ci/cd: trigger URL checks more, and limit amount
Key changes: - Run URL checks more frequently on every change. - Introduce environment variable to randomly select and limit URLs tested, this way the tests will provide quicker feedback on code changes. Other supporting changes: - Log more information about test before running the test to enable easier troubleshooting. - Move shuffle function for arrays for reusability and missing tests.
This commit is contained in:
undergroundwires
69
tests/checks/external-urls/DocumentationUrlExtractor.ts
Normal file
69
tests/checks/external-urls/DocumentationUrlExtractor.ts
Normal file
@@ -0,0 +1,69 @@
|
||||
import type { IApplication } from '@/domain/IApplication';
|
||||
import type { TestExecutionDetailsLogger } from './TestExecutionDetailsLogger';
|
||||
|
||||
interface UrlExtractionContext {
|
||||
readonly logger: TestExecutionDetailsLogger;
|
||||
readonly application: IApplication;
|
||||
readonly urlExclusionPatterns: readonly RegExp[];
|
||||
}
|
||||
|
||||
export function extractDocumentationUrls(
|
||||
context: UrlExtractionContext,
|
||||
): string[] {
|
||||
const urlsInApplication = extractUrlsFromApplication(context.application);
|
||||
context.logger.logLabeledInformation(
|
||||
'Extracted URLs from application',
|
||||
urlsInApplication.length.toString(),
|
||||
);
|
||||
const uniqueUrls = filterDuplicateUrls(urlsInApplication);
|
||||
context.logger.logLabeledInformation(
|
||||
'Unique URLs after deduplication',
|
||||
`${uniqueUrls.length} (duplicates removed)`,
|
||||
);
|
||||
context.logger.logLabeledInformation(
|
||||
'Exclusion patterns for URLs',
|
||||
context.urlExclusionPatterns.length === 0
|
||||
? 'None (all URLs included)'
|
||||
: context.urlExclusionPatterns.map((pattern, index) => `${index + 1}) ${pattern.toString()}`).join('\n'),
|
||||
);
|
||||
const includedUrls = filterUrlsExcludingPatterns(uniqueUrls, context.urlExclusionPatterns);
|
||||
context.logger.logLabeledInformation(
|
||||
'URLs extracted for testing',
|
||||
`${includedUrls.length} (after applying exclusion patterns; ${uniqueUrls.length - includedUrls.length} URLs ignored)`,
|
||||
);
|
||||
return includedUrls;
|
||||
}
|
||||
|
||||
function extractUrlsFromApplication(application: IApplication): string[] {
|
||||
return [ // Get all executables
|
||||
...application.collections.flatMap((c) => c.getAllCategories()),
|
||||
...application.collections.flatMap((c) => c.getAllScripts()),
|
||||
]
|
||||
// Get all docs
|
||||
.flatMap((documentable) => documentable.docs)
|
||||
// Parse all URLs
|
||||
.flatMap((docString) => extractUrlsExcludingCodeBlocks(docString));
|
||||
}
|
||||
|
||||
function filterDuplicateUrls(urls: readonly string[]): string[] {
|
||||
return urls.filter((url, index, array) => array.indexOf(url) === index);
|
||||
}
|
||||
|
||||
function filterUrlsExcludingPatterns(
|
||||
urls: readonly string[],
|
||||
patterns: readonly RegExp[],
|
||||
): string[] {
|
||||
return urls.filter((url) => !patterns.some((pattern) => pattern.test(url)));
|
||||
}
|
||||
|
||||
function extractUrlsExcludingCodeBlocks(textWithInlineCode: string): string[] {
|
||||
/*
|
||||
Matches URLs:
|
||||
- Excludes inline code blocks as they may contain URLs not intended for user interaction
|
||||
and not guaranteed to support expected HTTP methods, leading to false-negatives.
|
||||
- Supports URLs containing parentheses, avoiding matches within code that might not represent
|
||||
actual links.
|
||||
*/
|
||||
const nonCodeBlockUrlRegex = /(?<!`)(https?:\/\/[^\s`"<>()]+(?:\([^\s`"<>()]*\))?[^\s`"<>()]*)/g;
|
||||
return textWithInlineCode.match(nonCodeBlockUrlRegex) || [];
|
||||
}
|
||||
26
tests/checks/external-urls/TestExecutionDetailsLogger.ts
Normal file
26
tests/checks/external-urls/TestExecutionDetailsLogger.ts
Normal file
@@ -0,0 +1,26 @@
|
||||
import { indentText } from '@tests/shared/Text';
|
||||
|
||||
export class TestExecutionDetailsLogger {
|
||||
public logTestSectionStartDelimiter(): void {
|
||||
this.logSectionDelimiterLine();
|
||||
}
|
||||
|
||||
public logTestSectionEndDelimiter(): void {
|
||||
this.logSectionDelimiterLine();
|
||||
}
|
||||
|
||||
public logLabeledInformation(
|
||||
label: string,
|
||||
detailedInformation: string,
|
||||
): void {
|
||||
console.log([
|
||||
`${label}:`,
|
||||
indentText(detailedInformation),
|
||||
].join('\n'));
|
||||
}
|
||||
|
||||
private logSectionDelimiterLine(): void {
|
||||
const horizontalLine = '─'.repeat(40);
|
||||
console.log(horizontalLine);
|
||||
}
|
||||
}
|
||||
@@ -1,19 +1,26 @@
|
||||
import { test, expect } from 'vitest';
|
||||
import { parseApplication } from '@/application/Parser/ApplicationParser';
|
||||
import type { IApplication } from '@/domain/IApplication';
|
||||
import { indentText } from '@tests/shared/Text';
|
||||
import { formatAssertionMessage } from '@tests/shared/FormatAssertionMessage';
|
||||
import { shuffle } from '@/application/Common/Shuffle';
|
||||
import { type UrlStatus, formatUrlStatus } from './StatusChecker/UrlStatus';
|
||||
import { getUrlStatusesInParallel, type BatchRequestOptions } from './StatusChecker/BatchStatusChecker';
|
||||
import { TestExecutionDetailsLogger } from './TestExecutionDetailsLogger';
|
||||
import { extractDocumentationUrls } from './DocumentationUrlExtractor';
|
||||
|
||||
// arrange
|
||||
const logger = new TestExecutionDetailsLogger();
|
||||
logger.logTestSectionStartDelimiter();
|
||||
const app = parseApplication();
|
||||
const urls = collectUniqueUrls({
|
||||
application: app,
|
||||
excludePatterns: [
|
||||
let urls = extractDocumentationUrls({
|
||||
logger,
|
||||
urlExclusionPatterns: [
|
||||
/^https:\/\/archive\.ph/, // Drops HEAD/GET requests via fetch/curl, responding to Postman/Chromium.
|
||||
],
|
||||
application: app,
|
||||
});
|
||||
urls = filterUrlsToEnvironmentCheckLimit(urls);
|
||||
logger.logLabeledInformation('URLs submitted for testing', urls.length.toString());
|
||||
const requestOptions: BatchRequestOptions = {
|
||||
domainOptions: {
|
||||
sameDomainParallelize: false, // be nice to our third-party servers
|
||||
@@ -30,55 +37,65 @@ const requestOptions: BatchRequestOptions = {
|
||||
enableCookies: true,
|
||||
},
|
||||
};
|
||||
logger.logLabeledInformation('HTTP request options', JSON.stringify(requestOptions, null, 2));
|
||||
const testTimeoutInMs = urls.length * 60 /* seconds */ * 1000;
|
||||
logger.logLabeledInformation('Scheduled test duration', convertMillisecondsToHumanReadableFormat(testTimeoutInMs));
|
||||
logger.logTestSectionEndDelimiter();
|
||||
test(`all URLs (${urls.length}) should be alive`, async () => {
|
||||
// act
|
||||
console.log('URLS', urls); // TODO: Delete
|
||||
const results = await getUrlStatusesInParallel(urls, requestOptions);
|
||||
// assert
|
||||
const deadUrls = results.filter((r) => r.code === undefined || !isOkStatusCode(r.code));
|
||||
expect(deadUrls).to.have.lengthOf(0, formatAssertionMessage([formatUrlStatusReport(deadUrls)]));
|
||||
expect(deadUrls).to.have.lengthOf(
|
||||
0,
|
||||
formatAssertionMessage([createReportForDeadUrlStatuses(deadUrls)]),
|
||||
);
|
||||
}, testTimeoutInMs);
|
||||
|
||||
function isOkStatusCode(statusCode: number): boolean {
|
||||
return statusCode >= 200 && statusCode < 300;
|
||||
}
|
||||
|
||||
function collectUniqueUrls(
|
||||
options: {
|
||||
readonly application: IApplication,
|
||||
readonly excludePatterns?: readonly RegExp[],
|
||||
},
|
||||
): string[] {
|
||||
return [ // Get all nodes
|
||||
...options.application.collections.flatMap((c) => c.getAllCategories()),
|
||||
...options.application.collections.flatMap((c) => c.getAllScripts()),
|
||||
]
|
||||
// Get all docs
|
||||
.flatMap((documentable) => documentable.docs)
|
||||
// Parse all URLs
|
||||
.flatMap((docString) => extractUrls(docString))
|
||||
// Remove duplicates
|
||||
.filter((url, index, array) => array.indexOf(url) === index)
|
||||
// Exclude certain URLs based on patterns
|
||||
.filter((url) => !shouldExcludeUrl(url, options.excludePatterns ?? []));
|
||||
}
|
||||
|
||||
function shouldExcludeUrl(url: string, patterns: readonly RegExp[]): boolean {
|
||||
return patterns.some((pattern) => pattern.test(url));
|
||||
}
|
||||
|
||||
function formatUrlStatusReport(deadUrlStatuses: readonly UrlStatus[]): string {
|
||||
function createReportForDeadUrlStatuses(deadUrlStatuses: readonly UrlStatus[]): string {
|
||||
return `\n${deadUrlStatuses.map((status) => indentText(formatUrlStatus(status))).join('\n---\n')}\n`;
|
||||
}
|
||||
|
||||
function extractUrls(textWithInlineCode: string): string[] {
|
||||
/*
|
||||
Matches URLs:
|
||||
- Excludes inline code blocks as they may contain URLs not intended for user interaction
|
||||
and not guaranteed to support expected HTTP methods, leading to false-negatives.
|
||||
- Supports URLs containing parentheses, avoiding matches within code that might not represent
|
||||
actual links.
|
||||
*/
|
||||
const nonCodeBlockUrlRegex = /(?<!`)(https?:\/\/[^\s`"<>()]+(?:\([^\s`"<>()]*\))?[^\s`"<>()]*)/g;
|
||||
return textWithInlineCode.match(nonCodeBlockUrlRegex) || [];
|
||||
function filterUrlsToEnvironmentCheckLimit(originalUrls: string[]): string[] {
|
||||
const { RANDOMIZED_URL_CHECK_LIMIT } = process.env;
|
||||
logger.logLabeledInformation('URL check limit', RANDOMIZED_URL_CHECK_LIMIT || 'Unlimited');
|
||||
if (RANDOMIZED_URL_CHECK_LIMIT !== undefined && RANDOMIZED_URL_CHECK_LIMIT !== '') {
|
||||
const maxUrlsInTest = parseInt(RANDOMIZED_URL_CHECK_LIMIT, 10);
|
||||
if (Number.isNaN(maxUrlsInTest)) {
|
||||
throw new Error(`Invalid URL limit: ${RANDOMIZED_URL_CHECK_LIMIT}`);
|
||||
}
|
||||
if (maxUrlsInTest < originalUrls.length) {
|
||||
return shuffle(originalUrls).slice(0, maxUrlsInTest);
|
||||
}
|
||||
}
|
||||
return originalUrls;
|
||||
}
|
||||
|
||||
function convertMillisecondsToHumanReadableFormat(milliseconds: number): string {
|
||||
const timeParts: string[] = [];
|
||||
const addTimePart = (amount: number, label: string) => {
|
||||
if (amount === 0) {
|
||||
return;
|
||||
}
|
||||
timeParts.push(`${amount} ${label}`);
|
||||
};
|
||||
|
||||
const hours = milliseconds / (1000 * 60 * 60);
|
||||
const absoluteHours = Math.floor(hours);
|
||||
addTimePart(absoluteHours, 'hours');
|
||||
|
||||
const minutes = (hours - absoluteHours) * 60;
|
||||
const absoluteMinutes = Math.floor(minutes);
|
||||
addTimePart(absoluteMinutes, 'minutes');
|
||||
|
||||
const seconds = (minutes - absoluteMinutes) * 60;
|
||||
const absoluteSeconds = Math.floor(seconds);
|
||||
addTimePart(absoluteSeconds, 'seconds');
|
||||
|
||||
return timeParts.join(', ');
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user