diff --git a/src/ad.js b/src/ad.js new file mode 100644 index 0000000..44b9270 --- /dev/null +++ b/src/ad.js @@ -0,0 +1,19 @@ +export function insertAd() { + return ` +--- + +## **AI产品自荐: [AIClient2API ↗️](https://github.com/justlovemaki/AIClient-2-API)** + +厌倦了在各种AI模型间来回切换，被烦人的API额度限制束缚手脚？现在，你有了一个终极解决方案！🎉 'AIClient-2-API' 不仅仅是一个普通的API代理，它是一个能将 Gemini CLI 和 Kiro 客户端等工具“点石成金”，变为强大 OpenAI 兼容 API 的魔法盒子。 + +这个项目的核心魅力在于它的“逆向思维”和强大功能： + +✨ **客户端变API，解锁新姿势**：我们巧妙地利用 Gemini CLI 的 OAuth 登录，让你轻松**突破官方免费API的速率和额度限制**。更令人兴奋的是，通过封装 Kiro 客户端的接口，我们成功**破解其API，让你能免费丝滑地调用强大的 Claude 模型**！这为你提供了 **“使用免费Claude API加 Claude Code，开发编程的经济实用方案”**。 + +🔧 **系统提示词，由你掌控**：想让AI更听话？我们提供了强大的系统提示词（System Prompt）管理功能。你可以轻松**提取、替换（'overwrite'）或追加（'append'）**任何请求中的系统提示词，在服务端精细地调整AI的行为，而无需修改客户端代码。 + +💡 **顶级体验，平民成本**：想象一下，**在你的编辑器里用 Kilo 代码助手，加上 Cursor 的高效提示词，再配上任意顶级大模型——用 Cursor，又何必是 Cursor？** 本项目让你能以极低的成本，组合出媲美付费工具的开发体验。同时支持MCP协议和图片、文档等多模态输入，让你的创意不再受限。 + +告别繁琐配置和昂贵账单，拥抱这个集免费、强大、灵活于一身的AI开发新范式吧！ + `; +} \ No newline at end of file diff --git a/src/chatapi.js b/src/chatapi.js index 0855a3e..64b8f1d 100644 --- a/src/chatapi.js +++ b/src/chatapi.js @@ -32,7 +32,7 @@ async function callGeminiChatAPI(env, promptText, systemPromptText = null) { } try { - const response = await fetchWithTimeout(url, { + const response = await fetch(url, { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(payload) @@ -126,6 +126,10 @@ async function* callGeminiChatAPIStream(env, promptText, systemPromptText = null contents: [{ parts: [{ text: promptText }] }], + generationConfig: { + temperature: 1, + topP: 0.95 + } }; if (systemPromptText && typeof systemPromptText === 'string' && systemPromptText.trim() !== '') { @@ -137,7 +141,7 @@ async function* callGeminiChatAPIStream(env, promptText, systemPromptText = null let response; try { - response = await fetchWithTimeout(url, { + response = await fetch(url, { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(payload) @@ -334,7 +338,7 @@ async function callOpenAIChatAPI(env, promptText, systemPromptText = null) { }; try { - const response = await fetchWithTimeout(url, { + const response = await fetch(url, { method: 'POST', headers: { 'Content-Type': 'application/json', @@ -413,7 +417,7 @@ async function* callOpenAIChatAPIStream(env, promptText, systemPromptText = null let response; try { - response = await fetchWithTimeout(url, { + response = await fetch(url, { method: 'POST', headers: { 'Content-Type': 'application/json', @@ -574,7 +578,7 @@ export async function* callChatAPIStream(env, promptText, systemPromptText = nul * @param {number} timeout 超时时间，单位毫秒 * @returns {Promise} */ -async function fetchWithTimeout(resource, options = {}, timeout = 60000) { +async function fetchWithTimeout(resource, options = {}, timeout = 180000) { const controller = new AbortController(); const id = setTimeout(() => controller.abort(), timeout); diff --git a/src/dataFetchers.js b/src/dataFetchers.js index 8e65ae8..abbeb46 100644 --- a/src/dataFetchers.js +++ b/src/dataFetchers.js @@ -1,19 +1,16 @@ // src/dataFetchers.js -import AibaseDataSource from './dataSources/aibase.js'; +import NewsAggregatorDataSource from './dataSources/newsAggregator.js'; import GithubTrendingDataSource from './dataSources/github-trending.js'; -import HuggingfacePapersDataSource from './dataSources/huggingface-papers.js'; -import XinZhiYuanDataSource from './dataSources/xinzhiyuan.js'; -import QBitDataSource from './dataSources/qbit.js'; -import JiqizhixinDataSource from './dataSources/jiqizhixin.js'; -import XiaohuDataSource from './dataSources/xiaohu.js'; +import PapersDataSource from './dataSources/papers.js'; import TwitterDataSource from './dataSources/twitter.js'; import RedditDataSource from './dataSources/reddit.js'; + // Register data sources as arrays to support multiple sources per type export const dataSources = { - news: { name: '新闻', sources: [AibaseDataSource, XiaohuDataSource, QBitDataSource, XinZhiYuanDataSource] }, + news: { name: '新闻', sources: [NewsAggregatorDataSource] }, project: { name: '项目', sources: [GithubTrendingDataSource] }, - paper: { name: '论文', sources: [HuggingfacePapersDataSource, JiqizhixinDataSource] }, + paper: { name: '论文', sources: [PapersDataSource] }, socialMedia: { name: '社交平台', sources: [TwitterDataSource, RedditDataSource] }, // Add new data sources here as arrays, e.g., // newType: { name: '新类型', sources: [NewTypeDataSource1, NewTypeDataSource2] }, diff --git a/src/dataSources/newsAggregator.js b/src/dataSources/newsAggregator.js new file mode 100644 index 0000000..b8c4dc3 --- /dev/null +++ b/src/dataSources/newsAggregator.js @@ -0,0 +1,135 @@ +import { getRandomUserAgent, sleep, isDateWithinLastDays, stripHtml, formatDateToChineseWithTime, escapeHtml } from '../helpers'; + +const NewsAggregatorDataSource = { + type: 'news-aggregator', + async fetch(env, foloCookie) { + const listId = env.NEWS_AGGREGATOR_LIST_ID; + const fetchPages = parseInt(env.NEWS_AGGREGATOR_FETCH_PAGES || '1', 10); + const allNewsItems = []; + const filterDays = parseInt(env.FOLO_FILTER_DAYS || '3', 10); + + if (!listId) { + console.warn('NEWS_AGGREGATOR_LIST_ID is not set in environment variables. Skipping news aggregator fetch.'); + return { + version: "https://jsonfeed.org/version/1.1", + title: "Aggregated News", + home_page_url: "https://example.com/news", + description: "Aggregated news from various sources", + language: "zh-cn", + items: [] + }; + } + + let publishedAfter = null; + for (let i = 0; i < fetchPages; i++) { + const userAgent = getRandomUserAgent(); + const headers = { + 'User-Agent': userAgent, + 'Content-Type': 'application/json', + 'accept': 'application/json', + 'accept-language': 'zh-CN,zh;q=0.9', + 'baggage': 'sentry-environment=stable,sentry-release=5251fa921ef6cbb6df0ac4271c41c2b4a0ce7c50,sentry-public_key=e5bccf7428aa4e881ed5cb713fdff181,sentry-trace_id=2da50ca5ad944cb794670097d876ada8,sentry-sampled=true,sentry-sample_rand=0.06211835167903246,sentry-sample_rate=1', + 'origin': 'https://app.follow.is', + 'priority': 'u=1, i', + 'sec-ch-ua': '"Google Chrome";v="135", "Not-A.Brand";v="8", "Chromium";v="135"', + 'sec-ch-ua-mobile': '?1', + 'sec-ch-ua-platform': '"Android"', + 'sec-fetch-dest': 'empty', + 'sec-fetch-mode': 'cors', + 'sec-fetch-site': 'same-site', + 'x-app-name': 'Folo Web', + 'x-app-version': '0.4.9', + }; + + if (foloCookie) { + headers['Cookie'] = foloCookie; + } + + const body = { + listId: listId, + view: 1, + withContent: true, + }; + + if (publishedAfter) { + body.publishedAfter = publishedAfter; + } + + try { + console.log(`Fetching News Aggregator data, page ${i + 1}...`); + const response = await fetch(env.FOLO_DATA_API, { + method: 'POST', + headers: headers, + body: JSON.stringify(body), + }); + + if (!response.ok) { + console.error(`Failed to fetch News Aggregator data, page ${i + 1}: ${response.statusText}`); + break; + } + const data = await response.json(); + if (data && data.data && data.data.length > 0) { + const filteredItems = data.data.filter(entry => isDateWithinLastDays(entry.entries.publishedAt, filterDays)); + allNewsItems.push(...filteredItems.map(entry => ({ + id: entry.entries.id, + url: entry.entries.url, + title: entry.entries.title, + content_html: entry.entries.content, + date_published: entry.entries.publishedAt, + authors: [{ name: entry.entries.author }], + source: entry.entries.author ? `${entry.feeds.title} - ${entry.entries.author}` : entry.feeds.title, + }))); + publishedAfter = data.data[data.data.length - 1].entries.publishedAt; + } else { + console.log(`No more data for News Aggregator, page ${i + 1}.`); + break; + } + } catch (error) { + console.error(`Error fetching News Aggregator data, page ${i + 1}:`, error); + break; + } + + await sleep(Math.random() * 5000); + } + + return { + version: "https://jsonfeed.org/version/1.1", + title: "Aggregated News", + home_page_url: "https://example.com/news", + description: "Aggregated news from various sources", + language: "zh-cn", + items: allNewsItems + }; + }, + + transform(rawData, sourceType) { + if (!rawData || !rawData.items) { + return []; + } + + return rawData.items.map(item => ({ + id: item.id, + type: sourceType, + url: item.url, + title: item.title, + description: stripHtml(item.content_html || ""), + published_date: item.date_published, + authors: item.authors ? item.authors.map(author => author.name).join(', ') : 'Unknown', + source: item.source || 'Aggregated News', + details: { + content_html: item.content_html || "" + } + })); + }, + + generateHtml: (item) => { + return ` + ${escapeHtml(item.title)}
+ 来源: ${escapeHtml(item.source || '未知')} | 发布日期: ${formatDateToChineseWithTime(item.published_date)} +
${item.details.content_html || '无内容。'}
+ 阅读更多 + `; + } +}; + +export default NewsAggregatorDataSource; diff --git a/src/dataSources/papers.js b/src/dataSources/papers.js new file mode 100644 index 0000000..3c8572b --- /dev/null +++ b/src/dataSources/papers.js @@ -0,0 +1,137 @@ +import { getRandomUserAgent, sleep, isDateWithinLastDays, stripHtml, formatDateToChineseWithTime, escapeHtml } from '../helpers'; + +const PapersDataSource = { + type: 'papers', + async fetch(env, foloCookie) { + const hgPapersListId = env.HGPAPERS_LIST_ID; + const fetchPages = parseInt(env.HGPAPERS_FETCH_PAGES || '1', 10); + const allPaperItems = []; + const filterDays = parseInt(env.FOLO_FILTER_DAYS || '3', 10); + + if (!hgPapersListId) { + console.warn('HGPAPERS_LIST_ID is not set in environment variables. Skipping papers fetch.'); + return { + version: "https://jsonfeed.org/version/1.1", + title: "Aggregated Papers", + home_page_url: "https://example.com/papers", + description: "Aggregated papers from various sources", + language: "zh-cn", + items: [] + }; + } + + let publishedAfter = null; + for (let i = 0; i < fetchPages; i++) { + const userAgent = getRandomUserAgent(); + const headers = { + 'User-Agent': userAgent, + 'Content-Type': 'application/json', + 'accept': 'application/json', + 'accept-language': 'zh-CN,zh;q=0.9', + 'baggage': 'sentry-environment=stable,sentry-release=5251fa921ef6cbb6df0ac4271c41c2b4a0ce7c50,sentry-public_key=e5bccf7428aa4e881ed5cb713fdff181,sentry-trace_id=2da50ca5ad944cb794670097d876ada8,sentry-sampled=true,sentry-sample_rand=0.06211835167903246,sentry-sample_rate=1', + 'origin': 'https://app.follow.is', + 'priority': 'u=1, i', + 'sec-ch-ua': '"Google Chrome";v="135", "Not-A.Brand";v="8", "Chromium";v="135"', + 'sec-ch-ua-mobile': '?1', + 'sec-ch-ua-platform': '"Android"', + 'sec-fetch-dest': 'empty', + 'sec-fetch-mode': 'cors', + 'sec-fetch-site': 'same-site', + 'x-app-name': 'Folo Web', + 'x-app-version': '0.4.9', + }; + + if (foloCookie) { + headers['Cookie'] = foloCookie; + } + + const body = { + listId: hgPapersListId, + view: 1, + withContent: true, + }; + + if (publishedAfter) { + body.publishedAfter = publishedAfter; + } + + try { + console.log(`Fetching Papers data, page ${i + 1}...`); + const response = await fetch(env.FOLO_DATA_API, { + method: 'POST', + headers: headers, + body: JSON.stringify(body), + }); + + if (!response.ok) { + console.error(`Failed to fetch Papers data, page ${i + 1}: ${response.statusText}`); + break; + } + const data = await response.json(); + if (data && data.data && data.data.length > 0) { + const filteredItems = data.data.filter(entry => isDateWithinLastDays(entry.entries.publishedAt, filterDays)); + allPaperItems.push(...filteredItems.map(entry => ({ + id: entry.entries.id, + url: entry.entries.url, + title: entry.entries.title, + content_html: entry.entries.content, + date_published: entry.entries.publishedAt, + authors: [{ name: entry.entries.author }], + source: entry.feeds.title, + }))); + publishedAfter = data.data[data.data.length - 1].entries.publishedAt; + } else { + console.log(`No more data for Papers, page ${i + 1}.`); + break; + } + } catch (error) { + console.error(`Error fetching Papers data, page ${i + 1}:`, error); + break; + } + + await sleep(Math.random() * 5000); + } + + return { + version: "https://jsonfeed.org/version/1.1", + title: "Aggregated Papers", + home_page_url: "https://example.com/papers", + description: "Aggregated papers from various sources", + language: "zh-cn", + items: allPaperItems + }; + }, + + transform(rawData, sourceType) { + if (!rawData || !rawData.items) { + return []; + } + + return rawData.items.map(item => ({ + id: item.id, + type: sourceType, + url: item.url, + title: item.title, + description: stripHtml(item.content_html || ""), + published_date: item.date_published, + authors: item.authors ? item.authors.map(author => author.name).join(', ') : 'Unknown', + source: item.source || 'Aggregated Papers', + details: { + content_html: item.content_html || "" + } + })); + }, + + generateHtml: (item) => { + return ` + ${escapeHtml(item.title)}
+ 来源: ${escapeHtml(item.source || '未知')} | 发布日期: ${formatDateToChineseWithTime(item.published_date)} +
+ ${item.details.content_html || '无内容。'}
+
+ 在 ArXiv/来源 阅读 + `; + } +}; + +export default PapersDataSource; \ No newline at end of file diff --git a/src/dataSources/reddit.js b/src/dataSources/reddit.js index c229802..febb5e8 100644 --- a/src/dataSources/reddit.js +++ b/src/dataSources/reddit.js @@ -1,6 +1,4 @@ import { getRandomUserAgent, sleep, isDateWithinLastDays, stripHtml, formatDateToChineseWithTime, escapeHtml} from '../helpers'; -import { callChatAPI } from '../chatapi.js'; -import { removeMarkdownCodeBlock } from '../helpers.js'; const RedditDataSource = { async fetch(env, foloCookie) { @@ -107,61 +105,11 @@ const RedditDataSource = { return redditData; } - if (!env.OPEN_TRANSLATE === "true") { - console.warn("Skipping reddit translations."); - redditData.items = redditData.items.map(item => ({ - ...item, - title_zh: item.title || "" - })); - return redditData; - } - - const itemsToTranslate = redditData.items.map((item, index) => ({ - id: index, - original_title: item.title || "" + redditData.items = redditData.items.map(item => ({ + ...item, + title_zh: item.title || "" })); - const hasContentToTranslate = itemsToTranslate.some(item => item.original_title.trim() !== ""); - if (!hasContentToTranslate) { - console.log("No non-empty reddit titles to translate for today's posts."); - redditData.items = redditData.items.map(item => ({ ...item, title_zh: item.title || "" })); - return redditData; - } - - const promptText = `You will be given a JSON array of reddit data objects. Each object has an "id" and "original_title". -Translate "original_title" into Chinese. -Return a JSON array of objects. Each output object MUST have: -- "id": The same id from the input. -- "title_zh": Chinese translation of "original_title". Empty if original is empty. -Input: ${JSON.stringify(itemsToTranslate)} -Respond ONLY with the JSON array.`; - - let translatedItemsMap = new Map(); - try { - console.log(`Requesting translation for ${itemsToTranslate.length} reddit titles for today.`); - const chatResponse = await callChatAPI(env, promptText); - const parsedTranslations = JSON.parse(removeMarkdownCodeBlock(chatResponse)); - - if (parsedTranslations) { - parsedTranslations.forEach(translatedItem => { - if (translatedItem && typeof translatedItem.id === 'number' && - typeof translatedItem.title_zh === 'string') { - translatedItemsMap.set(translatedItem.id, translatedItem); - } - }); - } - } catch (translationError) { - console.error("Failed to translate reddit titles in batch:", translationError.message); - } - - redditData.items = redditData.items.map((originalItem, index) => { - const translatedData = translatedItemsMap.get(index); - return { - ...originalItem, - title_zh: translatedData ? translatedData.title_zh : (originalItem.title || "") - }; - }); - return redditData; }, @@ -174,7 +122,7 @@ Respond ONLY with the JSON array.`; id: item.id, type: sourceType, url: item.url, - title: item.title_zh || item.title, // Use translated title if available + title: item.title, description: stripHtml(item.content_html || ""), published_date: item.date_published, authors: item.authors ? item.authors.map(author => author.name).join(', ') : 'Unknown', diff --git a/src/foot.js b/src/foot.js index 3ee01c3..fd06fc6 100644 --- a/src/foot.js +++ b/src/foot.js @@ -3,7 +3,7 @@ export function insertFoot() { --- -## **收听语音版AI日报** +## **AI资讯日报语音版** | 🎙️ **小宇宙** | 📹 **抖音** | | --- | --- | diff --git a/src/handlers/commitToGitHub.js b/src/handlers/commitToGitHub.js index 03c23c5..9e31031 100644 --- a/src/handlers/commitToGitHub.js +++ b/src/handlers/commitToGitHub.js @@ -1,5 +1,5 @@ // src/handlers/commitToGitHub.js -import { getISODate, formatMarkdownText, replaceImageProxy,formatDateToGMT12WithTime } from '../helpers.js'; +import { getISODate, formatMarkdownText } from '../helpers.js'; import { getGitHubFileSha, createOrUpdateGitHubFile } from '../github.js'; import { storeInKV } from '../kv.js'; import { marked } from '../marked.esm.js'; @@ -13,23 +13,12 @@ export async function handleCommitToGitHub(request, env) { const dateStr = formData.get('date') || getISODate(); const dailyMd = formData.get('daily_summary_markdown'); const podcastMd = formData.get('podcast_script_markdown'); - // 从 "YYYY-MM-DD" 格式的 dateStr 中提取 "YYYY-MM" - const yearMonth = dateStr.substring(0, 7); - const report = { - report_date: dateStr, - title: dateStr+'日刊', - link: '/'+yearMonth+'/'+dateStr+'/', - content_html: null, - // 可以添加其他相關欄位，例如作者、來源等 - published_date: formatDateToGMT12WithTime(new Date()) // 記錄保存時間 - } + const filesToCommit = []; if (dailyMd) { filesToCommit.push({ path: `daily/${dateStr}.md`, content: formatMarkdownText(dailyMd), description: "Daily Summary File" }); - report.content_html = marked.parse(formatMarkdownText(env.IMG_PROXY, dailyMd)); - storeInKV(env.DATA_KV, `${dateStr}-report`, report); } if (podcastMd) { filesToCommit.push({ path: `podcast/${dateStr}.md`, content: podcastMd, description: "Podcast Script File" }); diff --git a/src/handlers/genAIContent.js b/src/handlers/genAIContent.js index 1ba9646..68b587a 100644 --- a/src/handlers/genAIContent.js +++ b/src/handlers/genAIContent.js @@ -4,12 +4,14 @@ import { getFromKV } from '../kv.js'; import { callChatAPIStream } from '../chatapi.js'; import { generateGenAiPageHtml } from '../htmlGenerators.js'; import { dataSources } from '../dataFetchers.js'; // Import dataSources -import { getSystemPromptSummarizationStepOne } from "../prompt/summarizationPromptStepOne"; +import { getSystemPromptSummarizationStepOne } from "../prompt/summarizationPromptStepZero"; import { getSystemPromptSummarizationStepTwo } from "../prompt/summarizationPromptStepTwo"; import { getSystemPromptSummarizationStepThree } from "../prompt/summarizationPromptStepThree"; import { getSystemPromptPodcastFormatting, getSystemPromptShortPodcastFormatting } from '../prompt/podcastFormattingPrompt.js'; import { getSystemPromptDailyAnalysis } from '../prompt/dailyAnalysisPrompt.js'; // Import new prompt import { insertFoot } from '../foot.js'; +import { insertAd } from '../ad.js'; +import { getDailyReportContent } from '../github.js'; // 导入 getDailyReportContent export async function handleGenAIPodcastScript(request, env) { let dateStr; @@ -26,7 +28,25 @@ export async function handleGenAIPodcastScript(request, env) { formData = await request.formData(); dateStr = formData.get('date'); selectedItemsParams = formData.getAll('selectedItems'); - outputOfCall1 = formData.get('summarizedContent'); // Get summarized content from form data + const readGithub = formData.get('readGithub') === 'true'; + + if (readGithub) { + const filePath = `daily/${dateStr}.md`; + console.log(`从 GitHub 读取文件: ${filePath}`); + try { + outputOfCall1 = await getDailyReportContent(env, filePath); + if (!outputOfCall1) { + throw new Error(`从 GitHub 读取文件 ${filePath} 失败或内容为空。`); + } + console.log(`成功从 GitHub 读取文件，内容长度: ${outputOfCall1.length}`); + } catch (error) { + console.error(`读取 GitHub 文件出错: ${error}`); + const errorHtml = generateGenAiPageHtml(env, '生成AI播客脚本出错', `

从 GitHub 读取文件失败: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, null, null, null, null, null, null, outputOfCall1, null); + return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); + } + } else { + outputOfCall1 = formData.get('summarizedContent'); // Get summarized content from form data + } if (!outputOfCall1) { const errorHtml = generateGenAiPageHtml(env, '生成AI播客脚本出错', '

Summarized content is missing. Please go back and generate AI content first.

', dateStr, true, null, null, null, null, null, null, outputOfCall1, null); @@ -156,7 +176,7 @@ export async function handleGenAIContent(request, env) { // Add new data sources switch (item.type) { case 'news': - itemText = `News Title: ${item.title}\nPublished: ${item.published_date}\nContent Summary: ${stripHtml(item.details.content_html)}`; + itemText = `News Title: ${item.title}\nPublished: ${item.published_date}\nUrl: ${item.url}\nContent Summary: ${stripHtml(item.details.content_html)}`; break; case 'project': itemText = `Project Name: ${item.title}\nPublished: ${item.published_date}\nUrl: ${item.url}\nDescription: ${item.description}\nStars: ${item.details.totalStars}`; @@ -191,43 +211,43 @@ export async function handleGenAIContent(request, env) { } //提示词内不能有英文引号，否则会存储数据缺失。 - fullPromptForCall1_System = getSystemPromptSummarizationStepOne(); - fullPromptForCall1_User = '\n\n------\n\n'+selectedContentItems.join('\n\n------\n\n')+'\n\n------\n\n'; // Keep this for logging/error reporting if needed + // fullPromptForCall1_System = getSystemPromptSummarizationStepOne(); + // fullPromptForCall1_User = '\n\n------\n\n'+selectedContentItems.join('\n\n------\n\n')+'\n\n------\n\n'; // Keep this for logging/error reporting if needed - console.log("Call 1 to Chat (Summarization): User prompt length:", fullPromptForCall1_User.length); - try { - const chunkSize = 3; - const summaryPromises = []; + // console.log("Call 1 to Chat (Summarization): User prompt length:", fullPromptForCall1_User.length); + // try { + // const chunkSize = 3; + // const summaryPromises = []; - for (let i = 0; i < selectedContentItems.length; i += chunkSize) { - const chunk = selectedContentItems.slice(i, i + chunkSize); - const chunkPrompt = chunk.join('\n\n---\n\n'); // Join selected items with the separator + // for (let i = 0; i < selectedContentItems.length; i += chunkSize) { + // const chunk = selectedContentItems.slice(i, i + chunkSize); + // const chunkPrompt = chunk.join('\n\n---\n\n'); // Join selected items with the separator - summaryPromises.push((async () => { - let summarizedChunks = []; - for await (const streamChunk of callChatAPIStream(env, chunkPrompt, fullPromptForCall1_System)) { - summarizedChunks.push(streamChunk); - } - return summarizedChunks.join(''); - })()); - } + // summaryPromises.push((async () => { + // let summarizedChunks = []; + // for await (const streamChunk of callChatAPIStream(env, chunkPrompt, fullPromptForCall1_System)) { + // summarizedChunks.push(streamChunk); + // } + // return summarizedChunks.join(''); + // })()); + // } - const allSummarizedResults = await Promise.all(summaryPromises); - outputOfCall1 = allSummarizedResults.join('\n\n'); // Join all summarized parts + // const allSummarizedResults = await Promise.all(summaryPromises); + // outputOfCall1 = allSummarizedResults.join('\n\n'); // Join all summarized parts - if (!outputOfCall1 || outputOfCall1.trim() === "") throw new Error("Chat summarization call returned empty content."); - outputOfCall1 = removeMarkdownCodeBlock(outputOfCall1); // Clean the output - console.log("Call 1 (Summarization) successful. Output length:", outputOfCall1.length); - } catch (error) { - console.error("Error in Chat API Call 1 (Summarization):", error); - const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(分段处理)', `

Failed during summarization: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall1_System, fullPromptForCall1_User); - return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); - } + // if (!outputOfCall1 || outputOfCall1.trim() === "") throw new Error("Chat summarization call returned empty content."); + // outputOfCall1 = removeMarkdownCodeBlock(outputOfCall1); // Clean the output + // console.log("Call 1 (Summarization) successful. Output length:", outputOfCall1.length); + // } catch (error) { + // console.error("Error in Chat API Call 1 (Summarization):", error); + // const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(分段处理)', `

Failed during summarization: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall1_System, fullPromptForCall1_User); + // return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); + // } // Call 2: Process outputOfCall1 let outputOfCall2 = null; - let fullPromptForCall2_System = getSystemPromptSummarizationStepTwo(); // Re-using summarization prompt for now - let fullPromptForCall2_User = outputOfCall1; // Input for Call 2 is output of Call 1 + let fullPromptForCall2_System = getSystemPromptSummarizationStepOne(); // Re-using summarization prompt for now + let fullPromptForCall2_User = '\n\n------\n\n'+selectedContentItems.join('\n\n------\n\n')+'\n\n------\n\n'; // Input for Call 2 is output of Call 1 console.log("Call 2 to Chat (Processing Call 1 Output): User prompt length:", fullPromptForCall2_User.length); try { @@ -241,25 +261,28 @@ export async function handleGenAIContent(request, env) { console.log("Call 2 (Processing Call 1 Output) successful. Output length:", outputOfCall2.length); } catch (error) { console.error("Error in Chat API Call 2 (Processing Call 1 Output):", error); - const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(格式化)', `

Failed during processing of summarized content: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall1_System, fullPromptForCall1_User, fullPromptForCall2_System, fullPromptForCall2_User); + const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(格式化)', `

Failed during processing of summarized content: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall2_System, fullPromptForCall2_User); return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); } let promptsMarkdownContent = `# Prompts for ${dateStr}\n\n`; - promptsMarkdownContent += `## Call 1: Content Summarization\n\n`; - if (fullPromptForCall1_System) promptsMarkdownContent += `### System Instruction\n\`\`\`\n${fullPromptForCall1_System}\n\`\`\`\n\n`; - if (fullPromptForCall1_User) promptsMarkdownContent += `### User Input\n\`\`\`\n${fullPromptForCall1_User}\n\`\`\`\n\n`; + // promptsMarkdownContent += `## Call 1: Content Summarization\n\n`; + // if (fullPromptForCall1_System) promptsMarkdownContent += `### System Instruction\n\`\`\`\n${fullPromptForCall1_System}\n\`\`\`\n\n`; + // if (fullPromptForCall1_User) promptsMarkdownContent += `### User Input\n\`\`\`\n${fullPromptForCall1_User}\n\`\`\`\n\n`; promptsMarkdownContent += `## Call 2: Summarized Content Format\n\n`; if (fullPromptForCall2_System) promptsMarkdownContent += `### System Instruction\n\`\`\`\n${fullPromptForCall2_System}\n\`\`\`\n\n`; if (fullPromptForCall2_User) promptsMarkdownContent += `### User Input (Output of Call 1)\n\`\`\`\n${fullPromptForCall2_User}\n\`\`\`\n\n`; let dailySummaryMarkdownContent = `## ${env.DAILY_TITLE} ${formatDateToChinese(dateStr)}` + '\n\n'; dailySummaryMarkdownContent += '> '+ env.DAILY_TITLE_MIN + '\n\n'; + + let fullPromptForCall3_System = getSystemPromptSummarizationStepThree(); // Re-using summarization prompt for now + let fullPromptForCall3_User = outputOfCall2; // Input for Call 2 is output of Call 1 let outputOfCall3 = null; - console.log("Call 3 to Chat (Processing Call 2 Output): User prompt length:", outputOfCall2.length); + console.log("Call 3 to Chat (Processing Call 1 Output): User prompt length:", fullPromptForCall3_User.length); try { let processedChunks = []; - for await (const chunk of callChatAPIStream(env, outputOfCall2, getSystemPromptSummarizationStepThree())) { + for await (const chunk of callChatAPIStream(env, fullPromptForCall3_User, fullPromptForCall3_System)) { processedChunks.push(chunk); } outputOfCall3 = processedChunks.join(''); @@ -268,12 +291,13 @@ export async function handleGenAIContent(request, env) { console.log("Call 3 (Processing Call 2 Output) successful. Output length:", outputOfCall3.length); } catch (error) { console.error("Error in Chat API Call 3 (Processing Call 2 Output):", error); - const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(摘要)', `

Failed during processing of summarized content: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall1_System, fullPromptForCall1_User, fullPromptForCall2_System, fullPromptForCall2_User); + const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错(摘要)', `

Failed during processing of summarized content: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, dateStr, true, selectedItemsParams, fullPromptForCall3_System, fullPromptForCall3_User); return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); } - dailySummaryMarkdownContent += '\n\n### **AI内容摘要**\n\n```\n' + outputOfCall3 + '\n```\n\n'; + dailySummaryMarkdownContent += '\n\n### **今日摘要**\n\n```\n' + outputOfCall3 + '\n```\n\n'; dailySummaryMarkdownContent += `\n\n${removeMarkdownCodeBlock(outputOfCall2)}`; + if (env.INSERT_AD=='true') dailySummaryMarkdownContent += insertAd() +`\n`; if (env.INSERT_FOOT=='true') dailySummaryMarkdownContent += insertFoot() +`\n\n`; const successHtml = generateGenAiPageHtml( @@ -281,12 +305,11 @@ export async function handleGenAIContent(request, env) { 'AI日报', // Title for Call 1 page escapeHtml(dailySummaryMarkdownContent), dateStr, false, selectedItemsParams, - fullPromptForCall1_System, fullPromptForCall1_User, + fullPromptForCall2_System, fullPromptForCall2_User, null, null, // Pass Call 2 prompts convertEnglishQuotesToChinese(removeMarkdownCodeBlock(promptsMarkdownContent)), convertEnglishQuotesToChinese(dailySummaryMarkdownContent), null, // No podcast script for this page - outputOfCall1 // Pass summarized content for the next step (original outputOfCall1) ); return new Response(successHtml, { headers: { 'Content-Type': 'text/html; charset=utf-8' } }); @@ -294,7 +317,7 @@ export async function handleGenAIContent(request, env) { console.error("Error in /genAIContent (outer try-catch):", error); const pageDateForError = dateStr || getISODate(); const itemsForActionOnError = Array.isArray(selectedItemsParams) ? selectedItemsParams : []; - const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错', `

Unexpected error: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, pageDateForError, true, itemsForActionOnError, fullPromptForCall1_System, fullPromptForCall1_User, fullPromptForCall2_System, fullPromptForCall2_User); + const errorHtml = generateGenAiPageHtml(env, '生成AI日报出错', `

Unexpected error: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, pageDateForError, true, itemsForActionOnError, fullPromptForCall2_System, fullPromptForCall2_User); return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); } } diff --git a/src/handlers/genAIDailyPage.js b/src/handlers/genAIDailyPage.js new file mode 100644 index 0000000..5ba2296 --- /dev/null +++ b/src/handlers/genAIDailyPage.js @@ -0,0 +1,42 @@ +import { getISODate, escapeHtml, formatDateToChinese, convertEnglishQuotesToChinese} from '../helpers.js'; +import { generateGenAiPageHtml } from '../htmlGenerators.js'; +import { insertFoot } from '../foot.js'; +import { insertAd } from '../ad.js'; + +export async function handleGenAIDailyPage(request, env) { + let dateStr; + try { + const url = new URL(request.url); + const dateParam = url.searchParams.get('date'); + dateStr = dateParam ? dateParam : getISODate(); + + let dailySummaryMarkdownContent = `## ${env.DAILY_TITLE} ${formatDateToChinese(dateStr)}` + '\n\n'; + dailySummaryMarkdownContent += '> '+ env.DAILY_TITLE_MIN + '\n\n'; + + dailySummaryMarkdownContent += '\n\n### **今日摘要**\n\n```\n' + '这里输入内容摘要' + '\n```\n\n'; + if (env.INSERT_AD=='true') dailySummaryMarkdownContent += insertAd() +`\n`; + if (env.INSERT_FOOT=='true') dailySummaryMarkdownContent += insertFoot() +`\n\n`; + + const successHtml = generateGenAiPageHtml( + env, + 'AI日报', // Title for the page + escapeHtml(dailySummaryMarkdownContent), + dateStr, + false, // isError + [], // selectedItemsParams (not applicable here) + null, null, // Call 1 prompts (not applicable here) + null, null, // Call 2 prompts (not applicable here) + 'webbuild', // promptsMarkdownContent (not applicable here) + convertEnglishQuotesToChinese(dailySummaryMarkdownContent), // dailySummaryMarkdownContent + null, // podcastScriptMarkdownContent (not applicable here) + true, // readGithub + ); + return new Response(successHtml, { headers: { 'Content-Type': 'text/html; charset=utf-8' } }); + + } catch (error) { + console.error("Error in /genAIDailyPage:", error); + const pageDateForError = dateStr || getISODate(); + const errorHtml = generateGenAiPageHtml(env, '生成AI日报页面出错', `

Unexpected error: ${escapeHtml(error.message)}

${error.stack ? `
${escapeHtml(error.stack)}
` : ''}`, pageDateForError, true, []); + return new Response(errorHtml, { status: 500, headers: { 'Content-Type': 'text/html; charset=utf-8' } }); + } +} diff --git a/src/handlers/writeRssData.js b/src/handlers/writeRssData.js index 27df630..8a796db 100644 --- a/src/handlers/writeRssData.js +++ b/src/handlers/writeRssData.js @@ -1,4 +1,4 @@ -import { replaceImageProxy, formatMarkdownText, formatDateToGMT12WithTime } from '../helpers.js'; +import { replaceImageProxy, formatMarkdownText, formatDateToGMT8WithTime, removeMarkdownCodeBlock } from '../helpers.js'; import { getDailyReportContent } from '../github.js'; import { storeInKV } from '../kv.js'; import { marked } from '../marked.esm.js'; @@ -26,7 +26,7 @@ export async function handleWriteRssData(request, env) { } console.log(`[writeRssData] Successfully retrieved content for ${path}. Content length: ${content.length}`); - //content = extractContentFromSecondHash(content); + // content = extractContentFromSecondHash(content); // 从 "YYYY-MM-DD" 格式的 dateStr 中提取 "YYYY-MM" const yearMonth = dateStr.substring(0, 7); const report = { @@ -35,9 +35,11 @@ export async function handleWriteRssData(request, env) { link: '/'+yearMonth+'/'+dateStr+'/', content_html: null, // 可以添加其他相關欄位，例如作者、來源等 - published_date: formatDateToGMT12WithTime(new Date()) // 記錄保存時間 + published_date: formatDateToGMT8WithTime(new Date()) // 記錄保存時間 } report.content_html = marked.parse(formatMarkdownText(replaceImageProxy(env, content))); + //report.content_html = marked.parse(formatMarkdownText(await generateAIContent(env, content))); + const kvKey = `${dateStr}-report`; console.log(`[writeRssData] Preparing to store report in KV. Key: ${kvKey}, Report object:`, JSON.stringify(report).substring(0, 200) + '...'); // Log first 200 chars @@ -64,11 +66,12 @@ export function extractContentFromSecondHash(content) { const parts = content.split('###'); if (parts.length > 2) { // 原始逻辑：重新组合从第二个 ### 开始的所有部分 - const newcontent = '###' + parts.slice(2).join('###'); - const lastHashIndex = newcontent.lastIndexOf('###'); + let newcontent = '###' + parts.slice(2).join('###'); + const lastHashIndex = newcontent.lastIndexOf('AI资讯日报语音版'); if (lastHashIndex !== -1) { - return newcontent.substring(0, lastHashIndex); + newcontent = newcontent.substring(0, lastHashIndex-10); } + return newcontent; } return content; // 如果没有找到 ### 或不符合上述条件，则返回原始内容 } @@ -87,8 +90,8 @@ export async function generateAIContent(env, promptText) { try { let result = await callChatAPI(env, promptText, getSummarizationSimplifyPrompt()); console.log(`[generateAIContent] AI model returned content. Length: ${result.length}`); - - result += "\n\n
"+env.INSERT_APP_URL; + result = removeMarkdownCodeBlock(result); + result += "\n\n
"+env.INSERT_APP_URL; return result; } catch (error) { console.error('[generateAIContent] Error calling AI model:', error.message, error.stack); diff --git a/src/helpers.js b/src/helpers.js index ca9445e..28a77b3 100644 --- a/src/helpers.js +++ b/src/helpers.js @@ -257,7 +257,7 @@ export function formatRssDate(date) { return new Intl.DateTimeFormat('zh-CN', options).format(date); } - export function formatDateToGMT12WithTime(isoDateString) { + export function formatDateToGMT8WithTime(isoDateString) { if (!isoDateString) return ''; const date = new Date(isoDateString); const options = { @@ -268,7 +268,7 @@ export function formatRssDate(date) { minute: '2-digit', second: '2-digit', hour12: false, // 使用24小时制 - timeZone: 'Asia/Kamchatka'// 指定东12时区 + timeZone: 'Asia/Shanghai'// 指定东8时区 }; // 使用 'zh-CN' 语言环境以确保中文格式 return new Intl.DateTimeFormat('zh-CN', options).format(date); diff --git a/src/htmlGenerators.js b/src/htmlGenerators.js index 8fdda7e..f5ac858 100644 --- a/src/htmlGenerators.js +++ b/src/htmlGenerators.js @@ -282,7 +282,8 @@ function generatePromptSectionHtmlForGenAI(systemPrompt, userPrompt, promptTitle export function generateGenAiPageHtml(env, title, bodyContent, pageDate, isErrorPage = false, selectedItemsForAction = null, systemP1 = null, userP1 = null, systemP2 = null, userP2 = null, - promptsMd = null, dailyMd = null, podcastMd = null) { + promptsMd = null, dailyMd = null, podcastMd = null, readGithub = null) { + let actionButtonHtml = ''; // Regenerate button for AI Content Summary page @@ -333,6 +334,7 @@ export function generateGenAiPageHtml(env, title, bodyContent, pageDate, isError generatePodcastButtonHtml = `
+ ${selectedItemsForAction.map(item => ``).join('')} @@ -347,7 +349,7 @@ export function generateGenAiPageHtml(env, title, bodyContent, pageDate, isError } let promptDisplayHtml = ''; - if (title === 'AI日报') { + if (title === 'AI日报' || title.includes('生成AI日报出错(')) { if (systemP1 || userP1) { promptDisplayHtml = `
diff --git a/src/index.js b/src/index.js index 4ef07f1..2629fa3 100644 --- a/src/index.js +++ b/src/index.js @@ -2,7 +2,8 @@ import { handleWriteData } from './handlers/writeData.js'; import { handleGetContent } from './handlers/getContent.js'; import { handleGetContentHtml } from './handlers/getContentHtml.js'; -import { handleGenAIContent, handleGenAIPodcastScript, handleGenAIDailyAnalysis } from './handlers/genAIContent.js'; // Import handleGenAIPodcastScript and handleGenAIDailyAnalysis +import { handleGenAIContent, handleGenAIPodcastScript, handleGenAIDailyAnalysis } from './handlers/genAIContent.js'; +import { handleGenAIDailyPage } from './handlers/genAIDailyPage.js'; // Import handleGenAIDailyPage import { handleCommitToGitHub } from './handlers/commitToGitHub.js'; import { handleRss } from './handlers/getRss.js'; import { handleWriteRssData } from './handlers/writeRssData.js'; @@ -18,9 +19,6 @@ export default { 'LOGIN_USERNAME', 'LOGIN_PASSWORD', 'PODCAST_TITLE','PODCAST_BEGIN','PODCAST_END', 'FOLO_COOKIE_KV_KEY','FOLO_DATA_API','FOLO_FILTER_DAYS', - 'AIBASE_FEED_ID', 'XIAOHU_FEED_ID', 'HGPAPERS_FEED_ID', 'TWITTER_LIST_ID', - 'AIBASE_FETCH_PAGES', 'XIAOHU_FETCH_PAGES', 'HGPAPERS_FETCH_PAGES', 'TWITTER_FETCH_PAGES', - //'AIBASE_API_URL', 'XIAOHU_API_URL','PROJECTS_API_URL','HGPAPERS_API_URL', 'TWITTER_API_URL', 'TWITTER_USERNAMES', ]; console.log(env); const missingVars = requiredEnvVars.filter(varName => !env[varName]); @@ -79,6 +77,8 @@ export default { response = await handleGenAIPodcastScript(request, env); } else if (path === '/genAIDailyAnalysis' && request.method === 'POST') { // New route for AI Daily Analysis response = await handleGenAIDailyAnalysis(request, env); + } else if (path === '/genAIDailyPage' && request.method === 'GET') { // New route for AI Daily Page + response = await handleGenAIDailyPage(request, env); } else if (path === '/commitToGitHub' && request.method === 'POST') { response = await handleCommitToGitHub(request, env); } else { diff --git a/src/prompt/summarizationPromptStepZero.js b/src/prompt/summarizationPromptStepZero.js new file mode 100644 index 0000000..546c2cb --- /dev/null +++ b/src/prompt/summarizationPromptStepZero.js @@ -0,0 +1,16 @@ +// Add new data sources +export function getSystemPromptSummarizationStepOne() { + return ` + 你是一名专业的文本摘要助理。你的任务是根据收到的文本类型（或其包含的多种内容类型）执行特定类型的摘要。 + + 重要通用原则：所有摘要内容必须严格来源于原文。不得捏造、歪曲或添加原文未提及的信息。 + + **最终输出要求：** + * 参照以上条件优化文本内容，按内容自动分段，段落数量要和原始一样。 + * 仅输出最终生成的摘要。不要包含任何关于你如何分析文本、确定其类型、分割文本或应用规则的解释性文字。如果合并了来自多个片段的摘要，请确保合并后的文本流畅自然。 + * 输出语言与格式：内容必须为简体中文，并严格采用 Markdown 格式进行排版。 + * 关键词高亮：请在内容中自动识别并对核心关键词或重要概念进行加黑加粗处理，以增强可读性和重点突出。 + * 给最终内容加上标题，前置标题为“### **今日AI资讯**”。 + * 段落序列化：在每个独立段落的开头，必须添加以“1.”开头的阿拉伯数字序列，确保数字正确递增（例如，1.、2.、3.、...）。 + `; +} diff --git a/wrangler.toml b/wrangler.toml index d576edb..b9b5c14 100644 --- a/wrangler.toml +++ b/wrangler.toml @@ -20,23 +20,15 @@ OPENAI_API_URL = "https://api.deepseek.com" # Or your OpenAI compatible API URL DEFAULT_OPEN_MODEL = "deepseek-chat" FOLO_COOKIE_KV_KEY = "folo_auth_cookie" FOLO_DATA_API = "https://api.follow.is/entries" -FOLO_FILTER_DAYS = 3 -AIBASE_FEED_ID = "69533603812632576" -AIBASE_FETCH_PAGES = "3" -XIAOHU_FEED_ID = "151846580097413120" -XIAOHU_FETCH_PAGES = "2" -HGPAPERS_FEED_ID = "41359648680482832" -HGPAPERS_FETCH_PAGES = "2" -JIQIZHIXIN_FEED_ID = "41459996870678583" -JIQIZHIXIN_FETCH_PAGES = "1" -QBIT_FEED_ID = "58864180026527744" -QBIT_FETCH_PAGES = "1" -XINZHIYUAN_FEED_ID = "60901577013168128" -XINZHIYUAN_FETCH_PAGES = "1" +FOLO_FILTER_DAYS = 1 +NEWS_AGGREGATOR_LIST_ID = "158437828119024640" +NEWS_AGGREGATOR_FETCH_PAGES = "1" +HGPAPERS_LIST_ID = "158437917409783808" +HGPAPERS_FETCH_PAGES = "1" TWITTER_LIST_ID = "153028784690326528" -TWITTER_FETCH_PAGES = "2" +TWITTER_FETCH_PAGES = "1" REDDIT_LIST_ID = "167576006499975168" -REDDIT_FETCH_PAGES = "2" +REDDIT_FETCH_PAGES = "1" PROJECTS_API_URL = "https://git-trending.justlikemaki.vip/topone/?since=daily" GITHUB_TOKEN = "github_pat_xxxxxx" GITHUB_REPO_OWNER = "justlovemaki" @@ -51,4 +43,5 @@ PODCAST_BEGIN = "嘿，亲爱的V，欢迎收听新一期的来生情报站， PODCAST_END = "今天的情报就到这里，注意隐蔽，赶紧撤离" BOOK_LINK = "" INSERT_FOOT = "false" +INSERT_AD = "false" INSERT_APP_URL = "

[查看完整版AI日报↗️ https://ai.hubtoday.app/](https://ai.hubtoday.app/)

" \ No newline at end of file