utils.js: reorder mermaid inside Changes

[thypon]

No description provided.

[github-actions]

anthropic debug - [puLL-Merge] - brave/pull-merge@208

Description

This PR modifies the token limit for AI models and enhances the handling of Mermaid diagrams in the pull request explanation output. The changes aim to improve the quality and comprehensiveness of the AI-generated explanations for pull requests.

Changes

Changes

1. action.cjs

   • Increased the max_tokens value from 2048 to 3072.

2. src/anthropicExplainPatch.js

   • Increased the default max_tokens value from 2048 to 3072 in the explainPatch function.

3. src/bedrockExplainPatch.js

   • Increased the default max_tokens value from 2048 to 3072 in the explainPatch function.

4. src/openaiExplainPatch.js

   • Increased the default max_tokens value from 2048 to 3072 in the explainPatch function.

5. src/utils.js

   • Added a new regular expression mermaidRe to match Mermaid diagram code blocks.
   • Modified the explainPatchHelper function to:
     • Extract Mermaid diagrams from the response.
     • Remove diagrams from their original position in the response.
     • Add the extracted diagrams at the end of the Changes section.
   • Updated the response formatting to include the Mermaid diagrams in the appropriate location.

graph TD
    A[AI Model] -->|Generate Response| B[explainPatchHelper]
    B -->|Extract| C[Mermaid Diagrams]
    B -->|Remove| D[Diagrams from original position]
    B -->|Format| E[Response]
    C -->|Add| F[Diagrams to Changes section]
    E -->|Include| F
    E -->|Add| G[Model Attribution]

Loading

sequenceDiagram
    participant User
    participant Action
    participant ExplainPatch
    participant AIModel
    participant Utils

    User->>Action: Trigger PR review
    Action->>ExplainPatch: Call with increased max_tokens
    ExplainPatch->>AIModel: Generate explanation
    AIModel->>ExplainPatch: Return response
    ExplainPatch->>Utils: Process response
    Utils->>Utils: Extract and reposition Mermaid diagrams
    Utils->>Utils: Format response
    Utils->>ExplainPatch: Return formatted response
    ExplainPatch->>Action: Return explanation
    Action->>User: Display PR explanation

Loading

[github-actions]

openai debug - [puLL-Merge] - brave/pull-merge@208

Description

This PR increases the max_tokens parameter from 2048 to 3072 across multiple script modules, allowing for more extensive text generation by the AI models. This change is aimed at enhancing the models' capabilities in generating longer and potentially more nuanced responses. Additionally, it modifies the response formatting in src/utils.js to include Mermaid diagrams within a collapsible "Changes" section.

Changes

Changes

• action.cjs

  • Updated max_tokens value from 2048 to 3072.

• src/anthropicExplainPatch.js

  • Updated max_tokens value from 2048 to 3072.

• src/bedrockExplainPatch.js

  • Updated max_tokens value from 2048 to 3072.

• src/openaiExplainPatch.js

  • Updated max_tokens value from 2048 to 3072.

• src/utils.js

  • Added regular expression to capture Mermaid diagrams.
  • Adjusted the response formatting to insert Mermaid diagrams inside the "Changes" section.

classDiagram
  class action_cjs {
    +async function(module.exports)
  }

  class src_anthropicExplainPatch_js {
    +async function(explainPatch)
  }

  class src_bedrockExplainPatch_js {
    +async function(explainPatch)
  }

  class src_openaiExplainPatch_js {
    +async function(explainPatch)
  }

  class src_utils_js {
    +function explainPatchHelper
  }

  action_cjs <|-- src_anthropicExplainPatch_js : uses
  action_cjs <|-- src_bedrockExplainPatch_js : uses
  action_cjs <|-- src_openaiExplainPatch_js : uses
  src_utils_js <|-- src_anthropicExplainPatch_js : uses
  src_utils_js <|-- src_bedrockExplainPatch_js : uses
  src_utils_js <|-- src_openaiExplainPatch_js : uses

Loading

sequenceDiagram
    participant User
    participant action.cjs
    participant src/anthropicExplainPatch.js as Anthropic
    participant src/bedrockExplainPatch.js as Bedrock
    participant src/openaiExplainPatch.js as OpenAI
    participant src/utils.js as Utils
    User->>action.cjs: Initiate request
    action.cjs->>Anthropic: Call explainPatch
    action.cjs->>Bedrock: Call explainPatch
    action.cjs->>OpenAI: Call explainPatch
    Anthropic->>Utils: use explainPatchHelper
    Bedrock->>Utils: use explainPatchHelper
    OpenAI->>Utils: use explainPatchHelper
    Utils->>User: Return modified response with diagrams

Loading

Possible Issues

• Token Limits: Increasing the max_tokens to 3072 may hit API token limits faster depending on the service's quota.
• Performance Impact: Generating more tokens may increase response times and computational load.

Security Hotspots

None detected.

[github-actions]

bedrock debug - [puLL-Merge] - brave/pull-merge@208

Description

This PR modifies the token limit for AI models and improves the handling of Mermaid diagrams in the pull request merge explanation process. The changes aim to enhance the capability of the AI models to provide more detailed explanations and to better integrate generated diagrams into the response format.

Changes

Changes

action.cjs

• Increased the max_tokens value from 2048 to 3072.

src/anthropicExplainPatch.js

• Increased the default max_tokens value from 2048 to 3072 in the explainPatch function parameters.

src/bedrockExplainPatch.js

• Increased the default max_tokens value from 2048 to 3072 in the explainPatch function parameters.

src/openaiExplainPatch.js

• Increased the default max_tokens value from 2048 to 3072 in the explainPatch function parameters.

src/utils.js

• Added a new regular expression mermaidRe to match Mermaid diagram code blocks.
• Modified the explainPatchHelper function to:
  1. Extract Mermaid diagrams from the response.
  2. Remove the diagrams from their original positions in the response.
  3. Add the extracted diagrams at the end of the Changes section.
• Updated the logic for adding the closing </details> tag to ensure it comes after the inserted diagrams.

graph TD
    A[User] --> B[GitHub Action]
    B --> C[explainPatchHelper]
    C --> D[AI Model API]
    D --> E[Generate Response]
    E --> F[Extract Mermaid Diagrams]
    F --> G[Reformat Response]
    G --> H[Add Diagrams to Changes Section]
    H --> I[Return Formatted Response]
    I --> B
    B --> J[Update PR Comment]

Loading

sequenceDiagram
    participant User
    participant GitHub as GitHub Action
    participant Helper as explainPatchHelper
    participant AI as AI Model API
    participant Response as Response Processor

    User->>GitHub: Trigger action
    GitHub->>Helper: Call explainPatchHelper
    Helper->>AI: Send patch for explanation
    AI->>Helper: Return generated response
    Helper->>Response: Process response
    Response->>Response: Extract Mermaid diagrams
    Response->>Response: Reformat response
    Response->>Response: Add diagrams to Changes section
    Response->>Helper: Return formatted response
    Helper->>GitHub: Return final explanation
    GitHub->>User: Update PR comment

Loading