AI Documentation Analyzer

This tool analyzes public documentation (e.g., MoEngage docs) and provides a structured report evaluating:

1. Readability
2. Structure & Flow
3. Completeness & Examples
4. Style Guidelines

Optionally, it can also revise the content based on suggestions generated.

---

Features

• Evaluate readability using metrics like Flesch Reading Ease and Gunning Fog Index.
• Assess structural quality: headings, paragraph flow, and list usage.
• Check completeness: presence of explanations, examples, and step-by-step instructions.
• Evaluate writing style against the Microsoft Style Guide.
• Optionally rewrite content to improve clarity, flow, and readability.

---

Project Structure

document_analyzer/
│
├── analyzer/                 # Core orchestration scripts
│   ├── __init__.py
│   ├── analyzer.py           # Main CLI entry point
│   ├── analyzer_logic.py     # Orchestrates various assessment modules
│   ├── report_generator.py   # Aggregates results into structured output
│   └── revision_agent.py     # Optional agent to revise text
│
├── modules/                  # Individual functional modules
│   ├── __init__.py
│   ├── completeness.py       # Completeness & examples checker
│   ├── readability.py        # Readability metrics calculator
│   ├── structure.py          # Checks structure and flow
│   ├── style.py              # Evaluates style guideline adherence
│   ├── scraper.py            # Scrapes content using Selenium
│   └── utils.py              # Helper functions (e.g., text cleaner)
│
├── reports/                  # Generated or example reports
│   ├── example_report.json
│   └── report.json
│
├── requirements.txt          # Python package dependencies
├── README.md                 # Project documentation
├── LICENSE
└── venv/                     # Virtual environment (optional)

---

Installation

1. Clone the repository

git clone https://github.com/saniyaacharya04/document_analyzer.git
cd document_analyzer

2. Create and activate a virtual environment (optional but recommended)

python -m venv venv
source venv/bin/activate   # On Windows: venv\Scripts\activate

3. Install dependencies

pip install --upgrade setuptools  # Ensure pkg_resources is available
pip install -r requirements.txt

---

Usage

Analyze a Documentation URL

python -m analyzer.analyzer "https://help.moengage.com/hc/en-us/articles/4415622460948-Push-Templates" -o reports/report.json

This command will:

• Fetch the article via Selenium
• Analyze it across Readability, Structure, Completeness, and Style
• Save a structured JSON report to reports/report.json

---

Optional: Use the shell shortcut script

If you created run_analysis.sh:

./run_analysis.sh "https://help.moengage.com/hc/en-us/articles/4415622460948-Push-Templates"

The report will be saved automatically in reports/report.json.

---

Example Output

{
  "url": "https://help.moengage.com/hc/en-us/articles/4415622460948-Push-Templates",
  "readability": {
    "flesch_reading_ease": 47.43,
    "gunning_fog_index": 9.47,
    "feedback": "Moderate difficulty."
  },
  "structure_and_flow": {
    "assessment": "Add more headings for better navigation. Use bullet or numbered lists to improve readability."
  },
  "completeness_and_examples": {
    "assessment": "No clear examples or step-by-step instructions detected. Consider adding some."
  },
  "style_guidelines": {
    "assessment": "Style is clear, concise, and user-friendly."
  }
}

---

Optional: Content Rewriting Agent

You can use revision_agent.py to revise documentation with suggested improvements:

python analyzer/revision_agent.py input_article.txt suggestions.json -o reports/revised_output.md

---

Notes

• Uses Selenium with headless Chrome for scraping content that blocks requests.
• Ensure Google Chrome is installed and chromedriver is configured in your PATH.
• The textstat warning about pkg_resources is harmless and can be ignored.

---

Style Guide

Style recommendations are based on the Microsoft Writing Style Guide:

• Clear and concise writing
• Friendly, conversational tone
• Step-by-step instructions instead of passive descriptions
• Consistent use of headings, lists, and examples

---

License

This project is licensed under the MIT License.