Compare the `page_hashes` from the parser output against the `pageHashes` stored in the processing log for this file: - **New file** (no log entry): all pages are new -- OCR all of them - **Updated file** (log entry exists): only OCR pages whose hash differs from or doesn't exist in the stored `pageHashes` - **Pages with identical hashes to a known blank page should be skipped** -- blank Supernote pages produce identical hashes (e.g., two blank pages will share the same hash). If multiple pages at the end of a note share the same hash, they are likely blank -- skip them unless their content hash differs from previously seen blank pages. - Track which page numbers are new/changed for use in steps 5 and 6 ### 4. OCR new/changed pages only Read each **new or changed** PNG file usi