/ 5 min read / 1139 words

Automated-Marketing - A Rust command-line tool that extracts business websites and contact emails and send them mockups.

aryansrao/automated-marketing

Automated Marketing System

A Rust command-line tool that crawls directory listings (e.g., Yellow Pages US/AU), extracts business websites and contact emails, optionally generates a visual mockup and sends a marketing email with an inline image. It demonstrates web scraping, DOM parsing, headless browser rendering, and SMTP email sending.

Repository: https://github.com/aryansrao/automated-marketing

Table of Contents

Overview

automated-marketing is a Rust-based command-line tool that crawls directory listings (e.g., Yellow Pages US/AU), extracts business websites and contact emails and generates a mockup image for each matched business, and sends a marketing email containing the mockup. The project demonstrates web scraping, HTML parsing, headless browser rendering, and SMTP email sending in Rust.

This repository contains a single binary implemented in src/main.rs.

Features

  • Configurable search targets and limits for US and Australia Yellow Pages.
  • Concurrent, cancellable run loop (press q + Enter to stop).
  • Email extraction using a safe regular expression with filtering to avoid false positives.
  • Optional email automation with HTML content and an inline image attachment.
  • Mockup generation via a headless Chrome instance (screenshotting a locally rendered HTML mockup).
  • Leads exported to CSV leads.csv at the end of a run.

Architecture

  • AutomatedMarketing struct: manages HTTP client, email regex, stop flag, and email config.
  • Scraping: reqwest for HTTP, scraper for DOM querying.
  • Mockup generation: headless_chrome to render HTML and capture PNG screenshots.
  • Emailing: lettre for SMTP send with inline image attachments.
  • Persistence: CSV file output via save_to_csv().

See the main implementation at src/main.rs.

Prerequisites

  • Rust toolchain (stable) and Cargo installed. See https://www.rust-lang.org/tools/install
  • Google Chrome or Chromium installed (required by headless_chrome).
  • Network access for web scraping and SMTP (if using email automation).

Installation

Clone the repository and build with Cargo:

bash 
git clone https://github.com/aryansrao/automated-marketing.git
cd automated-marketing
cargo build --release

Or run it directly in dev mode:

bash 
cargo run --release

Configuration (.env)

The program can optionally send emails. To enable email sending, create a .env file in the project root with the following variables. The program reads these variables via dotenv.

text 
SENDER_EMAIL=you@example.com
SENDER_PASSWORD=your_smtp_password_or_app_specific_password
SENDER_NAME=Your Name or Company
PHONE_NUMBER=+1234567890

Notes:

  • For Gmail, you should create an App Password or configure the account to allow SMTP with the credentials you provide. The code uses Gmail’s SMTP server (smtp.gmail.com with STARTTLS on port 587).
  • If .env is missing or variables are not set, email automation is disabled and the program will only collect leads and produce leads.csv.

Build & Run

To build:

bash 
cargo build --release

To run (release build recommended):

bash 
./target/release/automated-marketing

Or using Cargo directly:

bash 
cargo run --release

Operation:

  • The process prints status for each search and each found lead.
  • Press q and Enter at any time to request a graceful stop.

Output:

  • leads.csv is written to the working directory when the process completes or is stopped.

Usage

  1. Optionally configure .env with SMTP credentials if you want email automation.
  2. Run the binary.
  3. Monitor console output for found leads and email sending status.
  4. When finished (or stopped via q), check leads.csv for all collected leads.

Example run (no email):

bash 
cargo run --release

Example run (with email):

bash 
# ensure .env is present and configured
cargo run --release

How It Works

High-level flow in src/main.rs:

  1. Build an HTTP client with a timeout and UA string.
  2. Compile an email regular expression for extraction and prepare an optional EmailConfig from environment.
  3. get_all_searches() returns search URLs for US and AU Yellow Pages with categories and pages.
  4. For each search URL, search_directory() fetches the page and extracts business names/website links.
  5. For each site, scrape_business() fetches the business site and extracts emails from the HTML.
  6. If an email is found and email automation is enabled, generate_mockup_image() renders a simple HTML mockup and captures a PNG via headless_chrome.
  7. send_email() constructs a multipart HTML email with inline PNG and sends via SMTP using lettre.
  8. Leads are collected into memory and written to leads.csv at the end.

Key code locations:

  • Main program: src/main.rs
  • Entry point: fn main() (Tokio runtime and stop-signal loop)

Email & Mockup Generation

  • The mockup is generated by writing a temporary mockup.html to the system temp directory and using headless_chrome to open and screenshot the mockup element with id #mockup.
  • The screenshot is then attached inline to the email as a PNG (cid:mockup-image).
  • The email uses lettre with STARTTLS to smtp.gmail.com:587 by default.

Caveats:

  • headless_chrome requires a compatible Chrome/Chromium binary on PATH. If Chrome is missing, mockup generation will fail.
  • Running a headless browser is resource-intensive; consider limiting concurrency or running on a machine with sufficient memory and CPU.

Persistence & Output

  • The CSV output leads.csv contains a header and rows with company_name,company_type,country,website,email.
  • CSV fields containing quotes are escaped by doubling quotes.

Testing

There are no unit tests included in the repo currently. To perform basic verification manually:

  1. Run the program without .env and confirm it discovers leads and writes leads.csv.
  2. Add a .env with valid SMTP credentials and run again to verify emails are attempted and sent.

If you want tests added, open an issue or provide guidance on which components to cover (parsing, email validation, CSV output, etc.).

Troubleshooting

  • If scraping returns zero results for a site, check network connectivity and possible site blocking or changes to Yellow Pages DOM structure.
  • If headless_chrome fails to launch, ensure Chrome/Chromium is installed and accessible, and that you have necessary permissions.
  • If email sending fails, inspect the console error printed by the program. Common causes: invalid credentials, blocked SMTP by provider, or missing STARTTLS support.
  • If you see many false-positive emails, update the email regex in src/main.rs and refine is_valid_email() filtering.

Security & Privacy

  • This tool collects publicly available emails from business websites. Respect local laws and anti-spam regulations (e.g., CAN-SPAM, GDPR). Use responsible, opt-in practices.
  • Store SMTP credentials securely and avoid committing .env to source control. Consider using secrets management for production usage.

Contributing

Contributions are welcome. Suggested improvements:

  • Add unit and integration tests for parsing and email filters.
  • Make search targets configurable via a config file or command-line flags.
  • Add rate-limiting and retries with backoff for robustness.
  • Improve mockup template and allow HTML customization per campaign.

To contribute:

  1. Fork the repository.
  2. Create a feature branch.
  3. Open a pull request with a clear description of changes.
All posts