fix broken URLs and automate broken URL checks #70

This commit: - Fixes broken URLs using archive.org or other references. - Replaces tenforums.com URLs with better documentation as they tend to return HTTP status code 403 to tests and also are low quality source. - Changes all insecure http sources to https alternatives - Adds integration tests to check for broken URLs - There's logic implemented for having a delay inbetween when sending requests to same domains, however it's not used as the sources can respond to totally parallelized requests. - Run test pipeline weekly to get notified about broken URls without commits
2021-05-05 23:52:06 +02:00
parent 36f0805590
commit db62ed7f3a
14 changed files with 307 additions and 27 deletions
--- a/tests/integration/application/collections/NoDeadDocumentationUrls.spec.ts
+++ b/tests/integration/application/collections/NoDeadDocumentationUrls.spec.ts
@@ -0,0 +1,48 @@
+import 'mocha';
+import { expect } from 'chai';
+import { parseApplication } from '@/application/Parser/ApplicationParser';
+import { IApplication } from '@/domain/IApplication';
+import { IUrlStatus } from './StatusChecker/IUrlStatus';
+import { getUrlStatusesInParallelAsync, IBatchRequestOptions } from './StatusChecker/BatchStatusChecker';
+
+describe('collections', () => {
+    // arrange
+    const app = parseApplication();
+    const urls = collectUniqueUrls(app);
+    const options: IBatchRequestOptions = {
+        domainOptions: {
+            sameDomainParallelize: true, // no need to be so nice until sources start failing
+            // sameDomainDelayInMs: 2 /* sec */ * 1000,
+        },
+        requestOptions: {
+            retryExponentialBaseInMs: 3 /* sec */ * 1000,
+            additionalHeaders: { referer: app.info.homepage },
+        },
+    };
+    const testTimeoutInMs = urls.length * 60000 /* 1 minute */;
+    it('have no dead urls', async () => {
+        // act
+        const results = await getUrlStatusesInParallelAsync(urls, options);
+        // assert
+        const deadUrls = results.filter((r) => r.statusCode !== 200);
+        expect(deadUrls).to.have.lengthOf(0, printUrls(deadUrls));
+    }).timeout(testTimeoutInMs);
+});
+
+function collectUniqueUrls(app: IApplication): string[] {
+    return app
+        .collections
+        .flatMap((a) => a.getAllScripts())
+        .flatMap((script) => script.documentationUrls)
+        .filter((url, index, array) => array.indexOf(url) === index);
+}
+
+function printUrls(statuses: IUrlStatus[]): string {
+    return '\n' +
+        statuses.map((status) =>
+            `- ${status.url}\n` +
+            (status.statusCode ? `\tResponse code: ${status.statusCode}` : '') +
+            (status.error ? `\tException: ${JSON.stringify(status.error, null, '\t')}` : ''))
+        .join(`\n`)
+        + '\n';
+}
--- a/tests/integration/application/collections/StatusChecker/BatchStatusChecker.ts
+++ b/tests/integration/application/collections/StatusChecker/BatchStatusChecker.ts
@@ -0,0 +1,67 @@
+import { sleepAsync } from '@/infrastructure/Threading/AsyncSleep';
+import { IUrlStatus } from './IUrlStatus';
+import { getUrlStatusAsync, IRequestOptions } from './Requestor';
+import { groupUrlsByDomain } from './UrlPerDomainGrouper';
+
+export async function getUrlStatusesInParallelAsync(
+    urls: string[],
+    options?: IBatchRequestOptions): Promise<IUrlStatus[]> {
+    const uniqueUrls = Array.from(new Set(urls));
+    options = { ...DefaultOptions, ...options };
+    console.log('Options: ', options); // tslint:disable-line: no-console
+    const results = await requestAsync(uniqueUrls, options);
+    return results;
+}
+
+export interface IBatchRequestOptions {
+    domainOptions?: IDomainOptions;
+    requestOptions?: IRequestOptions;
+}
+
+interface IDomainOptions {
+    sameDomainParallelize?: boolean;
+    sameDomainDelayInMs?: number;
+}
+
+const DefaultOptions: IBatchRequestOptions = {
+    domainOptions: {
+        sameDomainParallelize: false,
+        sameDomainDelayInMs: 3 /* sec */ * 1000,
+    },
+    requestOptions: {
+        retryExponentialBaseInMs: 5 /* sec */ * 1000,
+        additionalHeaders: {},
+    },
+};
+
+function requestAsync(urls: string[], options: IBatchRequestOptions): Promise<IUrlStatus[]> {
+    if (!options.domainOptions.sameDomainParallelize) {
+        return runOnEachDomainWithDelayAsync(
+            urls,
+            (url) => getUrlStatusAsync(url, options.requestOptions),
+            options.domainOptions.sameDomainDelayInMs);
+    } else {
+        return Promise.all(
+            urls.map((url) => getUrlStatusAsync(url, options.requestOptions)));
+    }
+}
+
+async function runOnEachDomainWithDelayAsync(
+    urls: string[],
+    action: (url: string) => Promise<IUrlStatus>,
+    delayInMs: number): Promise<IUrlStatus[]> {
+    const grouped = groupUrlsByDomain(urls);
+    const tasks = grouped.map(async (group) => {
+        const results = new Array<IUrlStatus>();
+        for (const url of group) {
+            const status = await action(url);
+            results.push(status);
+            if (results.length !== group.length) {
+                await sleepAsync(delayInMs);
+            }
+        }
+        return results;
+    });
+    const r = await Promise.all(tasks);
+    return r.flat();
+}
--- a/tests/integration/application/collections/StatusChecker/ExponentialBackOffRetryHandler.ts
+++ b/tests/integration/application/collections/StatusChecker/ExponentialBackOffRetryHandler.ts
@@ -0,0 +1,44 @@
+import { sleepAsync } from '@/infrastructure/Threading/AsyncSleep';
+import { IUrlStatus } from './IUrlStatus';
+
+const DefaultBaseRetryIntervalInMs = 5 /* sec */ * 1000;
+
+export async function retryWithExponentialBackOffAsync(
+    action: () => Promise<IUrlStatus>,
+    baseRetryIntervalInMs: number = DefaultBaseRetryIntervalInMs,
+    currentRetry = 1): Promise<IUrlStatus> {
+    const maxTries: number = 3;
+    const status = await action();
+    if (shouldRetry(status)) {
+        if (currentRetry <= maxTries) {
+            const exponentialBackOffInMs = getRetryTimeoutInMs(currentRetry, baseRetryIntervalInMs);
+            // tslint:disable-next-line: no-console
+            console.log(`Retrying (${currentRetry}) in ${exponentialBackOffInMs / 1000} seconds`, status);
+            await sleepAsync(exponentialBackOffInMs);
+            return retryWithExponentialBackOffAsync(action, baseRetryIntervalInMs, currentRetry + 1);
+        }
+    }
+    return status;
+}
+
+function shouldRetry(status: IUrlStatus) {
+    if (status.error) {
+        return true;
+    }
+    return isTransientError(status.statusCode)
+        || status.statusCode === 429; // Too Many Requests
+}
+
+function isTransientError(statusCode: number) {
+    return statusCode >= 500 && statusCode <= 599;
+}
+
+function getRetryTimeoutInMs(currentRetry: number, baseRetryIntervalInMs: number = DefaultBaseRetryIntervalInMs) {
+    const retryRandomFactor = 0.5;  // Retry intervals are between 50% and 150%
+                                    // of the exponentially increasing base amount
+    const minRandom = 1 - retryRandomFactor;
+    const maxRandom = 1 + retryRandomFactor;
+    const randomization = (Math.random() * (maxRandom - minRandom)) + maxRandom;
+    const exponential = Math.pow(2, currentRetry - 1);
+    return Math.ceil(exponential * baseRetryIntervalInMs * randomization);
+}
--- a/tests/integration/application/collections/StatusChecker/IUrlStatus.ts
+++ b/tests/integration/application/collections/StatusChecker/IUrlStatus.ts
@@ -0,0 +1,5 @@
+export interface IUrlStatus {
+    url: string;
+    error?: any;
+    statusCode?: number;
+}
--- a/tests/integration/application/collections/StatusChecker/Requestor.ts
+++ b/tests/integration/application/collections/StatusChecker/Requestor.ts
@@ -0,0 +1,47 @@
+import { retryWithExponentialBackOffAsync } from './ExponentialBackOffRetryHandler';
+import { IUrlStatus } from './IUrlStatus';
+import fetch from 'cross-fetch';
+
+export interface IRequestOptions {
+    retryExponentialBaseInMs?: number;
+    additionalHeaders?: Record<string, string>;
+}
+
+export async function getUrlStatusAsync(
+    url: string,
+    options: IRequestOptions = DefaultOptions): Promise<IUrlStatus> {
+    options = { ...DefaultOptions, ...options };
+    const fetchOptions = getFetchOptions(options);
+    return retryWithExponentialBackOffAsync(async () => {
+        console.log('Requesting', url); // tslint:disable-line: no-console
+        try {
+            const response = await fetch(url, fetchOptions);
+            return { url, statusCode: response.status};
+        } catch (err) {
+            return { url, error: err};
+        }
+    }, options.retryExponentialBaseInMs);
+}
+
+const DefaultOptions: IRequestOptions = {
+    retryExponentialBaseInMs: 5000,
+    additionalHeaders: {},
+};
+
+function getFetchOptions(options: IRequestOptions) {
+    return {
+        method: 'GET',
+        headers: { ...DefaultHeaders, ...options.additionalHeaders },
+    };
+}
+
+const DefaultHeaders: Record<string, string> = {
+    /* Chrome on macOS */
+    'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/90.0.4430.93 Safari/537.36',
+    'upgrade-insecure-requests': '1',
+    'connection': 'keep-alive',
+    'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8',
+    'accept-encoding': 'gzip, deflate, br',
+    'cache-control': 'max-age=0',
+    'accept-language': 'en-US,en;q=0.9',
+};
--- a/tests/integration/application/collections/StatusChecker/UrlPerDomainGrouper.ts
+++ b/tests/integration/application/collections/StatusChecker/UrlPerDomainGrouper.ts
@@ -0,0 +1,19 @@
+export function groupUrlsByDomain(urls: string[]): string[][] {
+    const domains = new Set<string>();
+    const urlsWithDomain = urls.map((url) => ({
+        url,
+        domain: extractDomain(url),
+    }));
+    for (const url of urlsWithDomain) {
+        domains.add(url.domain);
+    }
+    return Array.from(domains).map((domain) => {
+        return urlsWithDomain
+            .filter((url) => url.domain === domain)
+            .map((url) => url.url);
+    });
+}
+
+function extractDomain(url: string): string {
+    return url.split('://')[1].split('/')[0].toLowerCase();
+}