Integrating API

In this section, we'll discuss how to integrate the Gmail API and the OpenAI API to make these two functionalities work together.

Main Program

Let's first take a look at the structure of the main program:

from argparse import ArgumentParser
from datetime import datetime, timedelta

from gmail_api import build_service, get_messages, parse_message
from openai_api import chatgpt_summary
from tqdm import tqdm


def generate_markdown_report(input_text, project_name, date):
    title = f"{project_name} Update Report - {date}"
    markdown_text = input_text.replace('\n', '\n\n')
    markdown_content = f"# {title}\n\n{markdown_text}"
    file_name = f"{project_name}-update-{date}.md"
    with open(file_name, "w", encoding="utf-8") as file:
        file.write(markdown_content)
    return file_name


def main(project_name, time_length):

    current_date = datetime.now()
    after_date = current_date - timedelta(days=time_length)
    after_date = after_date.strftime('%Y/%m/%d')

    service = build_service()
    messages = get_messages(
        service,
        after_date=after_date,
        subject_filter=project_name
    )

    results = [
        parse_message(service, msg['id'])
        for msg in tqdm(messages)
    ]

    # Call OpenAI API
    summary = chatgpt_summary(results)
    summary = f'{summary}\n\n---\n\nThe above report was automatically generated by OpenAI GPT-3.5 Turbo model.'

    # Generate Markdown report
    markdown_file = generate_markdown_report(
        summary, project_name, current_date.strftime('%Y-%m-%d'))
    print(f"Markdown file generated: {markdown_file}")


if __name__ == "__main__":
    parser = ArgumentParser(description="Generate a project update report.")
    parser.add_argument(
        "--project_name",
        type=str,
        help="The name of the project to track.",
        default="Albumentations"
    )
    parser.add_argument(
        "--time_length",
        type=int,
        help="The time length (in days) to track updates.",
        default=1
    )

    args = parser.parse_args()

    main(args.project_name, args.time_length)

Introduction to the Main Program

1. Importing Modules

At the beginning of the program, necessary Python modules and custom functions are imported:

• ArgumentParser for parsing command-line arguments.
• datetime and timedelta for handling date and time calculations.
• build_service, get_messages, parse_message for handling Gmail API operations.
• chatgpt_summary for invoking the OpenAI API to generate summaries of email contents.
• tqdm for displaying progress bars.

2. Generating Reports

The generate_markdown_report function is responsible for generating Markdown-formatted report files:

• It takes input_text (summary of email content), project_name (name of the project), and date as parameters.
• It replaces newline characters in input_text with Markdown line breaks.
• It generates the title and content of the Markdown file, then writes it to a .md file named based on the project name and date.

3. Main Program

The main program executes the following steps:

• Sets the current date and past date (calculated based on the time_length parameter).
• Builds the Gmail API service using the build_service function.
• Calls get_messages to retrieve emails containing the specified project name since the specified date.
• Uses tqdm to display a progress bar for email parsing.
• Parses the content of each email and sends it to the OpenAI API for summarization.
• Adds the summary content to a Markdown file and generates that file.

4. Command-Line Argument Handling

In the input section, the program parses two command-line arguments:

• --project_name: The name of the project specified by the user, defaulting to "Albumentations".
• --time_length: The length of time (in days) to track updates, defaulting to 1 day.

These parameters allow users to customize the project name and duration for generating reports.