ruby-link-checker 0.4.0 → 0.5.1

data/spec/ruby-link-checker/typhoeus/hydra/checker_spec.rb

@@ -15,7 +15,7 @@ describe LinkChecker::Typhoeus::Hydra::Checker do
   end
   before :all do
     VCR.configure do |config|
-      config.hook_into :typhoeus
+      config.hook_into :webmock
     end
   end
 
@@ -90,4 +90,91 @@ describe LinkChecker::Typhoeus::Hydra::Checker do
       checker.run
     end
   end
+
+  describe 'delayed retries' do
+    let(:checker) { described_class.new(methods: ['GET'], retries: 1, retry_after: 0.005) }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    it 'processes unrelated requests before waiting for a retry' do
+      completed = []
+      stub_request(:get, 'https://www.example.org/retry').to_return({ status: 503 }, { status: 200 })
+      stub_request(:get, 'https://www.example.org/other').to_return(status: 200)
+      checker.on(:success) { |result| completed << result.uri.path }
+      checker.check('https://www.example.org/retry')
+      checker.check('https://www.example.org/other')
+      checker.run
+      expect(completed).to eq(['/other', '/retry'])
+    end
+  end
+
+  describe 'transport failures' do
+    let(:checker) { described_class.new(methods: ['GET']) }
+
+    around do |example|
+      VCR.turned_off do
+        WebMock.disable!
+        Typhoeus::Config.block_connection = true
+        example.run
+      ensure
+        Typhoeus::Config.block_connection = false
+        WebMock.enable!
+      end
+    end
+
+    after do
+      Typhoeus::Expectation.clear
+    end
+
+    it 'reports connection errors separately from HTTP failures' do
+      response = Typhoeus::Response.new(code: 0, return_code: :couldnt_connect)
+      Typhoeus.stub(nil).and_return(response)
+      checker.check('https://www.example.org/')
+      checker.run
+      expect(checker.results[:failure]).to be_empty
+      expect(checker.results[:error].first.error_type).to eq(:connection)
+    end
+  end
+
+  describe 'in-flight deduplication' do
+    let(:checker) { described_class.new(methods: ['GET'], deduplicate: true) }
+    let(:url) { 'https://www.example.org/' }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    it 'fans out events to duplicates queued before completion' do
+      metadata = []
+      stub_request(:any, url).to_return(status: 200)
+      checker.on(:result) { |result| metadata << result.options }
+      checker.check(url, location: 'one.html')
+      checker.check(url, location: 'two.html')
+      checker.run
+      expect(a_request(:any, url)).to have_been_made.once
+      expect(metadata).to eq([{ location: 'one.html' }, { location: 'two.html' }])
+    end
+
+    it 'replays events when a subscriber joins during a result callback' do
+      stub_request(:any, url).to_return(status: 200)
+      checker.on(:result) { |result| checker.check(url, location: 'two.html') if result.options[:location] == 'one.html' }
+      checker.check(url, location: 'one.html')
+      checker.run
+      expect(checker.results[:success].map(&:options)).to eq([{ location: 'one.html' }, { location: 'two.html' }])
+      expect(a_request(:any, url)).to have_been_made.once
+    end
+
+    it 'does not memoize a response if settings changed while queued' do
+      stub_request(:any, url).to_return(status: 200)
+      checker.check(url)
+      checker.headers = { 'Accept' => 'text/html' }
+      checker.run
+      checker.headers = {}
+      checker.check(url)
+      checker.run
+      expect(a_request(:any, url)).to have_been_made.twice
+    end
+  end
 end

data/spec/support/link_checker.rb

@@ -1,4 +1,244 @@
 shared_context 'a link checker' do
+  context 'with redirects', :aggregate_failures do
+    let(:checker) { described_class.new(methods: ['GET'], max_redirects: 2) }
+    let(:url) { 'https://www.example.org/start' }
+    let(:result) { checker.results.values.flatten.first }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    it 'resolves each relative redirect against the current URL' do
+      stub_request(:any, url).to_return(status: 302, headers: { 'Location' => '/nested/next' })
+      stub_request(:any, 'https://www.example.org/nested/next').to_return(status: 302, headers: { 'Location' => 'final' })
+      stub_request(:any, 'https://www.example.org/nested/final').to_return(status: 200)
+      checker.check(url, location: 'page.html')
+      checker.run if checker.respond_to?(:run)
+      expect(result.uri.to_s).to eq(url)
+      expect(result.result_uri.to_s).to eq('https://www.example.org/nested/final')
+      expect(result.options).to eq(location: 'page.html')
+    end
+
+    it 'stops redirect chains at the configured limit' do
+      stub_request(:any, url).to_return(status: 302, headers: { 'Location' => '/one' })
+      stub_request(:any, 'https://www.example.org/one').to_return(status: 302, headers: { 'Location' => '/two' })
+      stub_request(:any, 'https://www.example.org/two').to_return(status: 302, headers: { 'Location' => '/three' })
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(result.error).to be_a(LinkChecker::Errors::RedirectLimitError)
+      expect(result.uri.to_s).to eq(url)
+      expect(result.result_uri.to_s).to eq('https://www.example.org/two')
+      expect(a_request(:any, 'https://www.example.org/three')).not_to have_been_made
+    end
+
+    it 'detects a redirect back to the original URL without requesting it again' do
+      stub_request(:any, url).to_return(status: 302, headers: { 'Location' => '/one' })
+      stub_request(:any, 'https://www.example.org/one').to_return(status: 302, headers: { 'Location' => url })
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:error].first.error).to be_a(LinkChecker::Errors::RedirectLoopError)
+      expect(a_request(:any, url)).to have_been_made.once
+    end
+
+    it 'allows redirects to be disabled' do
+      checker.max_redirects = 0
+      stub_request(:any, url).to_return(status: 302, headers: { 'Location' => '/one' })
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:error].first.error).to be_a(LinkChecker::Errors::RedirectLimitError)
+      expect(a_request(:any, 'https://www.example.org/one')).not_to have_been_made
+    end
+  end
+
+  context 'with transient retries' do
+    let(:checker_options) { {} }
+    let(:checker) { described_class.new({ methods: ['GET'], retries: 2 }.merge(checker_options)) }
+    let(:url) { 'https://www.example.org/retry' }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    it 'retries service errors and preserves metadata' do
+      stub_request(:get, url).to_return({ status: 503 }, { status: 200 })
+      checker.check(url, location: 'page.html')
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:success].first.options).to eq(location: 'page.html')
+      expect(a_request(:get, url)).to have_been_made.twice
+    end
+
+    it 'does not retry permanent HTTP failures' do
+      stub_request(:get, url).to_return(status: 404)
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:failure].first.code).to eq(404)
+      expect(a_request(:get, url)).to have_been_made.once
+    end
+
+    it 'retries timeouts and reports a timeout category' do
+      stub_request(:get, url).to_timeout
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:error].first.error_type).to eq(:timeout)
+      expect(a_request(:get, url)).to have_been_made.times(3)
+    end
+
+    it 'does not retry malformed URLs' do
+      checker.check('\/invalid-url')
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:error].first.error_type).to eq(:other)
+      expect(checker.results[:error].size).to eq(1)
+    end
+
+    context 'with a configured delay and backoff' do
+      let(:checker_options) { { retry_after: 0.005, retry_backoff: 2 } }
+
+      it 'waits before each retry without changing the retry budget' do
+        stub_request(:get, url).to_return({ status: 503 }, { status: 503 }, { status: 200 })
+        started = Process.clock_gettime(Process::CLOCK_MONOTONIC)
+        checker.check(url)
+        checker.run if checker.respond_to?(:run)
+        elapsed = Process.clock_gettime(Process::CLOCK_MONOTONIC) - started
+        expect(elapsed).to be >= 0.015
+        expect(a_request(:get, url)).to have_been_made.times(3)
+        expect(checker.results[:success].size).to eq(1)
+      end
+    end
+
+    context 'with Retry-After enabled' do
+      let(:checker_options) { { respect_retry_after: true, max_retry_delay: 0.005 } }
+
+      it 'honors a bounded server delay' do
+        stub_request(:get, url).to_return({ status: 429, headers: { 'Retry-After' => '5' } }, { status: 200 })
+        started = Process.clock_gettime(Process::CLOCK_MONOTONIC)
+        checker.check(url)
+        checker.run if checker.respond_to?(:run)
+        elapsed = Process.clock_gettime(Process::CLOCK_MONOTONIC) - started
+        expect(elapsed).to be >= 0.005
+        expect(elapsed).to be < 1
+        expect(checker.results[:success].size).to eq(1)
+      end
+
+      it 'accepts HTTP dates' do
+        stub_request(:get, url).to_return(
+          { status: 503, headers: { 'Retry-After' => (Time.now + 60).httpdate } }, { status: 200 }
+        )
+        checker.check(url)
+        checker.run if checker.respond_to?(:run)
+        expect(checker.results[:success].size).to eq(1)
+      end
+
+      it 'ignores invalid server delays' do
+        stub_request(:get, url).to_return({ status: 503, headers: { 'Retry-After' => 'invalid' } }, { status: 200 })
+        checker.check(url)
+        checker.run if checker.respond_to?(:run)
+        expect(checker.results[:success].size).to eq(1)
+      end
+    end
+  end
+
+  context 'with deduplication' do
+    let(:checker) { described_class.new(methods: ['GET'], deduplicate: true) }
+    let(:url) { 'https://www.example.org/' }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    it 'checks once and preserves metadata for every submission' do
+      stub_request(:any, url).to_return(status: 200)
+      checker.check(url, location: 'one.html')
+      checker.check(url, location: 'two.html')
+      checker.run if checker.respond_to?(:run)
+      expect(a_request(:any, url)).to have_been_made.once
+      expect(checker.results[:success].map(&:options)).to eq([{ location: 'one.html' }, { location: 'two.html' }])
+      expect(checker.results[:success].map(&:uri)).to eq([URI(url), URI(url)])
+    end
+
+    it 'rechecks failures instead of retaining them' do
+      stub_request(:any, url).to_return({ status: 404 }, { status: 200 })
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(a_request(:any, url)).to have_been_made.twice
+      expect(checker.results[:failure].size).to eq(1)
+      expect(checker.results[:success].size).to eq(1)
+    end
+
+    it 'keeps query values distinct' do
+      stub_request(:any, "#{url}?id=1").to_return(status: 200)
+      stub_request(:any, "#{url}?id=2").to_return(status: 404)
+      checker.check("#{url}?id=1")
+      checker.check("#{url}?id=2")
+      checker.run if checker.respond_to?(:run)
+      expect(checker.results[:success].first.uri.query).to eq('id=1')
+      expect(checker.results[:failure].first.uri.query).to eq('id=2')
+    end
+
+    it 'includes headers in request identity' do
+      stub_request(:any, url).to_return(status: 200)
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      checker.headers = { 'Accept' => 'text/html' }
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      expect(a_request(:any, url)).to have_been_made.twice
+    end
+
+    context 'without result collection' do
+      let(:checker) { described_class.new(methods: ['GET'], deduplicate: true, results: false) }
+
+      it 'still delivers results to every subscriber' do
+        completed = []
+        stub_request(:any, url).to_return(status: 200)
+        checker.on(:success) { |result| completed << result.options }
+        checker.check(url, location: 'one.html')
+        checker.check(url, location: 'two.html')
+        checker.run if checker.respond_to?(:run)
+        expect(completed).to eq([{ location: 'one.html' }, { location: 'two.html' }])
+        expect(checker.results).to be_nil
+      end
+    end
+  end
+
+  context 'with persistent caching' do
+    let(:directory) { Dir.mktmpdir('link-checker-spec-') }
+    let(:cache_options) { { file: File.join(directory, 'cache.json'), ttl: 60 } }
+    let(:checker) { described_class.new(methods: ['GET'], cache: cache_options) }
+    let(:next_checker) { described_class.new(methods: ['GET'], cache: cache_options) }
+    let(:url) { 'https://www.example.org/' }
+
+    around do |example|
+      VCR.turned_off { example.run }
+    end
+
+    after do
+      FileUtils.remove_entry(directory)
+    end
+
+    it 'reuses a success across checker instances with fresh metadata' do
+      stub_request(:any, url).to_return(status: 200)
+      checker.check(url, location: 'one.html')
+      checker.run if checker.respond_to?(:run)
+      next_checker.check(url, location: 'two.html')
+      next_checker.run if next_checker.respond_to?(:run)
+      expect(next_checker.results[:success].first.cached?).to be true
+      expect(next_checker.results[:success].first.options).to eq(location: 'two.html')
+      expect(a_request(:any, url)).to have_been_made.once
+    end
+
+    it 'does not persist failed checks' do
+      stub_request(:any, url).to_return({ status: 404 }, { status: 200 })
+      checker.check(url)
+      checker.run if checker.respond_to?(:run)
+      next_checker.check(url)
+      next_checker.run if next_checker.respond_to?(:run)
+      expect(next_checker.results[:success].first.cached?).to be false
+      expect(a_request(:any, url)).to have_been_made.twice
+    end
+  end
+
   context 'default' do
     subject do
       described_class.new
@@ -176,7 +416,7 @@ shared_context 'a link checker' do
 
             context 'with 1 retry' do
               subject do
-                described_class.new(methods: ['GET'], retries: 1)
+                described_class.new(methods: ['GET'], retries: 1, retry_status_codes: nil)
               end
 
               it 'fails' do
@@ -192,7 +432,7 @@ shared_context 'a link checker' do
 
             context 'with 2 retries' do
               subject do
-                described_class.new(methods: ['GET'], retries: 2)
+                described_class.new(methods: ['GET'], retries: 2, retry_status_codes: nil)
               end
 
               it 'fails' do
@@ -240,7 +480,7 @@ shared_context 'a link checker' do
               expect(result.error?).to be true
               expect(result.error).to be_a LinkChecker::Errors::RedirectLoopError
               expect(result.redirect?).to be false
-              expect(subject).to have_received(:called!).with(:redirect, anything).twice
+              expect(subject).to have_received(:called!).with(:redirect, anything).once
               expect(subject).to have_received(:called!).with(:error, result)
               expect(subject).not_to have_received(:called!).with(:failure, result)
               expect(subject).not_to have_received(:called!).with(:success, result)

metadata

@@ -1,13 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-link-checker
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.5.1
 platform: ruby
 authors:
 - Daniel Doubrovkine
+autorequire:
 bindir: bin
 cert_chain: []
-date: 1980-01-02 00:00:00.000000000 Z
+date: 2026-11-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: async-http
@@ -37,16 +38,27 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+description:
 email: dblock@dblock.org
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".github/workflows/coverage.yml"
+- ".github/workflows/danger-comment.yml"
+- ".github/workflows/danger.yml"
+- ".github/workflows/lint.yml"
+- ".github/workflows/release.yml"
+- ".github/workflows/test.yml"
+- ".gitignore"
+- ".rspec"
+- ".rubocop.yml"
+- ".rubocop_todo.yml"
+- AGENTS.md
 - CHANGELOG.md
 - CONTRIBUTING.md
 - Dangerfile
 - Gemfile
-- Gemfile.lock
 - LICENSE.md
 - README.md
 - RELEASING.md
@@ -54,46 +66,19 @@ files:
 - benchmarks/check.rb
 - benchmarks/data/opensearch.org/links.txt
 - benchmarks/data/opensearch.org/small.txt
-- coverage/assets/0.13.2/DataTables-1.10.20/images/sort_asc.png
-- coverage/assets/0.13.2/DataTables-1.10.20/images/sort_asc_disabled.png
-- coverage/assets/0.13.2/DataTables-1.10.20/images/sort_both.png
-- coverage/assets/0.13.2/DataTables-1.10.20/images/sort_desc.png
-- coverage/assets/0.13.2/DataTables-1.10.20/images/sort_desc_disabled.png
-- coverage/assets/0.13.2/application.css
-- coverage/assets/0.13.2/application.js
-- coverage/assets/0.13.2/colorbox/border.png
-- coverage/assets/0.13.2/colorbox/controls.png
-- coverage/assets/0.13.2/colorbox/loading.gif
-- coverage/assets/0.13.2/colorbox/loading_background.png
-- coverage/assets/0.13.2/favicon_green.png
-- coverage/assets/0.13.2/favicon_red.png
-- coverage/assets/0.13.2/favicon_yellow.png
-- coverage/assets/0.13.2/images/ui-bg_flat_0_aaaaaa_40x100.png
-- coverage/assets/0.13.2/images/ui-bg_flat_75_ffffff_40x100.png
-- coverage/assets/0.13.2/images/ui-bg_glass_55_fbf9ee_1x400.png
-- coverage/assets/0.13.2/images/ui-bg_glass_65_ffffff_1x400.png
-- coverage/assets/0.13.2/images/ui-bg_glass_75_dadada_1x400.png
-- coverage/assets/0.13.2/images/ui-bg_glass_75_e6e6e6_1x400.png
-- coverage/assets/0.13.2/images/ui-bg_glass_95_fef1ec_1x400.png
-- coverage/assets/0.13.2/images/ui-bg_highlight-soft_75_cccccc_1x100.png
-- coverage/assets/0.13.2/images/ui-icons_222222_256x240.png
-- coverage/assets/0.13.2/images/ui-icons_2e83ff_256x240.png
-- coverage/assets/0.13.2/images/ui-icons_454545_256x240.png
-- coverage/assets/0.13.2/images/ui-icons_888888_256x240.png
-- coverage/assets/0.13.2/images/ui-icons_cd0a0a_256x240.png
-- coverage/assets/0.13.2/loading.gif
-- coverage/assets/0.13.2/magnify.png
-- coverage/index.html
 - lib/ruby-link-checker.rb
 - lib/ruby-link-checker/async/http.rb
 - lib/ruby-link-checker/async/http/checker.rb
 - lib/ruby-link-checker/async/http/config.rb
 - lib/ruby-link-checker/async/http/result.rb
+- lib/ruby-link-checker/cache.rb
 - lib/ruby-link-checker/callbacks.rb
 - lib/ruby-link-checker/checker.rb
 - lib/ruby-link-checker/config.rb
 - lib/ruby-link-checker/errors.rb
 - lib/ruby-link-checker/errors/base_error.rb
+- lib/ruby-link-checker/errors/connection_error.rb
+- lib/ruby-link-checker/errors/redirect_limit_error.rb
 - lib/ruby-link-checker/errors/redirect_loop_error.rb
 - lib/ruby-link-checker/logger.rb
 - lib/ruby-link-checker/net/http.rb
@@ -144,6 +129,7 @@ licenses:
 - MIT
 metadata:
   rubygems_mfa_required: 'true'
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -158,7 +144,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: 1.3.6
 requirements: []
-rubygems_version: 3.6.9
+rubygems_version: 3.5.22
+signing_key:
 specification_version: 4
 summary: Fast ruby link checker.
 test_files: []