OddsHarvester

OddsHarvester is an application designed to scrape and process sports betting odds and match data from oddsportal.com website.

📖 Table of Contents

1. ✨ Features
2. 🛠️ Local Installation
3. ⚡ Usage
   • 🔧 CLI Commands
   • 🐳 Running Inside a Docker Container
   • ☁️ Cloud Deployment
4. ⚙️ Configuration
5. 🤝 Contributing
6. 📜 License
7. 💬 Feedback
8. ❗ Disclaimer

✨ Features

• 📅 Scrape Upcoming Matches: Fetch odds and event details for upcoming sports matches.
• 📊 Scrape Historical Odds: Retrieve historical odds and match results for analytical purposes.
• 🔍 Advanced Parsing: Extract structured data, including match dates, team names, scores, and venue details.
• 💾 Flexible Storage: Store scraped data in JSON or CSV locally, or upload it directly to a remote S3 bucket.
• 🐳 Docker Compatibility: Designed to work seamlessly inside Docker containers with minimal setup.
• 🕵️ Proxy Support: Route web requests through SOCKS/HTTP proxies for enhanced anonymity, geolocation bypass, and anti-blocking measures.

📚 Current Support

OddsHarvester supports a growing number of sports and their associated betting markets. All configurations are managed via dedicated enum and mapping files in the codebase.

✅ Supported Sports & Markets

🏅 Sport	🛒 Supported Markets
⚽ Football	1x2, btts, double_chance, draw_no_bet, over/under, european_handicap, asian_handicap
🎾 Tennis	match_winner, total_sets_over/under, total_games_over/under, asian_handicap, exact_score
🏀 Basketball	1x2, moneyline, asian_handicap, over/under

⚙️ Note: Each sport and its markets are declared in enums inside sport_market_constants.py.

🗺️ Leagues & Competitions

Leagues and tournaments are mapped per sport in:
sport_league_constants.py

You’ll find support for:

• 🏆 Top Football leagues (Premier League, La Liga, Serie A, etc.)
• 🎾 Major Tennis tournaments (ATP, WTA, Grand Slams, etc.)
• 🏀 Global Basketball leagues (NBA, EuroLeague, ACB, etc.)

🛠️ Local Installation

1. Clone the repository: Navigate to your desired folder and clone the repository. Then, move into the project directory:

   git clone https://github.com/jordantete/OddsHarvester.git
   cd OddsHarvester

2. Quick Setup with uv:

   Use uv, a lightweight package manager, to simplify the setup process. First, install uv with pip, then run the setup:

   pip install uv
   uv sync

3. Manual Setup (Optional):

   If you prefer to set up manually, follow these steps:

   • Create a virtual environment: Use Python's venv module to create an isolated environment (or virtualenv) for the project. Activate it depending on your operating system:

     • pvython3 -m venv .venv

     • On Unix/MacOS: source venv/bin/activate

     • On Windows:
       venv\Scripts\activate

   • Install dependencies with pip: Use pip with the --use-pep517 flag to install directly from the pyproject.toml file: pip install . --use-pep517.

   • Or install dependencies with poetry: If you prefer poetry for dependency management: poetry install

4. Verify Installation:

   Ensure all dependencies are installed and Playwright is set up by running the following command:

   cd src
   python main.py --help

By following these steps, you should have OddsHarvester set up and ready to use.

⚡ Usage

🔧 CLI Commands

OddsHarvester provides a Command-Line Interface (CLI) to scrape sports betting data from oddsportal.com. Use it to retrieve upcoming match odds, analyze historical data, or store results for further processing. Below are the available commands and their options:

1. Scrape Upcoming Matches

Retrieve odds and event details for upcoming sports matches.

Options:

🏷️ Option	📝 Description	🔐 Required	🔧 Default
--sport	Specify the sport to scrape (e.g., football).	✅	None
--date	Date for matches in YYYYMMDD format (e.g., 20250227).	❌	None
--league	Specify the league to scrape (e.g., england-premier-league).	❌	None
--markets	Comma-separated betting markets (e.g., 1x2,btts).	❌	1x2
--storage	Save data locally or to a remote S3 bucket (local or remote).	❌	local
--file_path	File path to save data locally (e.g., output.json).	❌	scraped_data.json
--format	Format for saving local data (json or csv).	❌	json
--headless	Run the browser in headless mode (True or False).	❌	False
--save_logs	Save logs for debugging purposes (True or False).	❌	False
--proxies	List of proxies in "server user pass" format. Multiple proxies supported.	❌	None
--browser_user_agent	Custom user agent string for browser requests.	❌	None
--browser_locale_timezone	Browser locale timezone (e.g., fr-BE).	❌	None
--browser_timezone_id	Browser timezone ID (e.g., Europe/Brussels).	❌	None
--match_links	List of specific match links to scrape (overrides other filters).	❌	None

📌 Important Notes:

• If both --league and --date are provided, the scraper will only consider the league, meaning all upcoming matches for that league will be scraped, regardless of the --date argument.
• If --match_links is provided, it overrides --sport, --date, and --league, and only the specified match links will be scraped.
• All match links must belong to the same sport when using --match_links.
• For best results, ensure the proxy’s region matches the BROWSER_LOCALE_TIMEZONE and BROWSER_TIMEZONE_ID settings.

Example Usage:

• Retrieve upcoming football matches for January 1, 2025, and save results locally:

python main.py scrape_upcoming –sport football –date 2025-01-01

• Scrapes English Premier League matches with odds for 1x2 and Both Teams to Score (BTTS):

python main.py scrape_upcoming --sport football --league england-premier-league --markets 1x2,btts --storage local

• Scrapes football matches using a rotating proxy setup:

python main.py scrape_upcoming --sport football --date 20250227 --proxies "http://proxy1.com:8080 user1 pass1" "http://proxy2.com:8080 user2 pass2"

2. Scrape Historical Odds

Retrieve historical odds and results for analytical purposes.

Options:

🏷️ Option	📝 Description	🔐 Required	🔧 Default
--sport	Specify the sport to scrape (e.g., football).	✅	None
--league	Specify the league to scrape (e.g., england-premier-league).	✅	None
--season	Target season in YYYY-YYYY format (e.g., 2022-2023).	✅	None
--markets	Comma-separated betting markets (e.g., 1x2,btts).	❌	1x2
--storage	Save data locally or to a remote S3 bucket (local or remote).	❌	local
--file_path	File path to save data locally (e.g., output.json).	❌	scraped_data.json
--format	Format for saving local data (json or csv).	❌	json
--max_pages	Maximum number of pages to scrape.	❌	None
--headless	Run the browser in headless mode (True or False).	❌	False
--save_logs	Save logs for debugging purposes (True or False).	❌	False
--proxies	List of proxies in "server user pass" format. Multiple proxies supported.	❌	None
--browser_user_agent	Custom user agent string for browser requests.	❌	None
--browser_locale_timezone	Browser locale timezone (e.g., fr-BE).	❌	None
--browser_timezone_id	Browser timezone ID (e.g., Europe/Brussels).	❌	None
--match_links	List of specific match links to scrape (overrides other filters).	❌	None

Example Usage:

• Retrieve historical odds for the Premier League's 2022-2023 season:

python main.py scrape_historic –league premier-league –season 2022-2023

• Scrapes only 3 pages of historical odds data:

python main.py scrape_historic --sport football --league england-premier-league --season 2022-2023 --max_pages 3

📌 Running the Help Command:

To display all available CLI commands and options, run:

uv run python main.py --help

🐳 Running Inside a Docker Container

OddsHarvester is compatible with Docker, allowing you to run the application seamlessly in a containerized environment.

Steps to Run with Docker:

1. Ensure Docker is Installed
   Make sure Docker is installed and running on your system. Visit Docker's official website for installation instructions specific to your operating system.

2. Build the Docker Image
   Navigate to the project's root directory, where the Dockerfile is located. Build the Docker image using the appropriate Docker build command.
   Assign a name to the image, such as odds-harvester: docker build -t odds-harvester:local --target local-dev .

3. Run the Container
   Start a Docker container based on the built image. Map the necessary ports if required and specify any volumes to persist data. Pass any CLI arguments (e.g., scrape_upcoming) as part of the Docker run command: docker run --rm odds-harvester:latest python3 -m main scrape_upcoming --sport football --date 20250125 --markets 1x2 --storage local --file_path output.json --headless

4. Interactive Mode for Debugging If you need to debug or run commands interactively: docker run --rm -it odds-harvester:latest /bin/bash

Tips:

• Volume Mapping: Use volume mapping to store logs or output data on the host machine.
• Container Reusability: Assign a unique container name to avoid conflicts when running multiple instances.

☁️ Cloud Deployment

OddsHarvester can also be deployed on a cloud provider using the Serverless Framework, with a Docker image to ensure compatibility with AWS Lambda (Dockerfile will need to be tweaked if you want to deploy on a different cloud provider).

Why Use a Docker Image?

1. AWS Lambda’s Deployment Size Limit: AWS Lambda has a hard limit of 50MB for direct deployment packages, which includes code, dependencies, and assets. Playwright and its browser dependencies far exceed this limit.

2. Playwright’s Incompatibility with Lambda Layers: Playwright cannot be installed as an AWS Lambda layer because: • Its browser dependencies require system libraries that are unavailable in Lambda’s standard runtime environment. • Packaging these libraries within Lambda layers would exceed the layer size limit.

3. Solution: Using a Docker image solves these limitations by bundling the entire runtime environment, including Playwright, its browsers, and all required libraries, into a single package. This ensures a consistent and compatible execution environment.

Serverless Framework Setup:

1. Serverless Configuration:
   The application includes a serverless.yaml file located at the root of the project. This file defines the deployment configuration for a serverless environment. Users can customize the configuration as needed, including:

   • Provider: Specify the cloud provider (e.g., AWS).
   • Region: Set the desired deployment region (e.g., eu-west-3).
   • Resources: Update the S3 bucket details or permissions as required.

2. Docker Integration:
   The app uses a Docker image (playwright_python_arm64) to ensure compatibility with the serverless architecture. The Dockerfile is already included in the project and configured in serverless.yaml. You'll need to build the image locally (see section above) and push the Docker image to ECR.

3. Permissions:
   By default, the app is configured with IAM roles to:

   • Upload (PutObject), retrieve (GetObject), and delete (DeleteObject) files from an S3 bucket.
     Update the Resource field in serverless.yaml with the ARN of your S3 bucket.

4. Function Details:

   • Function Name: scanAndStoreOddsPortalDataV2
   • Memory Size: 2048 MB
   • Timeout: 360 seconds
   • Event Trigger: Runs automatically every 2 hours (rate(2 hours)) via EventBridge.

Customizing Your Configuration: To tailor the serverless deployment for your needs:

• Open the serverless.yaml file in the root directory.
• Update the relevant fields:
  • S3 bucket ARN in the IAM policy.
  • Scheduling rate for the EventBridge trigger.
  • Resource limits (e.g., memory size or timeout).

Deploying to your prefered Cloud provider:

1. Install the Serverless Framework:
   • Follow the installation guide at Serverless Framework.
2. Deploy the application:
   • Use the sls deploy command to deploy the app to your cloud provider.
3. Verify the deployment:
   • Confirm that the function is scheduled correctly and check logs or S3 outputs.

⚙️ Configuration

Constants

OddsHarvester uses a constants.py file to define important parameters for browser configuration and scraping behavior. Users can customize these parameters directly in the file to suit their needs. Key configurable constants include:

• ODDS_FORMAT: Configure the desired odds format (e.g., Decimal Odds, Fractional Odds).
• SCRAPE_CONCURRENCY_TASKS: Adjust the number of concurrent tasks the scraper can handle. Controls how many pages or tasks are processed simultaneously. Increasing this value can speed up scraping but may increase the risk of being blocked by the target website. Use cautiously based on your network and system capabilities.

🤝 Contributing

Contributions are welcome! If you have ideas, improvements, or bug fixes, feel free to submit an issue or a pull request. Please ensure that your contributions follow the project’s coding standards and include clear descriptions for any changes.

📜 License

This project is licensed under the MIT License - see the LICENSE file for more details.

💬 Feedback

Have any questions or feedback? Feel free to reach out via the issues tab on GitHub. We’d love to hear from you!

❗ Disclaimer

This package is intended for educational purposes only and not for any commercial use in any way. The author is not affiliated with or endorsed by the oddsportal.com website. Use this application responsibly and ensure compliance with the terms of service of oddsportal.com and any applicable laws in your jurisdiction.