@blocklet/crawler 2.1.246 → 2.1.248

package/lib/cjs/middlewares/crawler.d.ts

@@ -3,10 +3,11 @@ import { PageOptions } from './types';
 export { getRelativePath };
 export declare const getPageContent: ({ url, formatPageContent }: PageOptions) => Promise<string>;
 export declare const getUrlInfoFromCache: (url: string) => Promise<any>;
-export declare const setUrlInfoToCache: ({ url, content, lastmod, }: {
+export declare const setUrlInfoToCache: ({ url, content, lastmod, nextDate, }: {
     url: string;
     content: string;
     lastmod?: string;
+    nextDate?: string;
 }) => Promise<any>;
 export declare const crawlUrl: ({ urls, lastmodMap, formatPageContent, autoCloseBrowserCount, }: {
     urls: string[] | string;

package/lib/cjs/middlewares/crawler.js

@@ -71,18 +71,41 @@ const getUrlInfoFromCache = async url => {
   return cache;
 };
 exports.getUrlInfoFromCache = getUrlInfoFromCache;
+function getNextCrawlDate(lastmod) {
+  const now = /* @__PURE__ */new Date();
+  const lastModTime = lastmod ? new Date(lastmod).getTime() || 0 : 0;
+  const daysDiff = Math.max(0, (now.getTime() - lastModTime) / (24 * 60 * 60 * 1e3));
+  const CRAWL_INTERVALS = /* @__PURE__ */new Map([[[-1, 0], 1],
+  // 无 lastmod
+  [[0, 3], 1],
+  // 3 天内活跃
+  [[3, 7], 3],
+  // 7 天内活跃
+  [[7, 30], 14],
+  // 30 天内活跃
+  [[30, 90], 30],
+  // 90 天内活跃
+  [[90, Infinity], 365]
+  // 长期不活跃
+  ]);
+  const interval = Array.from(CRAWL_INTERVALS.entries()).find(([[min, max]]) => lastmod ? daysDiff > min && daysDiff <= max : min === -1)?.[1] || 90;
+  return new Date(now.getTime() + interval * 24 * 60 * 60 * 1e3).toISOString();
+}
 const setUrlInfoToCache = async ({
   url,
   content,
-  lastmod
+  lastmod,
+  nextDate
 }) => {
   if (!content || !url) {
     return;
   }
+  const lastmodValue = lastmod || ( /* @__PURE__ */new Date()).toISOString();
   return await _utils.useCache.set((0, _utils.getRelativePath)(url), {
     content,
-    lastmod: lastmod || ( /* @__PURE__ */new Date()).toISOString(),
-    updatedAt: ( /* @__PURE__ */new Date()).toISOString()
+    lastmod: lastmodValue,
+    updatedAt: ( /* @__PURE__ */new Date()).toISOString(),
+    nextDate: nextDate || getNextCrawlDate(lastmodValue)
   });
 };
 exports.setUrlInfoToCache = setUrlInfoToCache;
@@ -90,7 +113,7 @@ const crawlUrl = async ({
   urls,
   lastmodMap,
   formatPageContent,
-  autoCloseBrowserCount = 30
+  autoCloseBrowserCount = 50
 }) => {
   if (typeof urls === "string") {
     urls = [urls];
@@ -104,7 +127,7 @@ const crawlUrl = async ({
       try {
         if (index % autoCloseBrowserCount === 0) {
           await (0, _utils.closeBrowser)({
-            trimCache: false
+            trimCache: index % (autoCloseBrowserCount * 5) === 0
           });
         }
         const canCrawl = await (0, _utils.isAcceptCrawler)(url);
@@ -114,10 +137,12 @@ const crawlUrl = async ({
             formatPageContent
           });
           if (pageContent) {
+            const lastmodValue = lastmodMap?.get(url) || ( /* @__PURE__ */new Date()).toISOString();
             await setUrlInfoToCache({
               url,
               content: pageContent,
-              lastmod: lastmodMap?.get(url)
+              lastmod: lastmodValue,
+              nextDate: getNextCrawlDate(lastmodValue)
             });
             _utils.logger.info(`Crawler[${index}] ${url} success`);
           } else if (retryCount < 3) {
@@ -245,10 +270,6 @@ const crawlBlocklet = async () => {
             } catch (error) {}
           }
         };
-        removeElements("script");
-        removeElements("style");
-        removeElements("link");
-        removeElements("noscript");
         removeElements('[id="uploader-container"]');
         removeElements('[class^="uppy-"]');
         removeElements('[id="point-up-component"]');
@@ -262,7 +283,9 @@ const crawlBlocklet = async () => {
     }
   });
   _utils.logger.info(...crawlerLogText("success"));
-  await (0, _utils.closeBrowser)();
+  await (0, _utils.closeBrowser)({
+    trimCache: true
+  });
 };
 exports.crawlBlocklet = crawlBlocklet;
 const CRON_CRAWL_BLOCKLET_KEY = "cron-crawl-blocklet";
@@ -280,7 +303,7 @@ const initCronCrawlBlocklet = ({
         time,
         fn: (0, _debounce.default)(crawlBlocklet),
         options: {
-          runOnInit: true,
+          runOnInit: false,
           ...options
         }
       }],

package/lib/cjs/middlewares/utils.d.ts

@@ -10,7 +10,7 @@ export declare const getBrowser: () => Promise<any>;
 export declare const CRAWLER_FLAG = "x-crawler";
 export declare const isSelfCrawler: (req: any) => boolean;
 export declare const initPage: ({ abortResourceTypes, }?: {
-    abortResourceTypes?: never[] | undefined;
+    abortResourceTypes?: string[] | undefined;
 }) => Promise<any>;
 export declare const getDefaultRobotsUrl: (req: any) => string;
 export declare const getDefaultSitemapUrl: (req: any) => string;

package/lib/cjs/middlewares/utils.js

@@ -111,6 +111,16 @@ const closeBrowser = async ({
     clearCheckBrowserTimer();
     if (trimCache) {
       await _puppeteer.default.trimCache();
+      const {
+        cacheDirectory,
+        temporaryDirectory
+      } = getPuppeteerrc();
+      if (cacheDirectory) {
+        _fsExtra.default.emptyDirSync(cacheDirectory);
+      }
+      if (temporaryDirectory) {
+        _fsExtra.default.emptyDirSync(temporaryDirectory);
+      }
     }
     logger.info("Close browser success");
     if (global.gc) {
@@ -147,15 +157,11 @@ const getBrowser = async () => {
           headless: true,
           args: [
           // docs: https://peter.sh/experiments/chromium-command-line-switches/
-          "--no-first-run",
-          // '--no-startup-window',
-          "--hide-scrollbars", "--no-sandbox", "--no-zygote",
-          // '--single-process',
-          "--disable-setuid-sandbox", "--disable-gpu", "--disable-dev-shm-usage", "--disable-site-isolation-trials", "--disable-accelerated-2d-canvas", "--disable-extensions", "--js-flags=--max_old_space_size=512",
+          "--no-first-run", "--hide-scrollbars", "--no-sandbox", "--no-zygote", "--disable-setuid-sandbox", "--disable-gpu", "--disable-dev-shm-usage", "--disable-site-isolation-trials", "--disable-accelerated-2d-canvas", "--disable-extensions", "--js-flags=--max_old_space_size=512",
           // 限制V8内存
-          "--disable-background-networking", "--disable-default-apps", "--disable-web-security",
-          // 允许跨域请求
-          "--disable-software-rasterizer", "--disable-crash-reporter"]
+          "--disable-background-networking", "--disable-default-apps",
+          // '--disable-web-security', // 允许跨域请求
+          "--disable-software-rasterizer", "--disable-crash-reporter", "--disable-service-workers", "--no-startup-window", "--single-process", "--disable-gpu", "--disable-notifications", "--disable-infobars"]
         });
         logger.info("Launch browser success");
         const browserWSEndpoint = await browser.wsEndpoint();
@@ -201,8 +207,19 @@ const isSelfCrawler = req => {
 };
 exports.isSelfCrawler = isSelfCrawler;
 const initPage = async ({
-  abortResourceTypes = []
-  // ['image', 'stylesheet', 'font']
+  abortResourceTypes = ["image",
+  // 图片
+  "media",
+  // 媒体文件
+  "font",
+  // 字体
+  "websocket",
+  // websocket 连接
+  "manifest",
+  // manifest 文件
+  "other"
+  // 其他资源
+  ]
 } = {}) => {
   const browser2 = await getBrowser();
   const page = await browser2.newPage();

package/lib/es/middlewares/crawler.d.ts

@@ -3,10 +3,11 @@ import { PageOptions } from './types';
 export { getRelativePath };
 export declare const getPageContent: ({ url, formatPageContent }: PageOptions) => Promise<string>;
 export declare const getUrlInfoFromCache: (url: string) => Promise<any>;
-export declare const setUrlInfoToCache: ({ url, content, lastmod, }: {
+export declare const setUrlInfoToCache: ({ url, content, lastmod, nextDate, }: {
     url: string;
     content: string;
     lastmod?: string;
+    nextDate?: string;
 }) => Promise<any>;
 export declare const crawlUrl: ({ urls, lastmodMap, formatPageContent, autoCloseBrowserCount, }: {
     urls: string[] | string;

package/lib/es/middlewares/crawler.js

@@ -58,25 +58,51 @@ export const getUrlInfoFromCache = async (url) => {
   const cache = await useCache.get(getRelativePath(url));
   return cache;
 };
+function getNextCrawlDate(lastmod) {
+  const now = /* @__PURE__ */ new Date();
+  const lastModTime = lastmod ? new Date(lastmod).getTime() || 0 : 0;
+  const daysDiff = Math.max(0, (now.getTime() - lastModTime) / (24 * 60 * 60 * 1e3));
+  const CRAWL_INTERVALS = /* @__PURE__ */ new Map([
+    [[-1, 0], 1],
+    // 无 lastmod
+    [[0, 3], 1],
+    // 3 天内活跃
+    [[3, 7], 3],
+    // 7 天内活跃
+    [[7, 30], 14],
+    // 30 天内活跃
+    [[30, 90], 30],
+    // 90 天内活跃
+    [[90, Infinity], 365]
+    // 长期不活跃
+  ]);
+  const interval = Array.from(CRAWL_INTERVALS.entries()).find(
+    ([[min, max]]) => lastmod ? daysDiff > min && daysDiff <= max : min === -1
+  )?.[1] || 90;
+  return new Date(now.getTime() + interval * 24 * 60 * 60 * 1e3).toISOString();
+}
 export const setUrlInfoToCache = async ({
   url,
   content,
-  lastmod
+  lastmod,
+  nextDate
 }) => {
   if (!content || !url) {
     return;
   }
+  const lastmodValue = lastmod || (/* @__PURE__ */ new Date()).toISOString();
   return await useCache.set(getRelativePath(url), {
     content,
-    lastmod: lastmod || (/* @__PURE__ */ new Date()).toISOString(),
-    updatedAt: (/* @__PURE__ */ new Date()).toISOString()
+    lastmod: lastmodValue,
+    updatedAt: (/* @__PURE__ */ new Date()).toISOString(),
+    nextDate: nextDate || getNextCrawlDate(lastmodValue)
   });
 };
 export const crawlUrl = async ({
   urls,
   lastmodMap,
   formatPageContent,
-  autoCloseBrowserCount = 30
+  autoCloseBrowserCount = 50
 }) => {
   if (typeof urls === "string") {
     urls = [urls];
@@ -86,7 +112,7 @@ export const crawlUrl = async ({
       try {
         if (index % autoCloseBrowserCount === 0) {
           await closeBrowser({
-            trimCache: false
+            trimCache: index % (autoCloseBrowserCount * 5) === 0
           });
         }
         const canCrawl = await isAcceptCrawler(url);
@@ -96,7 +122,13 @@ export const crawlUrl = async ({
             formatPageContent
           });
           if (pageContent) {
-            await setUrlInfoToCache({ url, content: pageContent, lastmod: lastmodMap?.get(url) });
+            const lastmodValue = lastmodMap?.get(url) || (/* @__PURE__ */ new Date()).toISOString();
+            await setUrlInfoToCache({
+              url,
+              content: pageContent,
+              lastmod: lastmodValue,
+              nextDate: getNextCrawlDate(lastmodValue)
+            });
             logger.info(`Crawler[${index}] ${url} success`);
           } else if (retryCount < 3) {
             retryCount++;
@@ -213,10 +245,6 @@ export const crawlBlocklet = async () => {
             }
           }
         };
-        removeElements("script");
-        removeElements("style");
-        removeElements("link");
-        removeElements("noscript");
         removeElements('[id="uploader-container"]');
         removeElements('[class^="uppy-"]');
         removeElements('[id="point-up-component"]');
@@ -230,7 +258,9 @@ export const crawlBlocklet = async () => {
     }
   });
   logger.info(...crawlerLogText("success"));
-  await closeBrowser();
+  await closeBrowser({
+    trimCache: true
+  });
 };
 const CRON_CRAWL_BLOCKLET_KEY = "cron-crawl-blocklet";
 let cronCrawlBlockletJob = null;
@@ -247,7 +277,7 @@ export const initCronCrawlBlocklet = ({
           name: CRON_CRAWL_BLOCKLET_KEY,
           time,
           fn: debounce(crawlBlocklet),
-          options: { runOnInit: true, ...options }
+          options: { runOnInit: false, ...options }
         }
       ],
       onError: (err) => {

package/lib/es/middlewares/utils.d.ts

@@ -10,7 +10,7 @@ export declare const getBrowser: () => Promise<any>;
 export declare const CRAWLER_FLAG = "x-crawler";
 export declare const isSelfCrawler: (req: any) => boolean;
 export declare const initPage: ({ abortResourceTypes, }?: {
-    abortResourceTypes?: never[] | undefined;
+    abortResourceTypes?: string[] | undefined;
 }) => Promise<any>;
 export declare const getDefaultRobotsUrl: (req: any) => string;
 export declare const getDefaultSitemapUrl: (req: any) => string;

package/lib/es/middlewares/utils.js

@@ -54,6 +54,13 @@ export const closeBrowser = async ({ trimCache = true } = {}) => {
     clearCheckBrowserTimer();
     if (trimCache) {
       await puppeteer.trimCache();
+      const { cacheDirectory, temporaryDirectory } = getPuppeteerrc();
+      if (cacheDirectory) {
+        fs.emptyDirSync(cacheDirectory);
+      }
+      if (temporaryDirectory) {
+        fs.emptyDirSync(temporaryDirectory);
+      }
     }
     logger.info("Close browser success");
     if (global.gc) {
@@ -90,11 +97,9 @@ export const getBrowser = async () => {
           args: [
             // docs: https://peter.sh/experiments/chromium-command-line-switches/
             "--no-first-run",
-            // '--no-startup-window',
             "--hide-scrollbars",
             "--no-sandbox",
             "--no-zygote",
-            // '--single-process',
             "--disable-setuid-sandbox",
             "--disable-gpu",
             "--disable-dev-shm-usage",
@@ -105,10 +110,15 @@ export const getBrowser = async () => {
             // 限制V8内存
             "--disable-background-networking",
             "--disable-default-apps",
-            "--disable-web-security",
-            // 允许跨域请求
+            // '--disable-web-security', // 允许跨域请求
             "--disable-software-rasterizer",
-            "--disable-crash-reporter"
+            "--disable-crash-reporter",
+            "--disable-service-workers",
+            "--no-startup-window",
+            "--single-process",
+            "--disable-gpu",
+            "--disable-notifications",
+            "--disable-infobars"
           ]
         });
         logger.info("Launch browser success");
@@ -153,8 +163,20 @@ export const isSelfCrawler = (req) => {
   return req.get(CRAWLER_FLAG) === "true" || `${ua}`.toLowerCase().indexOf("headless") !== -1;
 };
 export const initPage = async ({
-  abortResourceTypes = []
-  // ['image', 'stylesheet', 'font']
+  abortResourceTypes = [
+    "image",
+    // 图片
+    "media",
+    // 媒体文件
+    "font",
+    // 字体
+    "websocket",
+    // websocket 连接
+    "manifest",
+    // manifest 文件
+    "other"
+    // 其他资源
+  ]
 } = {}) => {
   const browser2 = await getBrowser();
   const page = await browser2.newPage();

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@blocklet/crawler",
-  "version": "2.1.246",
+  "version": "2.1.248",
   "description": "blocklet crawler lib",
   "publishConfig": {
     "access": "public"