newsletter-html

A newsletter HTML generator that auto-loads tech news using AI. Fetches structured content from the Perplexity API, then fills an HTML template to produce a ready-to-view or send newsletter page.

Features

• AI-powered content — Uses Perplexity’s API to pull recent tech news (past 24 hours, viral stories, company developments).
• Structured output — Response is validated and shaped with Pydantic so the HTML template always gets the expected fields.
• Template-based HTML — Placeholders in an HTML template are replaced with the fetched data.
• Dated outputs — Each run saves a copy under outputs/ with a date and run number (e.g. template_today__number_1_14_03_2025.html).
• Logging — Writes logs under logs/ for debugging and auditing.

Prerequisites

• Python 3.x
• Perplexity API key — Get one from Perplexity (used for the “sonar” model).

Installation

1. Clone the repository

   git clone https://github.com/prishadhoot/newsletter-html.git
   cd newsletter-html

2. Create a virtual environment (recommended)

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # macOS/Linux
   source venv/bin/activate

3. Install dependencies

   pip install -r requirements.txt

4. Configure environment

   Create a .env file in the project root:

   PERPLEXITY_API_KEY=your_perplexity_api_key_here

   Do not commit .env; it is listed in .gitignore.

Usage

From the project root, run:

python main.py

This will:

1. Fetch tech news from the Perplexity API using the prompt in utils/templates/pplx_prompt.txt.
2. Validate and normalize the response against utils/templates/response_template.json.
3. Load the HTML template from utils/templates/template_placeholder.html.
4. Replace placeholders ({24h_1} … {24h_6}, {viral_1} … {viral_3}, {company_developments}) with the API data.
5. Save the result as utils/templates/template_today.html and a dated copy in outputs/.

Open main.html (or the generated HTML files) in a browser to view the newsletter.

Project structure

Path	Description
main.py	Entry point: logging, API fetch, data processing, template fill, and file output.
newsletter.py	Newsletter logic (load data, template, fill, save) — used conceptually; main.py embeds this flow.
data_process.py / data_query.py	Logic is inlined in main.py: API client, Pydantic models, validation, and template matching.
main.html	Front-end page that displays or loads the newsletter.
utils/templates/	pplx_prompt.txt, response_template.json, template_placeholder.html, and generated template_today.html.
outputs/	Dated newsletter HTML files (one per run).
logs/	Created at runtime; log files for each run.

Configuration

• API key — Set PERPLEXITY_API_KEY in .env.
• Prompt — Edit utils/templates/pplx_prompt.txt to change what kind of tech news is requested.
• Response shape — Edit utils/templates/response_template.json to match the structure expected by your template and by main.py’s validation.
• Layout — Edit utils/templates/template_placeholder.html and use the same placeholder names so the script can fill them.