[FEATURE] Handle OpenAI downtime and large requests

[guibranco]

This feature aims to enhance the resilience and user experience of dotnet-aicommitmessage by addressing two scenarios:

1. OpenAI Downtime/Unresponsiveness:

   • Detect if the OpenAI API is down or unresponsive.
   • Abort the process gracefully, displaying a clear error message to the user.

2. Large Requests (Exceeding OpenAI Limits):

   • If the Git diff output size exceeds acceptable limits:
     • Abort the commit process.
     • Provide an error message instructing the user to reduce the number of staged files or the size of the changes.

Acceptance Criteria

• OpenAI Downtime Handling:

  • The tool should detect when the OpenAI API is unreachable or unresponsive (e.g., network issues, server downtime).
  • An appropriate error message should be returned, clearly informing the user of the issue.

• Handling Large Requests:

  • Implement a mechanism to calculate and validate the size of the Git diff output.
  • If the size exceeds a predefined limit, return an error message and abort the commit process.
  • The error message should suggest staging fewer files or smaller changes.

Proposed Error Messages

• OpenAI Downtime:

  ⚠️ OpenAI API is currently unavailable. Please try again later.

• Large Request:

  🚫 The staged changes are too large to process. Please reduce the number of files or size of changes and try again.

Additional Notes

• Ensure robust error handling to prevent crashes during these scenarios.
• Add relevant unit tests to verify both downtime handling and request size validation.
• Document these behaviors in the README.md or a dedicated error-handling section.

Sample example code

public string GenerateCommitMessage(GenerateCommitMessageOptions options)
{
    const int MaxDiffSizeBytes = 10240; // Limit Git diff size to 10 KB
    var model = EnvironmentLoader.LoadOpenAiModel();
    var url = EnvironmentLoader.LoadOpenAiApiUrl();
    var key = EnvironmentLoader.LoadOpenAiApiKey();

    var client = new ChatClient(
        model,
        new ApiKeyCredential(key),
        new OpenAIClientOptions { Endpoint = new Uri(url) }
    );

    var branch = string.IsNullOrEmpty(options.Branch)
        ? GitHelper.GetBranchName()
        : options.Branch;

    var diff = string.IsNullOrEmpty(options.Diff) ? GitHelper.GetGitDiff() : options.Diff;

    // Check if the Git diff is too large
    if (Encoding.UTF8.GetByteCount(diff) > MaxDiffSizeBytes)
    {
        throw new InvalidOperationException(
            "🚫 The staged changes are too large to process. Please reduce the number of files or size of changes and try again."
        );
    }

    var message = options.Message; // Use the provided message from the prepare-commit-msg hook

    if (string.IsNullOrEmpty(branch) && string.IsNullOrEmpty(diff))
    {
        throw new InvalidOperationException(
            "Unable to generate commit message: Both branch and diff are empty."
        );
    }

    var formattedMessage =
        "Branch: "
        + (string.IsNullOrEmpty(branch) ? "<unknown>" : branch)
        + "\n\n"
        + "Original message: "
        + message
        + "\n\n"
        + "Git Diff: "
        + (string.IsNullOrEmpty(diff) ? "<no changes>" : diff);

    // Call OpenAI API and handle downtime
    try
    {
        var chatCompletion = client.CompleteChat(
            new SystemChatMessage(Constants.SystemMessage),
            new UserChatMessage(formattedMessage)
        );

        var text = chatCompletion.Value.Content[0].Text;

        // Post-processing the generated message
        if (text.Length >= 7 && text[..7] == "type - ")
        {
            text = text[7..];
        }

        var provider = GetGitProvider();
        if (provider == GitProvider.GitHub)
        {
            var issueNumber = BranchNameUtility.ExtractIssueNumber(branch);
            if (!string.IsNullOrWhiteSpace(issueNumber))
            {
                text = $"#{issueNumber} {text}";
            }
        }
        else
        {
            var jiraTicketNumber = BranchNameUtility.ExtractJiraTicket(branch);
            if (!string.IsNullOrWhiteSpace(jiraTicketNumber))
            {
                text = $"[{jiraTicketNumber}] {text}";
            }
        }

        var gitVersionCommand = GitVersionUtility.ExtractGitVersionBumpCommand(message);
        if (!string.IsNullOrEmpty(gitVersionCommand))
        {
            text = $"{text} {gitVersionCommand}";
        }

        return text;
    }
    catch (Exception ex)
    {
        throw new InvalidOperationException(
            $"⚠️ OpenAI API is currently unavailable or an error occurred. Please try again later.\nError: {ex.Message}"
        );
    }
}

[gitauto-ai]

Click the checkbox below to generate a PR!

• Generate PR

@guibranco, You have 16 requests left in this cycle which refreshes on 2024-12-26 00:19:30+00:00.
If you have any questions or concerns, please contact us at [email protected].

[gitauto-ai]

@guibranco Pull request completed! Check it out here #151 🚀

Note: I automatically create a pull request for an unassigned and open issue in order from oldest to newest once a day at 00:00 UTC, as long as you have remaining automation usage. Should you have any questions or wish to change settings or limits, please feel free to contact [email protected] or invite us to Slack Connect.