#!/usr/bin/env python3
"""
Generate sample test documents for RAGStack-Lambda testing.
Requires:
- PyMuPDF (pip install pymupdf)
- Pillow (pip install Pillow)
- python-docx (pip install python-docx)
- openpyxl (pip install openpyxl)
"""
from pathlib import Path
def generate_text_native_pdf():
"""Generate text-native PDF with embedded text."""
try:
import fitz # PyMuPDF
except ImportError:
print("⚠️ PyMuPDF not installed. Install with: pip install pymupdf")
return False
print("Creating text-native.pdf...")
doc = fitz.open()
page = doc.new_page()
# Add text content
text_content = """Sample Text-Native PDF
This is a test document with embedded text.
It should be processed using direct text extraction, not OCR.
Date: 2025-01-15
Status: Test Document
Type: Text-Native PDF
Content:
This document contains regular text that is embedded in the PDF format.
When processed by RAGStack-Lambda, it should be detected as text-native
and skip the OCR step entirely.
Features tested:
- Direct text extraction
- Fast processing time (< 5 seconds)
- Cost optimization (no OCR charges)
- Page rendering for image embeddings
Additional Information:
This multi-page document tests pagination and chunking.
Each page should be processed sequentially, and text should be
extracted efficiently without requiring OCR services.
Expected Results:
- is_text_native: true
- total_pages: 1
- processing_time: < 5 seconds
- ocr_cost: $0.00
- embedding_cost: ~$0.001
Test Scenario:
Upload this file via the UI and verify it processes correctly.
Check the dashboard to confirm text-native detection worked.
"""
# Insert text at position (50, 50)
page.insert_text((50, 50), text_content, fontsize=11)
# Save PDF
doc.save("text-native.pdf")
doc.close()
print("✓ Created text-native.pdf")
return True
def generate_invoice_image():
"""Generate sample invoice image."""
try:
from PIL import Image, ImageDraw, ImageFont
except ImportError:
print("⚠️ Pillow not installed. Install with: pip install Pillow")
return False
print("Creating invoice.jpg...")
# Create image
img = Image.new("RGB", (800, 1000), color="white")
draw = ImageDraw.Draw(img)
# Try to use default font
try:
font_large = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf", 24)
font_normal = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf", 14)
except OSError:
font_large = ImageFont.load_default()
font_normal = ImageFont.load_default()
# Draw invoice content
y = 50
# Header
draw.text((50, y), "INVOICE", fill="black", font=font_large)
y += 60
# Invoice details
invoice_text = [
"Invoice #: INV-2025-001",
"Date: January 15, 2025",
"Due Date: February 15, 2025",
"",
"Bill To:",
"Acme Corporation",
"123 Main Street",
"Springfield, USA 12345",
"",
"Items:",
"",
"1. Professional Services $1,500.00",
"2. Cloud Infrastructure $750.00",
"3. Technical Support $250.00",
"",
" Subtotal: $2,500.00",
" Tax (10%): $250.00",
" Total: $2,750.00",
"",
"Payment Terms: Net 30 days",
"Thank you for your business!",
]
for line in invoice_text:
draw.text((50, y), line, fill="black", font=font_normal)
y += 25
# Save image
img.save("invoice.jpg", "JPEG", quality=95)
print("✓ Created invoice.jpg")
return True
def generate_word_document():
"""Generate sample Word document."""
try:
from docx import Document
except ImportError:
print("⚠️ python-docx not installed. Install with: pip install python-docx")
return False
print("Creating document.docx...")
doc = Document()
# Add title
doc.add_heading("Sample Word Document", 0)
# Add content
doc.add_paragraph("This is a sample Microsoft Word document for testing RAGStack-Lambda.")
doc.add_heading("Purpose", level=1)
doc.add_paragraph(
"This document tests the format conversion pipeline. "
"Word documents should be converted to PDF format before OCR processing."
)
doc.add_heading("Test Criteria", level=1)
doc.add_paragraph("1. Document should be converted to PDF")
doc.add_paragraph("2. Text should be extracted correctly")
doc.add_paragraph("3. Formatting should be preserved where possible")
doc.add_heading("Expected Results", level=1)
doc.add_paragraph("• File format: DOCX → PDF")
doc.add_paragraph("• Text extraction: Successful")
doc.add_paragraph("• Processing time: < 15 seconds")
# Save document
doc.save("document.docx")
print("✓ Created document.docx")
return True
def generate_excel_spreadsheet():
"""Generate sample Excel spreadsheet."""
try:
from openpyxl import Workbook
except ImportError:
print("⚠️ openpyxl not installed. Install with: pip install openpyxl")
return False
print("Creating spreadsheet.xlsx...")
wb = Workbook()
ws = wb.active
ws.title = "Sales Data"
# Add headers
headers = ["Date", "Product", "Quantity", "Price", "Total"]
ws.append(headers)
# Add data
data = [
["2025-01-01", "Widget A", 10, 25.00, 250.00],
["2025-01-02", "Widget B", 5, 50.00, 250.00],
["2025-01-03", "Widget C", 15, 10.00, 150.00],
["2025-01-04", "Widget A", 20, 25.00, 500.00],
["2025-01-05", "Widget B", 8, 50.00, 400.00],
]
for row in data:
ws.append(row)
# Add totals
ws.append([])
ws.append(["", "", "Total", "", "=SUM(E2:E6)"])
# Save spreadsheet
wb.save("spreadsheet.xlsx")
print("✓ Created spreadsheet.xlsx")
return True
def main():
print("=" * 60)
print("RAGStack-Lambda Sample Document Generator")
print("=" * 60)
print()
# Change to script directory
script_dir = Path(__file__).parent
script_dir.mkdir(parents=True, exist_ok=True)
import os
os.chdir(script_dir)
results = {
"text-native.pdf": generate_text_native_pdf(),
"invoice.jpg": generate_invoice_image(),
"document.docx": generate_word_document(),
"spreadsheet.xlsx": generate_excel_spreadsheet(),
}
print()
print("=" * 60)
print("Summary:")
print("=" * 60)
success_count = sum(results.values())
total_count = len(results)
for filename, success in results.items():
status = "✓" if success else "✗"
print(f"{status} {filename}")
print()
print(f"Created {success_count}/{total_count} files successfully")
if success_count < total_count:
print()
print("Install missing dependencies with:")
print(" pip install pymupdf Pillow python-docx openpyxl")
print()
print("Note: For scanned.pdf, you'll need to provide an actual scanned document")
print("or use an online tool to convert an image to a scanned PDF.")
if __name__ == "__main__":
main()
