Azure AI Document Intelligence to HTML Parser

A lightweight Python library for converting Azure AI Document Intelligence AnalyzeResult objects into well-formatted HTML documents.

Features

• 📄 Clean HTML Generation: Converts document analysis results to semantic HTML
• 🎨 Customizable Styling: Support for custom CSS styling
• 📊 Table Support: Handles complex tables with row/column spans
• 📝 Section Preservation: Maintains document structure with sections and paragraphs
• 🔧 Flexible API: Works with JSON files, stdin/stdout, or Python objects

Installation

pip install azure-ai-documentintelligence-html

Quick Start

From Python Code

from azure.ai.documentintelligence import DocumentIntelligenceClient
from azure.core.credentials import AzureKeyCredential
from azure_ai_doc_intel_html import parse_analyze_result

# Analyze document with Azure Document Intelligence
client = DocumentIntelligenceClient(
    endpoint="<your-endpoint>",
    credential=AzureKeyCredential("<your-key>")
)

poller = client.begin_analyze_document("prebuilt-layout", document_url)
result = poller.result()

# Convert to HTML
html = parse_analyze_result(result)

# Save to file
with open("output.html", "w") as f:
    f.write(html)

# Or save directly
parse_analyze_result(result, output_html="output.html")

From Command Line

# From JSON file to HTML file
python -m azure_ai_doc_intel_html input.json output.html

# From stdin to stdout
cat input.json | python -m azure_ai_doc_intel_html > output.html

From JSON File

from azure_ai_doc_intel_html import parse

# Parse from JSON file
html = parse("analyze_result.json")

# Save to file
parse("analyze_result.json", output_html="output.html")

# With custom title
html = parse("analyze_result.json", title="My Document Analysis")

# With custom CSS
custom_css = """
    body { font-family: 'Segoe UI', Tahoma, sans-serif; }
    h1 { color: #0078D4; }
    table { border-color: #0078D4; }
"""
html = parse("analyze_result.json", custom_css=custom_css)

API Reference

Core Functions

parse_analyze_result(analyze_result, output_html=None, title=None, custom_css=None)

Convert an AnalyzeResult object to HTML.

Parameters:

• analyze_result: AnalyzeResult object from Azure AI Document Intelligence
• output_html: Optional path to save HTML file
• title: Optional custom title (defaults to auto-detection from document)
• custom_css: Optional custom CSS styles

Returns: HTML string

parse(path_or_stdin, output_html=None, title=None, custom_css=None)

Parse JSON file or stdin to HTML.

Parameters:

• path_or_stdin: Path to JSON file or None for stdin
• output_html: Optional path to save HTML file
• title: Optional custom title
• custom_css: Optional custom CSS styles

Returns: HTML string

Helper Functions

• load_json(path_or_stdin): Load JSON from file or stdin
• write_html(path_or_stdout, html): Write HTML to file or stdout
• render_table_html(table): Render table object to HTML
• render_paragraph_html(paragraph): Render paragraph object to HTML
• render_sections_html(data): Render all sections to HTML
• build_html_doc(title, sections_html, custom_css): Build complete HTML document

Output Structure

The generated HTML includes:

• Document Title: Extracted from page headers or custom-specified
• Sections: Logical document sections with proper hierarchy
• Paragraphs: Text content with role-based styling (headers, footers, body text)
• Tables: Complex tables with proper row/column spans and headers
• Semantic HTML: Uses appropriate HTML5 elements for better accessibility

Default CSS Classes

• .doc-table: Document tables
• .section: Document sections
• .page-header: Page header content
• .page-footer: Page footer content
• .summary: Summary sections with definition lists

Custom Styling Example

custom_css = """
    body { 
        font-family: 'Segoe UI', Tahoma, Geneva, sans-serif;
        max-width: 1200px;
        margin: 0 auto;
        padding: 20px;
    }
    h1 { 
        color: #0078D4;
        border-bottom: 2px solid #0078D4;
        padding-bottom: 10px;
    }
    .doc-table { 
        box-shadow: 0 2px 4px rgba(0,0,0,0.1);
        border-radius: 4px;
        overflow: hidden;
    }
    .doc-table th { 
        background: #0078D4;
        color: white;
    }
    .section { 
        background: #f9f9f9;
        padding: 15px;
        margin: 20px 0;
        border-radius: 4px;
    }
"""

html = parse_analyze_result(result, custom_css=custom_css)

Integration with Azure Document Intelligence

This library is designed to work seamlessly with Azure AI Document Intelligence. Here's a complete example:

from azure.ai.documentintelligence import DocumentIntelligenceClient
from azure.core.credentials import AzureKeyCredential
from azure_ai_doc_intel_html import parse_analyze_result

# Setup client
endpoint = "https://your-resource.cognitiveservices.azure.com/"
key = "your-api-key"
client = DocumentIntelligenceClient(endpoint, AzureKeyCredential(key))

# Analyze document
with open("document.pdf", "rb") as f:
    poller = client.begin_analyze_document(
        "prebuilt-layout",
        analyze_request=f,
        content_type="application/pdf"
    )
result = poller.result()

# Convert to HTML with custom title
html = parse_analyze_result(
    result,
    title="Financial Report Analysis",
    output_html="report.html"
)

print(f"HTML generated: {len(html)} characters")

Supported Document Elements

• ✅ Paragraphs with roles (headers, footers, section headings)
• ✅ Tables with complex spans
• ✅ Sections with hierarchical structure
• ✅ Text content with proper escaping
• ✅ Page structure preservation

Requirements

• Python 3.7+
• azure-ai-documentintelligence>=1.0.0

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

For issues and questions, please use the GitHub issue tracker.