Add Language Selection for Review

action.md

@@ -28,9 +28,10 @@ jobs:
       - name: Code Review GPT
         uses: mattzcarey/[email protected]
         with:
+          GITHUB_TOKEN: ${{ github.token }}
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
           MODEL: 'gpt-4o'
-          GITHUB_TOKEN: ${{ github.token }}
+          REVIEW_LANGUAGE: 'English'
 ```
 
 ### Workflow yml option 2: Add a code review bot
@@ -60,7 +61,10 @@ jobs:
       - name: Code Review GPT
         uses: mattzcarey/[email protected]
         with:
+          GITHUB_TOKEN: ${{ github.token }}
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+          MODEL: 'gpt-4o'
+          REVIEW_LANGUAGE: 'English'
           MODEL: 'gpt-4o-mini' # reference: https://platform.openai.com/settings/organization/limits
           GITHUB_TOKEN: ${{ github.token }}
 ```

action.yml

@@ -6,6 +6,10 @@ inputs:
     description: 'The GPT model to use'
     required: true
     default: 'gpt-4o'
+  REVIEW_LANGUAGE:
+    description: 'Target natural language for translation'
+    required: false
+    default: 'English'
   OPENAI_API_KEY:
     description: 'OpenAI API Key'
     required: true
@@ -27,10 +31,11 @@ runs:
 
     - name: Run Code Review GPT
       shell: bash
-      run: npx code-review-gpt review --ci=github --model=$MODEL
+      run: npx code-review-gpt review --ci=github --model=$MODEL --reviewLanguage=$REVIEW_LANGUAGE
 
       env:
         MODEL: ${{ inputs.MODEL }}
+        REVIEW_LANGUAGE: ${{ inputs.REVIEW_LANGUAGE }}
         OPENAI_API_KEY: ${{ inputs.OPENAI_API_KEY }}
         BASE_SHA: ${{ github.event.pull_request.base.sha }}
         GITHUB_TOKEN: ${{ inputs.GITHUB_TOKEN }}

packages/code-review-gpt/README.md

@@ -105,6 +105,8 @@ You can now run `code-review-gpt review` in the root directory of any git-enable
 
 - `--reviewType` - Used with the 'review' command. The options are --reviewType=("changed" | "full" | "costOptimized). Defaults to "changed" if no option is specified. Specifies whether the review is for the full file or just the changed lines. costOptimized limits the context surrounding the changed lines to 5 lines.
 
+- `--reviewLanguage` - Specifies the target natural language for translation, such as "German", "Spanish", or "French". Use this parameter to set the language in which the utility will provide output.
+
 - `--remote` - Used with the 'review' command. Usage `--remote=mattzcarey/code-review-gpt#96`. Review a remote GitHub Pull Request.
 
 - `--commentPerFile` - Used when the `--ci` flag is set. Defaults to false. It enables the bot to comment the feedback on a file-by-file basis.

packages/code-review-gpt/src/args.ts

@@ -58,6 +58,10 @@ export const getYargs = async (): Promise<ReviewArgs> => {
       type: "string",
       default: "changed",
     })
+    .option("reviewLanguage", {
+      description: "Specifies the target natural language for translation",
+      type: "string",
+    })
     .option("remote", {
       description: "The identifier of a remote Pull Request to review",
       type: "string",

packages/code-review-gpt/src/common/types.ts

@@ -42,6 +42,7 @@ export type ReviewArgs = {
   commentPerFile: boolean;
   model: string;
   reviewType: string;
+  reviewLanguage: string | undefined;
   org: string | undefined;
   remote: string | undefined;
   provider: string;

packages/code-review-gpt/src/review/index.ts

@@ -34,6 +34,7 @@ export const review = async (
   const reviewType = yargs.reviewType;
   const organization = yargs.org;
   const provider = yargs.provider;
+  const reviewLanguage = yargs.reviewLanguage;
 
   const filteredFiles = filterFiles(files);
 
@@ -54,7 +55,8 @@ export const review = async (
   const prompts = constructPromptsArray(
     filteredFiles,
     maxPromptLength,
-    reviewType
+    reviewType,
+    reviewLanguage
   );
 
   logger.debug(`Prompts used:\n ${prompts.toString()}`);

packages/code-review-gpt/src/review/prompt/constructPrompt/constructPrompt.ts

@@ -8,7 +8,8 @@ import { instructionPrompt } from "../prompts";
 export const constructPromptsArray = (
   files: ReviewFile[],
   maxPromptLength: number,
-  reviewType: string
+  reviewType: string,
+  reviewLanguage = 'English',
 ): string[] => {
   const maxPromptPayloadLength = maxPromptLength - instructionPrompt.length;
   let promptPayloads: PromptFile[][];
@@ -33,10 +34,9 @@ export const constructPromptsArray = (
       );
   }
 
-  const languageToInstructionPrompt = instructionPrompt.replace(
-    "{Language}",
-    getLanguageName(files[0].fileName) //assume the first file is representative of the language
-  );
+  const languageToInstructionPrompt = instructionPrompt
+      .replace("{ProgrammingLanguage}", getLanguageName(files[0].fileName))
+      .replace("{ReviewLanguage}", reviewLanguage);
 
   const prompts = promptPayloads.map((payload) => {
     return languageToInstructionPrompt + JSON.stringify(payload);

packages/code-review-gpt/src/review/prompt/prompts.ts

@@ -1,4 +1,4 @@
-export const instructionPrompt = `You are an expert {Language} developer, your task is to review a set of pull requests.
+export const instructionPrompt = `You are an expert {ProgrammingLanguage} developer, your task is to review a set of pull requests.
 You are given a list of filenames and their partial contents, but note that you might not have the full context of the code.
 
 Only review lines of code which have been changed (added or removed) in the pull request. The code looks similar to the output of a git diff command. Lines which have been removed are prefixed with a minus (-) and lines which have been added are prefixed with a plus (+). Other lines are added to provide context but should be ignored in the review.
@@ -11,7 +11,7 @@ Do not comment on breaking functions down into smaller, more manageable function
 
 Use markdown formatting for the feedback details. Also do not include the filename or risk level in the feedback details.
 
-Ensure the feedback details are brief, concise, accurate. If there are multiple similar issues, only comment on the most critical.
+Ensure the feedback details are brief, concise, accurate, and in {ReviewLanguage}. If there are multiple similar issues, only comment on the most critical.
 
 Include brief example code snippets in the feedback details for your suggested changes when you're confident your suggestions are improvements. Use the same programming language as the file under review.
 If there are multiple improvements you suggest in the feedback details, use an ordered list to indicate the priority of the changes.
@@ -38,36 +38,3 @@ You are a senior developer and have just reviewed a pull request. This was your
 {feedback}
 Please summarise the review using 3 emojis.
 `;
-
-export const demoPrompt = `You are an senior developer, your task is to review a code snippet.
-Note that you do not have the full context of the code.
-
-Begin your review by evaluating the code using a risk score similar to a LOGAF score but measured from 1 to 5, where 1 is the lowest risk to the code base if the code is merged and 5 is the highest risk which would likely break something or be unsafe.
-
-In your feedback, focus on highlighting potential bugs, improving readability if it is a problem, making code cleaner, and maximising the performance of the programming language. Flag any API keys or secrets present in the code in plain text immediately as highest risk. Rate the changes based on SOLID principles if applicable.
-
-Do not comment on breaking functions down into smaller, more manageable functions unless it is a huge problem. Also be aware that there will be libraries and techniques used which you are not familiar with, so do not comment on those unless you are confident that there is a problem.
-
-Use markdown formatting for the feedback details. Also do not include the risk level in the feedback details.
-
-Ensure the feedback details are brief, concise, accurate. If there are multiple similar issues, only comment on the most critical.
-
-Include brief example code snippets in the feedback details for your suggested changes when you're confident your suggestions are improvements. Use the same programming language as the file under review.
-If there are multiple improvements you suggest in the feedback details, use an ordered list to indicate the priority of the changes.
-
-Format the response in a valid JSON format as a list of feedbacks, where the value is an object containing the risk score ("riskScore") and the feedback ("details"). Also add the filename ("filename") which will always be "demo code". The schema of the JSON feedback object must be:
-{
-  "fileName": {
-    "type": "string"
-  },
-  "riskScore": {
-    "type": "number"
-  },
-  "details": {
-    "type": "string"
-  }
-}
-
-The code to review is provided below:
-
-`;

packages/code-review-gpt/src/test/index.ts

@@ -12,7 +12,7 @@ import { loadOrGenerateCodeSnippets } from "./load/loadTestCodeSnippets";
 import { runTests } from "./run/runTest";
 
 export const test = async (
-  { ci, model, reviewType }: ReviewArgs,
+  { ci, model, reviewType, reviewLanguage }: ReviewArgs,
   openAIApiKey: string
 ): Promise<void> => {
   const maxPromptLength = getMaxPromptLength(model);
@@ -40,12 +40,13 @@ export const test = async (
 
   // Run the review on the code snippets and compare the results to the expected results.
   const testSummary = await runTests(
+    openAIApiKey,
     testCasesWithSnippets,
     model,
     maxPromptLength,
     vectorStore,
     reviewType,
-    openAIApiKey
+    reviewLanguage,
   );
 
   if (ci === PlatformOptions.GITHUB) {

packages/code-review-gpt/src/test/run/runTest.ts

@@ -13,20 +13,24 @@ import {
 
 /**
  * Run a single test case.
+ * @param openAIApiKey Open AI API Key.
  * @param testCase The test case.
  * @param modelName The name of the model.
  * @param maxPromptLength The maximum prompt length.
  * @param vectorStore The vector store.
  * @param reviewType The review type.
+ * @param reviewLanguage Target natural language for translation.
  * @returns The test result.
  */
 const runTest = async (
+  openAIApiKey: string,
   testCase: TestCase,
   modelName: string,
   maxPromptLength: number,
   vectorStore: MemoryVectorStore,
   reviewType: string,
-  openAIApiKey: string
+  reviewLanguage?: string,
+  // eslint-disable-next-line max-params
 ): Promise<testResult> => {
   if (!testCase.snippet) {
     throw new Error(`Test case ${testCase.name} does not have a snippet.`);
@@ -38,7 +42,8 @@ const runTest = async (
   const prompts = constructPromptsArray(
     [testCase.snippet],
     maxPromptLength,
-    reviewType
+    reviewType,
+    reviewLanguage
   );
 
   const { markdownReport: reviewResponse } = await askAI(
@@ -74,20 +79,24 @@ const runTest = async (
 
 /**
  * Run all the test cases.
+ * @param openAIApiKey Open AI API Key.
  * @param testCases The test cases.
  * @param modelName The name of the model.
  * @param maxPromptLength The maximum prompt length.
  * @param vectorStore The vector store.
  * @param reviewType The review type.
+ * @param reviewLanguage Target natural language for translation.
  * @returns The test results.
  */
 export const runTests = async (
+  openAIApiKey: string,
   testCases: TestCase[],
   modelName: string,
   maxPromptLength: number,
   vectorStore: MemoryVectorStore,
   reviewType: string,
-  openAIApiKey: string
+  reviewLanguage?: string,
+  // eslint-disable-next-line max-params
 ): Promise<string> => {
   if (testCases.length === 0) {
     return "No test cases found.";
@@ -101,12 +110,13 @@ export const runTests = async (
   for (const testCase of testCases) {
     try {
       const result = await runTest(
+        openAIApiKey,
         testCase,
         modelName,
         maxPromptLength,
         vectorStore,
         reviewType,
-        openAIApiKey
+        reviewLanguage
       );
       testResults[testCase.name] = result;
     } catch (error) {