cd /home/daytona/workspace/attachments/e784f551-31ce-41d8-b413-65d344b1ffc0/ && python3 -c " import pdfplumber with pdfplumber.open('1. IBE-I (PHB-I)-compressed.pdf') as pdf: print(f'Total pages: {len(pdf.pages)}') for i, page in enumerate(pdf.pages): print(f'--- PAGE {i+1} ---') text = page.extract_text() if text: print(text) print() " 2>&1 | head -300

python3 -c " import subprocess result = subprocess.run(['which', 'tesseract'], capture_output=True, text=True) print('tesseract:', result.stdout.strip()) result2 = subprocess.run(['which', 'pdftoppm'], capture_output=True, text=True) print('pdftoppm:', result2.stdout.strip()) result3 = subprocess.run(['python3', '-c', 'import pytesseract; print(pytesseract.__version__)'], capture_output=True, text=True) print('pytesseract:', result3.stdout.strip(), result3.stderr.strip()) result4 = subprocess.run(['python3', '-c', 'from pdf2image import convert_from_path; print(\"pdf2image ok\")'], capture_output=True, text=True) print('pdf2image:', result4.stdout.strip(), result4.stderr.strip()) "

python3 << 'EOF' from pdf2image import convert_from_path import pytesseract import os pdf_path = '/home/daytona/workspace/attachments/e784f551-31ce-41d8-b413-65d344b1ffc0/1. IBE-I (PHB-I)-compressed.pdf' output_dir = '/home/daytona/workspace/ocr_output' os.makedirs(output_dir, exist_ok=True) print("Converting PDF to images and running OCR...") # Process in batches to avoid memory issues pages = convert_from_path(pdf_path, dpi=200, first_page=1, last_page=30) full_text = "" for i, page in enumerate(pages): text = pytesseract.image_to_string(page, lang='eng') full_text += f"\n\n=== PAGE {i+1} ===\n{text}" print(f"Done page {i+1}") with open(f'{output_dir}/pages_1_30.txt', 'w') as f: f.write(full_text) print("Pages 1-30 done") EOF

mkdir -p /home/daytona/workspace/ocr_output && cd /home/daytona/workspace/ocr_output && pdftoppm -r 150 '/home/daytona/workspace/attachments/e784f551-31ce-41d8-b413-65d344b1ffc0/1. IBE-I (PHB-I)-compressed.pdf' page && ls page-*.ppm | head -20

cd /home/daytona/workspace/ocr_output && ls page-*.ppm | wc -l && for f in page-0{01..30}.ppm; do echo "=== $f ===" >> all_text.txt; tesseract "$f" stdout -l eng 2>/dev/null >> all_text.txt; done && echo "Done pages 1-30"

python3 << 'PYEOF' import subprocess import os ocr_dir = '/home/daytona/workspace/ocr_output' out_file = os.path.join(ocr_dir, 'all_text.txt') all_text = "" for i in range(1, 106): ppm_file = os.path.join(ocr_dir, f'page-{i:03d}.ppm') if os.path.exists(ppm_file): result = subprocess.run( ['tesseract', ppm_file, 'stdout', '-l', 'eng'], capture_output=True, text=True, timeout=30 ) page_text = result.stdout all_text += f"\n\n=== PAGE {i} ===\n{page_text}" if i % 10 == 0: print(f"Done page {i}") with open(out_file, 'w', encoding='utf-8') as f: f.write(all_text) print(f"Total chars: {len(all_text)}") print("DONE") PYEOF

python3 << 'PYEOF' import subprocess, os ocr_dir = '/home/daytona/workspace/ocr_output' all_text = "" # Only do first 20 pages to get the questions for i in range(1, 21): ppm_file = os.path.join(ocr_dir, f'page-{i:03d}.ppm') if os.path.exists(ppm_file): r = subprocess.run(['tesseract', ppm_file, 'stdout', '-l', 'eng'], capture_output=True, text=True, timeout=25) all_text += f"\n\n=== PAGE {i} ===\n{r.stdout}" print(f"p{i} done") with open(f'{ocr_dir}/pages_1_20.txt','w') as f: f.write(all_text) print("DONE") PYEOF

python3 << 'PYEOF' import subprocess, os ocr_dir = '/home/daytona/workspace/ocr_output' all_text = "" for i in range(21, 50): ppm_file = os.path.join(ocr_dir, f'page-{i:03d}.ppm') if os.path.exists(ppm_file): r = subprocess.run(['tesseract', ppm_file, 'stdout', '-l', 'eng'], capture_output=True, text=True, timeout=25) all_text += f"\n\n=== PAGE {i} ===\n{r.stdout}" if i % 10 == 0: print(f"p{i} done") with open(f'{ocr_dir}/pages_21_49.txt','w') as f: f.write(all_text) print("DONE") PYEOF

ls /home/daytona/workspace/ocr_output/ 2>/dev/null | head -20 && echo "---" && ls /home/daytona/workspace/ocr_output/*.txt 2>/dev/null

wc -l /home/daytona/workspace/ocr_output/pages_1_20.txt /home/daytona/workspace/ocr_output/pages_21_49.txt 2>/dev/null && echo "---" && ls /home/daytona/workspace/ocr_output/page-*.ppm | wc -l

python3 << 'PYEOF' import subprocess, os ocr_dir = '/home/daytona/workspace/ocr_output' all_text = "" for i in range(50, 106): ppm_file = os.path.join(ocr_dir, f'page-{i:03d}.ppm') if os.path.exists(ppm_file): r = subprocess.run(['tesseract', ppm_file, 'stdout', '-l', 'eng'], capture_output=True, text=True, timeout=25) all_text += f"\n\n=== PAGE {i} ===\n{r.stdout}" if i % 10 == 0: print(f"p{i} done") with open(f'{ocr_dir}/pages_50_105.txt','w') as f: f.write(all_text) print("DONE") PYEOF

cat /home/daytona/workspace/ocr_output/pages_1_20.txt /home/daytona/workspace/ocr_output/pages_21_49.txt /home/daytona/workspace/ocr_output/pages_50_105.txt > /home/daytona/workspace/ocr_output/FULL_TEXT.txt && wc -l /home/daytona/workspace/ocr_output/FULL_TEXT.txt