Internal change

GitOrigin-RevId: e0f6d40a13463dd73d6a0b228a7701ba9d6263b2
Jigsaw-Code · Dec 11, 2024 · ea62440 · ea62440
1 parent 2f7bc48
commit ea62440
Show file tree

Hide file tree

Showing 5 changed files with 121 additions and 53 deletions.
diff --git a/package-lock.json b/package-lock.json
diff --git a/package.json b/package.json
@@ -29,6 +29,7 @@
     "husky": "^9.1.6",
     "jest": "^29.7.0",
     "lint-staged": "^15.2.10",
+    "marked": "^15.0.3",
     "nodemon": "^3.1.4",
     "papaparse": "^5.4.1",
     "prettier": "^3.3.3",

diff --git a/runner-cli/rerunner.ts b/runner-cli/rerunner.ts
@@ -0,0 +1,51 @@
+// Rerun summarize 5x using a CSV file as input and outputting the summaries to another CSV.
+// Run like:
+// npx ts-node ./evaluations/rerunner.ts --outputFile "data1.csv" \
+// --vertexProject "<your project name here>" \
+// --inputFile "/usr/local/google/home/achvasta/Downloads/comments-with-vote-tallies.csv"
+// --rerunCount 3
+
+import { Command } from "commander";
+import { createObjectCsvWriter } from "csv-writer";
+import { getCommentsFromCsv, getSummary } from "./runner_utils";
+
+interface outputCsvFormat {
+  run: number;
+  summaryType: string;
+  text: string;
+}
+
+async function main(): Promise<void> {
+  // Parse command line arguments.
+  const program = new Command();
+  program
+    .option("-o, --outputFile <file>", "The output file name.")
+    .option("-i, --inputFile <file>", "The input file name.")
+    .option("-r, --rerunCount <count>", "The number of times to rerun.")
+    .option("-v, --vertexProject <project>", "The Vertex Project name.");
+  program.parse(process.argv);
+  const options = program.opts();
+
+  const comments = await getCommentsFromCsv(options.inputFile);
+
+  let outputTexts: outputCsvFormat[] = [];
+  const csvWriter = createObjectCsvWriter({
+    path: options.outputFile,
+    header: ["run", "summaryType", "text"],
+  });
+
+  for (let i = 0; i < options.rerunCount; i++) {
+    const summary = await getSummary(options.vertexProject, comments);
+    outputTexts = outputTexts.concat([
+      {
+        run: i,
+        summaryType: "VoteTally",
+        text: summary.getText("MARKDOWN"),
+      },
+    ]);
+  }
+
+  csvWriter.writeRecords(outputTexts).then(() => console.log("CSV file written successfully."));
+}
+
+main();
diff --git a/runner-cli/runner.ts b/runner-cli/runner.ts
@@ -0,0 +1,47 @@
+// Run the summarizer based on a CSV input and output the result as an hmtl page.
+
+import { Command } from "commander";
+import * as fs from "fs";
+import { marked } from "marked";
+import { getCommentsFromCsv, getSummary } from "./runner_utils";
+
+async function main(): Promise<void> {
+  // Parse command line arguments.
+  const program = new Command();
+  program
+    .option("-o, --outputFile <file>", "The output file name.")
+    .option("-i, --inputFile <file>", "The input file name.")
+    .option("-v, --vertexProject <project>", "The Vertex Project name.");
+  program.parse(process.argv);
+  const options = program.opts();
+
+  const comments = await getCommentsFromCsv(options.inputFile);
+
+  const summary = await getSummary(options.vertexProject, comments);
+  const markdownContent = summary.getText("MARKDOWN");
+  const htmlContent = `
+<!DOCTYPE html>
+<html>
+<head>
+    <title>Summary</title>
+    <style>
+        body {
+            font-family: Arial, sans-serif;
+            line-height: 1.6;
+            max-width: 800px;
+            margin: 0 auto;
+            padding: 20px;
+        }
+    </style>
+</head>
+<body>
+    ${marked(markdownContent)}
+</body>
+</html>`;
+
+  const outputPath = `${options.outputFile}.html`;
+  fs.writeFileSync(outputPath, htmlContent);
+  console.log(`Written summary to ${outputPath}`);
+}
+
+main();
diff --git a/evaluations/rerunner.ts → runner-cli/runner_utils.ts b/evaluations/rerunner.ts → runner-cli/runner_utils.ts
@@ -1,25 +1,10 @@
-// Rerun summarize 5x using a CSV file as input and outputting the summaries to another CSV.
-// Run like:
-// npx ts-node ./evaluations/rerunner.ts --outputFile "data1.csv" \
-// --vertexProject "<your project name here>" \
-// --inputFile "/usr/local/google/home/achvasta/Downloads/comments-with-vote-tallies.csv"
-// --rerunCount 3
-
-import { Command } from "commander";
 import { Sensemaker } from "../src/sensemaker";
 import { VertexModel } from "../src/models/vertex_model";
-import { Comment, SummarizationType, Summary, VoteTally } from "../src/types";
-import { createObjectCsvWriter } from "csv-writer";
+import { Summary, VoteTally, Comment, SummarizationType } from "../src/types";
 import * as path from "path";
 import * as fs from "fs";
 import { parse } from "csv-parse";
 
-interface outputCsvFormat {
-  run: number;
-  summaryType: string;
-  text: string;
-}
-
 // TODO: remove this and make it more general
 type VoteTallyCsvRow = {
   index: number;
@@ -40,14 +25,14 @@ type VoteTallyCsvRow = {
   "group-1-agree-count": number;
 };
 
-async function getSummary(project: string, comments: Comment[]): Promise<Summary> {
+export async function getSummary(project: string, comments: Comment[]): Promise<Summary> {
   const sensemaker = new Sensemaker({
     defaultModel: new VertexModel(project, "us-central1", "gemini-1.5-pro-002"),
   });
   return await sensemaker.summarize(comments, SummarizationType.VOTE_TALLY);
 }
 
-async function getCommentsFromCsv(inputFilePath: string): Promise<Comment[]> {
+export async function getCommentsFromCsv(inputFilePath: string): Promise<Comment[]> {
   const filePath = path.resolve(inputFilePath);
   const fileContent = fs.readFileSync(filePath, { encoding: "utf-8" });
 
@@ -85,38 +70,3 @@ async function getCommentsFromCsv(inputFilePath: string): Promise<Comment[]> {
       .on("end", () => resolve(data));
   });
 }
-
-async function main(): Promise<void> {
-  // Parse command line arguments.
-  const program = new Command();
-  program
-    .option("-o, --outputFile <file>", "The output file name.")
-    .option("-i, --inputFile <file>", "The input file name.")
-    .option("-r, --rerunCount <count>", "The number of times to rerun.")
-    .option("-v, --vertexProject <project>", "The Vertex Project name.");
-  program.parse(process.argv);
-  const options = program.opts();
-
-  const comments = await getCommentsFromCsv(options.inputFile);
-
-  let outputTexts: outputCsvFormat[] = [];
-  const csvWriter = createObjectCsvWriter({
-    path: options.outputFile,
-    header: ["run", "summaryType", "text"],
-  });
-
-  for (let i = 0; i < options.rerunCount; i++) {
-    const summary = await getSummary(options.vertexProject, comments);
-    outputTexts = outputTexts.concat([
-      {
-        run: i,
-        summaryType: "VoteTally",
-        text: summary.getText("MARKDOWN"),
-      },
-    ]);
-  }
-
-  csvWriter.writeRecords(outputTexts).then(() => console.log("CSV file written successfully."));
-}
-
-main();