cwv-lighthouse-scores

This project processes URLs to fetch Lighthouse scores using the Google PageSpeed Insights API.

Setup

1. Clone the repository:

   git clone https://github.com/yourusername/my_project.git
   cd my_project

2. Create a virtual environment and activate it:

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

3. Install the dependencies:

   pip install -r requirements.txt

4. Add your Google API key: Update the API_KEY in config.py with your Google PageSpeed Insights API key.

5. Place your input data: Ensure that your cwv.csv file is in the data directory.

URL Format Requirements

• All URLs in the cwv.csv file must be in the format https://www.example.com.
• Ensure that each URL starts with https:// and includes www. to avoid any issues with API requests.

Platform Column

• The cwv.csv file should include a platform column, which differentiates between "Carrot" and "Non-Carrot" sites.
• This data is directly gleaned from the script output at carrot-serp-compare. The TRUE and FALSE values from this script need to be turned into "Carrot" and "Non-Carrot" respectively.
• The reason for this differentiation is to provide mean CWV (Core Web Vitals) scores for Carrot vs Non-Carrot sites in the comparison file at the end of the processing.

Running the Script

1. Run the main script:

   python main.py

Output

• The processed Lighthouse scores will be saved in data/lighthouse_scores.csv.
• The comparison results will be saved in data/comparison_results.csv.
• Errors and logs will be saved in logs/errors.log.

Parallel Processing

• The current setup uses parallel processing to speed up the fetching of Lighthouse scores.
• URLs are processed concurrently using Python's concurrent.futures.ThreadPoolExecutor, which significantly reduces the total processing time. That said, the script does take a long time to execute when there are thousands of URLs. When ran in a virtual environment, your console will provide updates, such as X/Y processed.
• By default, the script uses 10 threads to handle multiple requests in parallel. This can be adjusted by modifying the max_workers parameter in the main.py script.