UtilsText.mjs

import OpenAI from "openai";
import { logError, logInfo } from "#lib/UtilsLog";

/**
 * Utilities functions for text
 * @module Text
 */

/**
 * Generates a summary of the provided content using OpenAI's chat completions API.
 *
 * @param {Object} options - The options for generating the summary.
 * @param {string} options.apiKey - The API key for OpenAI.
 * @param {string} options.baseUrl - The base URL for the OpenAI API.
 * @param {string} options.model - The model to use for generating the summary.
 * @param {string} options.language - The language in which the summary should be written.
 * @param {number} options.maxWords - The maximum number of words the summary should contain.
 * @param {string} options.content - The content to be summarized.
 * @returns {Promise<string>} - A promise that resolves to the generated summary.
 * @throws {Error} - Throws an error if there is a problem generating the summary.
 *
 * @example
 * import { generateSummary } from "nsuite";
 * const summary = generateSummary({
 *   apiKey: "",
 *   baseUrl: "https://dashscope.aliyuncs.com/compatible-mode/v1",
 *   model: "qwen-turbo",
 *   language: "English",
 *   maxWords: 200,
 *   content: "",
 * });
 */
export async function generateSummary({
  apiKey,
  baseUrl,
  model,
  language,
  maxWords,
  content,
}) {
  try {
    const openai = new OpenAI({
      apiKey,
      baseURL: baseUrl,
    });
    const completion = await openai.chat.completions.create({
      messages: [
        {
          role: "system",
          content: `You are a professional writer, you can write a summary of the article. The summary should be in ${language} and should be less than ${maxWords} words (each English letter, number, and symbol is considered as one word).`,
        },
        { role: "user", content },
      ],
      model,
      /**
       * 采样温度，控制模型生成文本的多样性。
       * temperature越高，生成的文本更多样，反之，生成的文本更确定。
       * 取值范围： [0, 2)
       *
       * @ignore
       */
      temperature: 0.7,
      max_tokens: 200,
      /**
       * 控制模型生成文本时的内容重复度。
       * 取值范围：[-2.0, 2.0]。正数会减少重复度，负数会增加重复度。
       * 适用场景：
       * 较高的presence_penalty适用于要求多样性、趣味性或创造性的场景，如创意写作或头脑风暴。
       * 较低的presence_penalty适用于要求一致性或专业术语的场景，如技术文档或其他正式文档。
       *
       * @ignore
       */
      presence_penalty: -2.0,
    });
    logInfo(JSON.stringify(completion));
    return completion.choices[0].message.content || "";
  } catch (error) {
    logError("Error generating summary:", error);
    throw error;
  }
}