mirror of
https://github.com/webrecorder/browsertrix-crawler.git
synced 2025-10-19 06:23:16 +00:00

- check for urls that are wrapped in quotes, eg. 'https://example.com/' or "https://example.com/" and trim and remove the quotes before adding seed - tests: add quoted URL to tests, fix old.webrecorder.net test - deps: update wabac.js, RWP to latest - logging: reduce error logging for seed lists, only log once that there are duplicates or page limit is reached - fix for #882
93 lines
2.4 KiB
JavaScript
93 lines
2.4 KiB
JavaScript
import util from "util";
|
|
import { spawn, exec as execCallback } from "child_process";
|
|
import fs from "fs";
|
|
|
|
const exec = util.promisify(execCallback);
|
|
|
|
let proc = null;
|
|
|
|
const DOCKER_HOST_NAME = process.env.DOCKER_HOST_NAME || "host.docker.internal";
|
|
const TEST_HOST = `http://${DOCKER_HOST_NAME}:31502`;
|
|
|
|
beforeAll(() => {
|
|
proc = spawn("../../node_modules/.bin/http-server", ["-p", "31502"], {cwd: "tests/fixtures/"});
|
|
});
|
|
|
|
afterAll(() => {
|
|
if (proc) {
|
|
proc.kill();
|
|
}
|
|
});
|
|
|
|
|
|
|
|
test("check that URLs in seed-list are crawled", async () => {
|
|
try {
|
|
await exec(
|
|
"docker run -v $PWD/test-crawls:/crawls -v $PWD/tests/fixtures:/tests/fixtures webrecorder/browsertrix-crawler crawl --collection filelisttest --urlFile /tests/fixtures/urlSeedFile.txt --timeout 90000 --scopeType page",
|
|
);
|
|
} catch (error) {
|
|
console.log(error);
|
|
}
|
|
|
|
let crawled_pages = fs.readFileSync(
|
|
"test-crawls/collections/filelisttest/pages/pages.jsonl",
|
|
"utf8",
|
|
);
|
|
let seed_file = fs
|
|
.readFileSync("tests/fixtures/urlSeedFile.txt", "utf8")
|
|
.split("\n")
|
|
.sort();
|
|
|
|
let seed_file_list = [];
|
|
for (var j = 0; j < seed_file.length; j++) {
|
|
if (seed_file[j] != undefined) {
|
|
seed_file_list.push(seed_file[j]);
|
|
}
|
|
}
|
|
|
|
let foundSeedUrl = true;
|
|
|
|
for (var i = 1; i < seed_file_list.length; i++) {
|
|
if (crawled_pages.indexOf(seed_file_list[i]) == -1) {
|
|
foundSeedUrl = false;
|
|
}
|
|
}
|
|
expect(foundSeedUrl).toBe(true);
|
|
});
|
|
|
|
|
|
test("check that URLs in seed-list hosted at URL are crawled", async () => {
|
|
try {
|
|
await exec(
|
|
`docker run -v $PWD/test-crawls:/crawls -v $PWD/tests/fixtures:/tests/fixtures webrecorder/browsertrix-crawler crawl --collection onlinefilelisttest --urlFile "${TEST_HOST}/urlSeedFile.txt" --timeout 90000 --scopeType page`,
|
|
);
|
|
} catch (error) {
|
|
console.log(error);
|
|
}
|
|
|
|
let crawled_pages = fs.readFileSync(
|
|
"test-crawls/collections/onlinefilelisttest/pages/pages.jsonl",
|
|
"utf8",
|
|
);
|
|
let seed_file = fs
|
|
.readFileSync("tests/fixtures/urlSeedFile.txt", "utf8")
|
|
.split("\n")
|
|
.sort();
|
|
|
|
let seed_file_list = [];
|
|
for (var j = 0; j < seed_file.length; j++) {
|
|
if (seed_file[j] != undefined) {
|
|
seed_file_list.push(seed_file[j]);
|
|
}
|
|
}
|
|
|
|
let foundSeedUrl = true;
|
|
|
|
for (var i = 1; i < seed_file_list.length; i++) {
|
|
if (crawled_pages.indexOf(seed_file_list[i]) == -1) {
|
|
foundSeedUrl = false;
|
|
}
|
|
}
|
|
expect(foundSeedUrl).toBe(true);
|
|
});
|