CI: Test for the health of all code base URLs #2127
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,138 @@ | ||
| # This file is distributed under New Relic's license terms. | ||
| # See https://github.com/newrelic/newrelic-ruby-agent/blob/main/LICENSE for complete details. | ||
| # frozen_string_literal: true | ||
|
|
||
| # The unit test class defined herein will verify the health of all URLs found | ||
| # in the project source code. | ||
| # | ||
| # To run the URL health tests by themselves: | ||
| # TEST=test/new_relic/healthy_urls_test bundle exec rake test | ||
| # | ||
| # A file will be scanned for URLs if the file's basename is found in the | ||
| # FILENAMES array OR the file's extension is found in the EXTENSIONS array | ||
| # unless the file's absolute path matches the IGNORED_FILE_PATTERN regex. | ||
| # | ||
| # NOTE that CHANGELOG.md is handled with special logic so that only the most | ||
| # recent 2 versions mentioned in the changelog are scannned for URLs. | ||
| # | ||
| # See TIMEOUT for the number of seconds permitted for a GET request to a given | ||
| # URL to be completed. | ||
| # | ||
| # Enable DEBUG for additional verbosity | ||
|
|
||
| require_relative '../test_helper' | ||
|
|
||
| class HealthyUrlsTest < Minitest::Test | ||
| ROOT = File.expand_path('../../..', __FILE__).freeze | ||
| FILENAMES = %w[ | ||
| baselines | ||
|
There was a problem hiding this comment. I don't think I'm familiar with this file. Where is it and what does it do? There was a problem hiding this comment. It's at |
||
| Brewfile | ||
| Capfile | ||
| Dockerfile | ||
| Envfile | ||
| Gemfile | ||
| Guardfile | ||
| install_mysql55 | ||
| LICENSE | ||
| mega-runner | ||
| newrelic | ||
| newrelic_cmd | ||
| nrdebug | ||
|
Comment on lines
+38
to
+40
There was a problem hiding this comment. Oh, are these executables as well as files? There was a problem hiding this comment. Yeah, these ones are really just Ruby scripts so they're eligible for having URLs in the source code. For now I went with an allowlist approach of filenames and extensions for specifying what to scan for URLs. In the future we may want to reverse this and have a blocklist of things like image files or perhaps even take a quick peek at the file's encoding and make a judgment call from that. |
||
| Rakefile | ||
| run_tests | ||
| runner | ||
| Thorfile | ||
| ].freeze | ||
| EXTENSIONS = %w[ | ||
| css | ||
| erb | ||
| gemspec | ||
| haml | ||
| html | ||
| js | ||
| json | ||
| md | ||
| proto | ||
|
There was a problem hiding this comment. Can we make changes to the proto files? There was a problem hiding this comment. Yes! Not only can we, but we have, and with a URL in fact! There is currently only 1 file at |
||
| rake | ||
| readme | ||
| rb | ||
| sh | ||
| txt | ||
| thor | ||
| tt | ||
| yml | ||
| ].freeze | ||
| FILE_PATTERN = /(?:^(?:#{FILENAMES.join('|')})$)|\.(?:#{EXTENSIONS.join('|')})$/.freeze | ||
| IGNORED_FILE_PATTERN = %r{/(?:coverage|test)/}.freeze | ||
| URL_PATTERN = %r{(https?://.*?)[^a-zA-Z0-9/\.\-_#]}.freeze | ||
| IGNORED_URL_PATTERN = %r{(?:\{|\(|\$|169\.254|\.\.\.|metadata\.google)} | ||
| TIMEOUT = 5 | ||
| DEBUG = false | ||
|
|
||
| def test_all_urls | ||
| skip_unless_ci_cron | ||
| skip_unless_newest_ruby | ||
| load_httparty | ||
|
|
||
| urls = gather_urls | ||
| errors = urls.each_with_object({}) do |(url, _files), hash| | ||
| error = verify_url(url) | ||
| hash[url] = error if error | ||
| end | ||
|
|
||
| msg = "#{errors.keys.size} URLs were unreachable!\n\n" | ||
| msg += errors.map { |url, error| " #{url} - #{error}\n files: #{urls[url].join(',')}" }.join("\n") | ||
|
|
||
| assert_empty errors, msg | ||
| end | ||
|
|
||
| private | ||
|
|
||
| def load_httparty | ||
| require 'httparty' | ||
| rescue | ||
| skip 'Skipping URL health tests in this context, as HTTParty is not available' | ||
| end | ||
|
|
||
| def real_url?(url) | ||
| return false if url.match?(IGNORED_URL_PATTERN) | ||
|
|
||
| true | ||
| end | ||
|
|
||
| def gather_urls | ||
| Dir.glob(File.join(ROOT, '**', '*')).each_with_object({}) do |file, urls| | ||
|
There was a problem hiding this comment. Minor: Can default the Example from docs: https://rubyapi.org/3.2/o/hash Change we can do: Which then would mean that There was a problem hiding this comment. Thanks, @olleolleolle! Implemented with #2132. Using LETTERS = ('a'..'z').to_a
hash1 = LETTERS.each_with_object(Hash.new { |h, k| h[k] = [] }) { |l, h| h[l] << true }
hash2 = LETTERS.each_with_object(Hash.new([])) { |l, h| h[l] += [true] }
hash1 == hash2 |
||
| next unless File.file?(file) && File.basename(file).match?(FILE_PATTERN) && !file.match?(IGNORED_FILE_PATTERN) | ||
|
|
||
| changelog_entries_seen = 0 | ||
| File.open(file).each do |line| | ||
| changelog_entries_seen += 1 if File.basename(file).eql?('CHANGELOG.md') && line.start_with?('##') | ||
| break if changelog_entries_seen > 2 | ||
| next unless line =~ URL_PATTERN | ||
|
|
||
| url = Regexp.last_match(1).sub(%r{(?:/|\.)$}, '') | ||
| if real_url?(url) | ||
| urls[url] ||= [] | ||
| urls[url] << file | ||
| end | ||
| end | ||
| end | ||
| end | ||
|
|
||
| def verify_url(url) | ||
| puts "Testing '#{url}'..." if DEBUG | ||
| res = HTTParty.get(url, timeout: TIMEOUT) | ||
| if res.success? | ||
| puts ' OK.' if DEBUG | ||
| return | ||
| end | ||
|
|
||
| msg = "HTTP #{res.code}: #{res.message}" | ||
| puts " FAILED. #{msg}" if DEBUG | ||
| msg | ||
| rescue StandardError => e | ||
| msg = "#{e.class}: #{e.message}" | ||
| puts " FAILED. #{msg}" if DEBUG | ||
| msg | ||
| end | ||
| end | ||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Didn't know he wrote memcache-client!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah, I'd be interested in knowing more about the history there. I think that repo is a fork of another fork of the original RubyForge code (which Mike also has commit access to). For whatever reason, it's what's linked to from the Rubygems.org page's "Homepage" link, so I'm using it here.