make_eigen_release_notes_json.py

import csv
import os
import json
import time
from openai import OpenAI

# Initialize the OpenAI client using the new API interface.
client = OpenAI(
    api_key=os.environ['OPENAI_API_KEY']
)

def read_csv(file_path):
    """Read CSV file and return a list of dictionaries."""
    with open(file_path, newline="", encoding="utf-8") as csvfile:
        reader = csv.DictReader(csvfile)
        return list(reader)

def generate_json_for_mr(mr):
    """
    Generate a JSON object for a single merge request.

    The output JSON object has the following structure:
      {
        "#<MR_Number>": {
          "<category>": "[#<MR_Number>](<Link>): <One line summary>"
        }
      }
    where <category> is one of: breaking_changes, major_features, improvements,
    backend_improvements, or other_changes.
    """
    prompt = (
        "You are a release notes categorizer for the Eigen C++ library. "
        "Given the following merge request details, output a JSON object with a single key-value pair. "
        "The key should be the merge request number prefixed with a '#' (for example, '#606'). "
        "The value should be an object with one key, which must be one of: "
        "'breaking_changes', 'major_features', 'improvements', 'backend_improvements', or 'other_changes'. "
        "The value corresponding to that key should be a single-line markdown summary that includes the merge request number "
        "as a markdown link (e.g., [#606](https://gitlab.com/libeigen/eigen/-/merge_requests/606)) followed by a concise description. "
        """
            The output JSON object has the following structure:
          {
            "#<MR_Number>": {
              "<category>": "[#<MR_Number>](<Link>): <One line summary>"
            }
          }
        where <category> is one of: breaking_changes, major_features, improvements,
        backend_improvements, or other_changes.
        """
        "Do not include any extra text or commentary. "
    )
    input_v = ("Here are the merge request details:\n"
        f"Title: {mr.get('title_subject', 'NA')}\n"
        f"Link: {mr.get('mr_id_link', 'NA')}\n"
        f"Summary: {mr.get('summary', 'NA')}\n"
        f"Merge Date: {mr.get('merge_date', 'NA')}\n\n"
        "Output only a valid JSON object.")

    response = client.responses.create(
        model="gpt-4o",
        instructions=prompt,
        input=input_v
    )
    return response.output_text.strip()

def main():
    input_csv = "eigen_release_notes_with_summary.csv"
    output_file = "mr_release_notes.jsonl"
    bad_output_file = "mr_release_notes.txt"
    first_try = True
    # Read the CSV file containing merge request details.
    rows = read_csv(input_csv)

    with open(output_file, "w", encoding="utf-8") as outfile:
      with open(bad_output_file, "w", encoding="utf-8") as bad_outfile:
          for idx, mr in enumerate(rows, start=1):
              if first_try:
                  import pdb; pdb.set_trace()
                  first_try = False
              try:
                  mr_json_str = generate_json_for_mr(mr)
                  mr_json_str = mr_json_str.replace("```json", "")
                  mr_json_str = mr_json_str.replace("```", "")

                  # Validate that the output is valid JSON.
                  try:
                    mr_json = json.loads(mr_json_str)
                    # Write the JSON object as one line.
                    outfile.write(json.dumps(mr_json) + "\n")
                  except Exception as e:
                    print(f"Error parsing JSON for MR {idx}: {e}")
                    bad_outfile.write(f"{mr_json_str}\n")
                  print(f"Processed MR {idx}/{len(rows)}")
                  print("Output:", mr_json_str)
              except Exception as e:
                  print(f"Error processing MR {idx}: {e}")
                  import pdb; pdb.set_trace()
              time.sleep(0.1)  # Adjust delay if necessary to avoid rate limits.

    print(f"All MR release notes have been written to {output_file}")

if __name__ == "__main__":
    main()