Node js Scraper
我在typescript中写了一个scrapper，在节点上运行：10.12.0，
问题：代码在几个小时后随机进入睡眠状态。我不得不重新启动它。我最好的猜测是它坚持网址请求
工具/包使用：
- 木偶
- Cheerio
- 打字稿
码：
import * as cheerio from "cheerio";
import * as request from "request";
import * as fs from "fs";
import * as shell from "shelljs";
import pup = require("puppeteer");
class App {
// @ts-ignore
public browser: pup.Browser;
public appendToFile(file: string, content: string): Promise < string > {
return new Promise < string > ((resolve, reject) => {
try {
fs.appendFileSync(file, content);
resolve("DONE");
} catch (e) {
reject(e);
}
});
}
public loadPage(url: string): Promise < any > {
return new Promise < any > ((resolve, reject) => {
request.get(url, async (err, res, html) => {
if (!err && res.statusCode === 200) {
resolve(html);
} else {
if (err) {
reject(err);
} else {
reject(res);
}
}
});
});
}
public step1(url: string): Promise < string > {
return new Promise < string > (async (resolve, reject) => {
let page: pup.Page | undefined;
try {
let next = false;
let urlLink = url;
let first = true;
let header = "unknown";
let f = url.split("/");
let folder = f[f.length - 3];
folder = folder || header;
let path = "data/" + folder;
shell.mkdir("-p", path);
page = await this.browser.newPage();
await page.goto(url, {
timeout: 0
});
let count = 1;
do {
next = false;
let res = await page.evaluate(() => {
let e = document.querySelectorAll(".ch-product-view-list-container.list-view li ul > li > h6 > a");
let p: string[] = [];
e.forEach((v) => {
p.push(("") + (v.getAttribute("href") as string));
});
return p;
});
// for(const l of res) {
// try {
// await this.step2(l, "" , "")
// } catch(er) {
// this.appendToFile("./error.txt", l + "::" + url + "\n").catch(e=>e)
// }
// }
let p = [];
let c = 1;
for (const d of res) {
p.push(await this.step2(d, folder, c.toString()).catch((_e) => {
console.log(_e);
fs.appendFileSync("./error-2.txt", urlLink + " ### " + d + "\n");
}));
c++;
}
await Promise.all(p);
await this.appendToFile("./processed.txt", urlLink + ":" + count.toString() + "\n").catch(e => e);
count++;
console.log(urlLink + ":" + count);
let e = await page.evaluate(() => {
let ele = document.querySelector("#pagination-next") as Element;
let r = ele.getAttribute("style");
return r || "";
});
if (e === "") {
next = true;
await page.click("#pagination-next");
// console.log('waitng')
await page.waitFor(1000);
// console.log('done wait')
// await page.waitForNavigation({waitUntil: 'load'}).catch(e=> console.log(e));
// await Promise.all([
// page.click("#pagination-next"),
// page.waitForNavigation({ waitUntil: 'networkidle0'}), // ]);
}
} while (next);
// await page.close();
resolve("page all scrapped");
} catch (errrr) {
reject(errrr);
} finally {
if (page !== undefined) {
await page.close().catch(e => e);
}
}
});
}
public step2(url: string, folder: string, file: string): Promise < string > {
return new Promise < string > (async (resolve, reject) => {
try {
let html = await this.loadPage(url).catch(e => reject(e));
let $ = cheerio.load(html);
let ress: any = {};
let t = $(".qal_title_heading").text();
if (t) {
ress.header = t.replace(/"/g, "'").replace(/\n|\r|\t/g, "");
}
let d = $("div.ch_formatted_text.qal_thread-content_text.asker").html();
if (d) {
ress.body = d.replace(/"/g, "'").replace(/\n|\r|\t/g, "");
}
// let sprit = "-------------------------------";
let filename = "data" + file + ".json"; // ((t.replace(/[^\w\s]/gi, "")).substring(0,250)+".txt")
let data = JSON.stringify(ress) // t +sprit + d + "\n---end---\n"; await this.appendToFile("./data/"+ folder + "/" +filename, data+",\n")
.then((r) => {
resolve(r);
});
} catch (err) {
reject(err);
}
});
}
}
async function main() {
process.on("SIGTERM", () => {
console.log("SigTerm received");
process.exit(1);
});
process.on("SIGINT", () => {
console.log("SigInt received");
process.exit(1);
});
let path = "data/unknown";
shell.mkdir("-p", path);
let c = new App();
let list: string[] = [];
console.log(process.argv[2]);
require("fs").readFileSync(process.argv[2], "utf-8").split(/\r?\n/).forEach((line: string) => {
list.push(line);
});
console.log("total links->" + list.length);
c.browser = await pup.launch({
headless: true
});
for (const l of list) {
await c.step1(l).then(e => {
fs.appendFileSync("./processed.txt", l);
}).catch(e => {
fs.appendFileSync("./error.txt", l);
});
}
}
main();
如果你需要我的其他东西，请告诉我。这也是所有代码。
回答如下：所以，我想到了两个问题。
- Chrome（在木偶操纵者下）消耗高CPU，这给出了这样的趋势： 在开始它是在适度使用。它逐渐增加。我的趋势是从4％的使用率开始，一天后，它达到了100％。我在他们的git上提交了一个问题
- 我没有在请求中指定超时
是：
request.get(url, async (err, res, html) => {应该：request.get(url,{timeout: 1500} async (err, res, html) => {
到目前为止，我的代码运行正常超过一天。唯一的问题是高CPU使用率。但这不是我现在关注的问题。
Node js Scraper
我在typescript中写了一个scrapper，在节点上运行：10.12.0，
问题：代码在几个小时后随机进入睡眠状态。我不得不重新启动它。我最好的猜测是它坚持网址请求
工具/包使用：
- 木偶
- Cheerio
- 打字稿
码：
import * as cheerio from "cheerio";
import * as request from "request";
import * as fs from "fs";
import * as shell from "shelljs";
import pup = require("puppeteer");
class App {
// @ts-ignore
public browser: pup.Browser;
public appendToFile(file: string, content: string): Promise < string > {
return new Promise < string > ((resolve, reject) => {
try {
fs.appendFileSync(file, content);
resolve("DONE");
} catch (e) {
reject(e);
}
});
}
public loadPage(url: string): Promise < any > {
return new Promise < any > ((resolve, reject) => {
request.get(url, async (err, res, html) => {
if (!err && res.statusCode === 200) {
resolve(html);
} else {
if (err) {
reject(err);
} else {
reject(res);
}
}
});
});
}
public step1(url: string): Promise < string > {
return new Promise < string > (async (resolve, reject) => {
let page: pup.Page | undefined;
try {
let next = false;
let urlLink = url;
let first = true;
let header = "unknown";
let f = url.split("/");
let folder = f[f.length - 3];
folder = folder || header;
let path = "data/" + folder;
shell.mkdir("-p", path);
page = await this.browser.newPage();
await page.goto(url, {
timeout: 0
});
let count = 1;
do {
next = false;
let res = await page.evaluate(() => {
let e = document.querySelectorAll(".ch-product-view-list-container.list-view li ul > li > h6 > a");
let p: string[] = [];
e.forEach((v) => {
p.push(("") + (v.getAttribute("href") as string));
});
return p;
});
// for(const l of res) {
// try {
// await this.step2(l, "" , "")
// } catch(er) {
// this.appendToFile("./error.txt", l + "::" + url + "\n").catch(e=>e)
// }
// }
let p = [];
let c = 1;
for (const d of res) {
p.push(await this.step2(d, folder, c.toString()).catch((_e) => {
console.log(_e);
fs.appendFileSync("./error-2.txt", urlLink + " ### " + d + "\n");
}));
c++;
}
await Promise.all(p);
await this.appendToFile("./processed.txt", urlLink + ":" + count.toString() + "\n").catch(e => e);
count++;
console.log(urlLink + ":" + count);
let e = await page.evaluate(() => {
let ele = document.querySelector("#pagination-next") as Element;
let r = ele.getAttribute("style");
return r || "";
});
if (e === "") {
next = true;
await page.click("#pagination-next");
// console.log('waitng')
await page.waitFor(1000);
// console.log('done wait')
// await page.waitForNavigation({waitUntil: 'load'}).catch(e=> console.log(e));
// await Promise.all([
// page.click("#pagination-next"),
// page.waitForNavigation({ waitUntil: 'networkidle0'}), // ]);
}
} while (next);
// await page.close();
resolve("page all scrapped");
} catch (errrr) {
reject(errrr);
} finally {
if (page !== undefined) {
await page.close().catch(e => e);
}
}
});
}
public step2(url: string, folder: string, file: string): Promise < string > {
return new Promise < string > (async (resolve, reject) => {
try {
let html = await this.loadPage(url).catch(e => reject(e));
let $ = cheerio.load(html);
let ress: any = {};
let t = $(".qal_title_heading").text();
if (t) {
ress.header = t.replace(/"/g, "'").replace(/\n|\r|\t/g, "");
}
let d = $("div.ch_formatted_text.qal_thread-content_text.asker").html();
if (d) {
ress.body = d.replace(/"/g, "'").replace(/\n|\r|\t/g, "");
}
// let sprit = "-------------------------------";
let filename = "data" + file + ".json"; // ((t.replace(/[^\w\s]/gi, "")).substring(0,250)+".txt")
let data = JSON.stringify(ress) // t +sprit + d + "\n---end---\n"; await this.appendToFile("./data/"+ folder + "/" +filename, data+",\n")
.then((r) => {
resolve(r);
});
} catch (err) {
reject(err);
}
});
}
}
async function main() {
process.on("SIGTERM", () => {
console.log("SigTerm received");
process.exit(1);
});
process.on("SIGINT", () => {
console.log("SigInt received");
process.exit(1);
});
let path = "data/unknown";
shell.mkdir("-p", path);
let c = new App();
let list: string[] = [];
console.log(process.argv[2]);
require("fs").readFileSync(process.argv[2], "utf-8").split(/\r?\n/).forEach((line: string) => {
list.push(line);
});
console.log("total links->" + list.length);
c.browser = await pup.launch({
headless: true
});
for (const l of list) {
await c.step1(l).then(e => {
fs.appendFileSync("./processed.txt", l);
}).catch(e => {
fs.appendFileSync("./error.txt", l);
});
}
}
main();
如果你需要我的其他东西，请告诉我。这也是所有代码。
回答如下：所以，我想到了两个问题。
- Chrome（在木偶操纵者下）消耗高CPU，这给出了这样的趋势： 在开始它是在适度使用。它逐渐增加。我的趋势是从4％的使用率开始，一天后，它达到了100％。我在他们的git上提交了一个问题
- 我没有在请求中指定超时
是：
request.get(url, async (err, res, html) => {应该：request.get(url,{timeout: 1500} async (err, res, html) => {
到目前为止，我的代码运行正常超过一天。唯一的问题是高CPU使用率。但这不是我现在关注的问题。