parklife 0.8.1 → 0.9.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: b76fbfd9c4d1a56c6b2498e1e5470e5ac43bd06fc68ef5e944ed2f2c5b0ab81c
-  data.tar.gz: '0808d1e25b27429f521002f7e28d41c254dcee31e2ea4b76a639f1bbd61d15b0'
+  metadata.gz: 98577bf55e3f709d741dfe6df388c5d61f6c93d12d53b88983ad10dbf0552fbd
+  data.tar.gz: 6dbe3f4d37e80846b461a6b29981b56cd15050bdce4f9ee517385ac7be08d5a3
 SHA512:
-  metadata.gz: 950ab97e849c11e69181093a0dd110f51fcafbc22ab869dbff07dd1ce79f69cb78dfbd9ffe5311b0b1d34fbd770f531042d5a646f349ab387b00b44fe6c0cc97
-  data.tar.gz: 23d920c50159706d18f20c958abb98f59e320b841a7ab00cde022c57ef8fb9b4b9e14486965ae55f538d0820c378f4c0b41c3bc30a37d8eee5250101ab5cee5a
+  metadata.gz: d2b8b5540f43681e1b0ea92fff99fc14e5d25bde63b1e06909e648522be0baba8577be1f8981de0c92238b030791933cdad8a42b2d4cc5827a89937633393f07
+  data.tar.gz: c66a33ad08986f3e94fc0477eefdbab1c98ca8ea7ff87e87d9a26ea01f1d0f09fd152230632b462dbea42d54ddaf18bf5624f6e15710f136be22149b2fa9cd0f

data/.github/dependabot.yml

@@ -0,0 +1,11 @@
+version: 2
+
+updates:
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+    groups:
+      github-actions:
+        patterns:
+          - "*"

data/.github/workflows/examples.yml

@@ -7,13 +7,13 @@ jobs:
     runs-on: ubuntu-latest
     name: Rack example
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v6
     - uses: ruby/setup-ruby@v1
       with:
         bundler-cache: true
         ruby-version: '2.7'
         working-directory: examples/rack
-    - run: bundle exec parklife build
+    - run: bundle exec parklife build --reporter log
       working-directory: examples/rack
     - run: test -f build/index.html
       working-directory: examples/rack
@@ -22,13 +22,13 @@ jobs:
     runs-on: ubuntu-latest
     name: Roda example
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v6
     - uses: ruby/setup-ruby@v1
       with:
         bundler-cache: true
         ruby-version: '3.2'
         working-directory: examples/roda
-    - run: bundle exec parklife build
+    - run: bundle exec parklife build --reporter log
       working-directory: examples/roda
     - run: test -f build/index.html
       working-directory: examples/roda

data/.github/workflows/tests.yml

@@ -17,7 +17,7 @@ jobs:
           - '3.4'
     name: RSpec (Ruby ${{ matrix.ruby }})
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v6
     - uses: ruby/setup-ruby@v1
       with:
         bundler-cache: true
@@ -28,7 +28,7 @@ jobs:
     runs-on: ubuntu-latest
     name: RuboCop
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v6
     - uses: ruby/setup-ruby@v1
       with:
         bundler-cache: true

data/.rubocop.yml

@@ -10,9 +10,6 @@ AllCops:
 Layout:
   Enabled: false
 
-Layout/EmptyLineAfterMagicComment:
-  Enabled: true
-
 Lint/AmbiguousBlockAssociation:
   Enabled: false
 

data/CHANGELOG.md

@@ -1,4 +1,24 @@
-## Unreleased
+<!-- ## Unreleased -->
+
+## Version 0.9.0 - 2026-05-16
+
+- Introduce the log reporter and add colour to the output by default (or pass `--no-colour` to disable). Pass `--reporter log` to the build) Supported reporters are `log` (one line per visited route), `null` (only errors), and `progress` (dots). <https://github.com/benpickles/parklife/pull/138>
+
+- Add inter-build caching via support HTTP ETags. <https://github.com/benpickles/parklife/pull/138>
+
+  Parklife now saves build metadata (in the file `BUILD_DIR/.parklife/build.yml`) which is used to support HTTP caching via ETags (it can be skipped by setting the config `skip_build_meta = false` or using `--skip-build-meta` via the CLI). ETag generation and the actual cache hit/miss responsibility rests on the app itself but hopefully your framework can help with this (certainly Rails has lots of HTTP caching helpers). Tell Parklife to use a previous build as a cache source at build time with `parklife build --cache-dir build`.
+
+- Introduce response handlers and make it possible for them to be customised without having to resort to monkeypatching. <https://github.com/benpickles/parklife/pull/136>
+
+- Only store a text/html response as an .html file <https://github.com/benpickles/parklife/pull/136>
+
+  Prior to this change all responses were expected to be HTML unless the request path included a file extension which led to the following issue:
+
+  The path `/dry-types/v1.8` would be detected as having a file extension (`.8`) and create the _file_ `dry-types/v1.8`, then the path `/dry-types/v1.8/something-else` would attempt to create the file `dry-types/v1.8/something-else.html` but encounter a `Errno::EEXIST` when it attempted to create the _directory_ `dry-types/v1.8`.
+
+  With this change a response is only saved as HTML when its content type is `text/html`. This fixes the above case and should mean that files in the final build and how they're treated by a static server is better aligned with how they're used in the dynamic development environment.
+
+- Fix including a 404 response in the build when `on_404` was `:skip` or `:warn`. <https://github.com/benpickles/parklife/pull/135>
 
 ## Version 0.8.1 - 2025-12-21
 

data/README.md

@@ -116,10 +116,13 @@ Parklife.application.config.build_dir = 'my/build/dir'
 
 ### Handling a 404
 
-By default if Parklife encounters a 404 response when fetching a route it will raise an exception (the `:error` setting) and stop the build. Other values are:
+By default if Parklife encounters a 404 response when fetching a route it will raise an exception which stops the build (the `:error` setting).
 
-- `:warn` - output a message to `stderr`, save the response, and continue processing.
-- `:skip` - silently ignore and not save the response, and continue processing.
+Possible values are:
+
+- `:error` (default) - raise an exception which stops the build.
+- `:skip` - do not save the response, continue processing.
+- `:warn` - output a message to `stderr`, do not save the response, continue processing.
 
 ```ruby
 Parklife.application.config.on_404 = :warn

data/lib/parklife/application.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require 'fileutils'
+require 'parklife/build'
 require 'parklife/config'
 require 'parklife/crawler'
 require 'parklife/errors'
@@ -26,14 +27,10 @@ module Parklife
     end
 
     def build
-      raise BuildDirNotDefinedError if config.build_dir.nil?
       raise RackAppNotDefinedError if config.app.nil?
 
-      if Dir.exist?(config.build_dir)
-        FileUtils.rm_rf(Pathname.new(config.build_dir).children)
-      else
-        Dir.mkdir(config.build_dir)
-      end
+      prepare_cache_dir if config.cache_dir
+      prepare_build_dir
 
       @before_build_callbacks.each do |callback|
         callback.call(self)
@@ -51,7 +48,11 @@ module Parklife
     end
 
     def crawler
-      @crawler ||= Crawler.new(config, @route_set)
+      @crawler ||= Crawler.new(
+        config,
+        @route_set,
+        config.cache_dir ? Build.from_dir(config.cache_dir) : nil
+      )
     end
 
     def load_Parkfile(path)
@@ -66,5 +67,37 @@ module Parklife
         @route_set
       end
     end
+
+    private
+      def prepare_build_dir
+        if config.build_dir.directory?
+          FileUtils.rm_rf(config.build_dir.children)
+        else
+          config.build_dir.mkdir
+        end
+      end
+
+      def prepare_cache_dir
+        # Nothing to do unless the previous build is being used as a cache.
+        return unless config.cache_dir.expand_path == config.build_dir.expand_path
+
+        if config.build_dir.exist?
+          config.cache_dir = Config::CACHE_TMPDIR
+
+          if config.cache_dir.exist?
+            config.cache_dir.rmtree
+          else
+            config.cache_dir.dirname.mkpath
+          end
+
+          # Move the existing previous build to the tmp location to clear the
+          # way for a fresh new build.
+          config.build_dir.rename(config.cache_dir)
+        else
+          # The build/cache directories are set to the same thing but don't
+          # exist so there is no cache.
+          config.cache_dir = nil
+        end
+      end
   end
 end

data/lib/parklife/browser.rb

@@ -11,24 +11,23 @@ module Parklife
       @app = app
       @base = base
       @session = Rack::Test::Session.new(app)
-      set_env
+      @env = {
+        'HTTP_HOST' => Utils.host_with_port(base),
+        'HTTPS' => base.scheme == 'https' ? 'on' : 'off',
+        script_name: base.path.chomp('/'),
+      }
     end
 
-    def get(path)
-      session.get(path, nil, env)
+    def get(path, headers: nil)
+      session.get(
+        path,
+        nil,
+        headers ? headers.merge(env) : env
+      )
     end
 
     def uri_for(path)
       base.dup.tap { |uri| uri.path = path }
     end
-
-    private
-      def set_env
-        @env = {
-          'HTTP_HOST' => Utils.host_with_port(base),
-          'HTTPS' => base.scheme == 'https' ? 'on' : 'off',
-          script_name: base.path.chomp('/'),
-        }
-      end
   end
 end

data/lib/parklife/build.rb

@@ -0,0 +1,109 @@
+# frozen_string_literal: true
+require 'fileutils'
+require 'yaml'
+
+module Parklife
+  class Build
+    META_PATH = File.join('.parklife', 'build.yml')
+
+    def self.from_dir(dir)
+      return unless dir.exist?
+      path = dir.join(META_PATH)
+      return unless path.exist?
+      data = YAML.safe_load(path.read)
+
+      build = new(dir, nested_index: data.dig('config', 'nested_index'))
+      build.paths.merge!(data['paths'])
+      build
+    end
+
+    def self.path_for(path, media_type: nil, nested_index:)
+      # Remove leading/trailing slashes.
+      path = path.gsub(/^\/|\/$/, '')
+
+      # The root is always expected to be an HTML page regardless of the
+      # response's content type.
+      return 'index.html' if path.empty?
+
+      text_html = media_type.nil? || media_type == 'text/html'
+
+      # Store a text/html response in an .html file.
+      if text_html && !path.end_with?('.html')
+        if nested_index
+          path << '/index.html'
+        else
+          path << '.html'
+        end
+      end
+
+      path
+    end
+
+    attr_reader :dir, :nested_index, :paths
+
+    def initialize(dir, nested_index:)
+      @dir = dir
+      @nested_index = nested_index
+      @paths = {}
+    end
+
+    def add(route, response)
+      build_path = self.build_path(route, response)
+      write(build_path, response.body)
+      add_path_meta(route, response, build_path)
+    end
+
+    def build_path(route, response)
+      self.class.path_for(
+        route.path,
+        media_type: response.media_type,
+        nested_index: nested_index,
+      )
+    end
+
+    def copy(src, route, response)
+      build_path = self.build_path(route, response)
+
+      dest = dir.join(build_path)
+      dest.dirname.mkpath
+      FileUtils.cp(src, dest)
+
+      add_path_meta(route, response, build_path)
+    end
+
+    def etag(path)
+      paths.dig(path, 'etag')
+    end
+
+    def get(route, response)
+      dir.join(build_path(route, response))
+    end
+
+    def to_yaml
+      YAML.dump({
+        'config' => {
+          'nested_index' => nested_index,
+        },
+        'paths' => paths,
+      })
+    end
+
+    def write(path, content)
+      file = dir.join(path)
+      file.dirname.mkpath
+      file.write(content, mode: 'wb')
+    end
+
+    def write_meta
+      write(META_PATH, to_yaml)
+    end
+
+    private
+      def add_path_meta(route, response, build_path)
+        paths[route.path] = {
+          'build_path' => build_path,
+          'etag' => response['Etag'],
+        }.compact
+      end
+  end
+end

data/lib/parklife/cli.rb

@@ -11,23 +11,31 @@ module Parklife
     class_option :base, desc: 'Override config.base configured in the Parkfile'
 
     desc 'build', 'Create a production build'
+    option :cache_dir, desc: 'Path to an existing build directory (which must include build meta)', type: :string
+    option :no_colour, aliases: '--no-color', desc: "Don't include colours in terminal output", type: :boolean
+    option :reporter, desc: 'Output formatter: log (one line per visited route), null (only errors), progress (dots)'
+    option :skip_build_meta, desc: 'Do not include Parklife build metadata', type: :boolean
     def build
+      application.config.cache_dir = options[:cache_dir] if options.key?(:cache_dir)
+      application.config.no_colour = options[:no_colour] if options.key?(:no_colour)
+      application.config.reporter = options[:reporter] if options.key?(:reporter)
+      application.config.skip_build_meta = options[:skip_build_meta] if options.key?(:skip_build_meta)
       application.build
     end
 
     desc 'config', 'Output the full Parklife config'
     def config
-      reporter = application.config.reporter
-
       shell.print_table([
         ['app', application.config.app.inspect],
         ['base', application.config.base.to_s],
         ['build_dir', application.config.build_dir],
+        ['cache_dir', application.config.cache_dir],
         ['nested_index', application.config.nested_index],
+        ['no_colour', application.config.no_colour],
         ['on_404', application.config.on_404.inspect],
         ['parklife-rails', defined?(::Parklife::Rails) ? 'enabled' : '-'],
         ['parklife-sinatra', defined?(::Parklife::Sinatra) ? 'enabled' : '-'],
-        ['reporter', reporter == $stdout ? '$stdout' : reporter],
+        ['reporter', application.config.reporter],
       ])
     end
 
@@ -65,8 +73,8 @@ module Parklife
     private
       def application
         @application ||= Parklife.application.tap { |app|
-          # Default output to stdout (can be overridden in the Parkfile).
-          app.config.reporter = $stdout
+          # Default to dots for progress (can be overridden in the Parkfile).
+          app.config.reporter = 'progress'
 
           # Reach inside the consuming app's directory to apply its Parklife
           # config.

data/lib/parklife/config.rb

@@ -1,22 +1,31 @@
 # frozen_string_literal: true
-
+require 'pathname'
 require 'stringio'
 require 'uri'
+require_relative 'logger'
+require_relative 'reporter/log'
+require_relative 'reporter/null'
+require_relative 'reporter/progress'
 
 module Parklife
   class Config
+    CACHE_TMPDIR = 'tmp/parklife/cache'
     DEFAULT_HOST = 'example.com'
     DEFAULT_SCHEME = 'http'
 
-    attr_accessor :app, :build_dir, :nested_index, :on_404, :reporter
-    attr_reader :base
+    attr_accessor :app, :logger, :nested_index, :on_404, :skip_build_meta
+    attr_reader :base, :build_dir, :cache_dir, :no_colour, :reporter
 
     def initialize
       self.base = nil
       self.build_dir = 'build'
+      self.cache_dir = nil
+      self.logger = Logger.new
       self.nested_index = true
+      self.no_colour = false
       self.on_404 = :error
-      self.reporter = StringIO.new
+      self.reporter = 'null'
+      self.skip_build_meta = false
     end
 
     def base=(value)
@@ -25,5 +34,28 @@ module Parklife
       uri.scheme ||= DEFAULT_SCHEME
       @base = uri
     end
+
+    def build_dir=(value)
+      @build_dir = Pathname.new(value)
+    end
+
+    def cache_dir=(value)
+      @cache_dir = value ? Pathname.new(value) : nil
+    end
+
+    def no_colour=(value)
+      @no_colour = logger.no_colour = value
+    end
+
+    def reporter=(value)
+      @reporter = case value
+      when 'log'
+        Reporter::Log.new(logger)
+      when 'progress'
+        Reporter::Progress.new(logger)
+      else
+        Reporter::Null.new(logger)
+      end
+    end
   end
 end

data/lib/parklife/crawler.rb

@@ -1,98 +1,97 @@
 # frozen_string_literal: true
-
-require 'parklife/browser'
-require 'parklife/route'
-require 'parklife/utils'
 require 'set'
+require_relative 'browser'
+require_relative 'build'
+require_relative 'responder/not_found'
+require_relative 'responder/not_modified'
+require_relative 'responder/ok'
+require_relative 'responder/redirect'
+require_relative 'responder/unknown'
+require_relative 'route'
+require_relative 'utils'
 
 module Parklife
   class Crawler
-    attr_reader :browser, :config, :route_set
+    RESPONDERS = {
+      200 => Responder::Ok,
+      301 => Responder::Redirect,
+      302 => Responder::Redirect,
+      304 => Responder::NotModified,
+      404 => Responder::NotFound,
+    }
+
+    attr_reader :browser, :build, :cache, :config, :routes, :visited
 
-    def initialize(config, route_set)
+    def initialize(config, routes, cache)
       @config = config
-      @route_set = route_set
+      @routes = routes.to_a
+      @cache = cache
       @browser = Browser.new(config.app, config.base)
+      @build = Build.new(config.build_dir, nested_index: config.nested_index)
+      @visited = Set.new
+      @responder_for_status = {}
     end
 
-    def get(path)
-      browser.get(path)
+    def crawl(html)
+      Utils.scan_for_links(html) do |path|
+        # If the app is mounted at a subdirectory then it responds to paths that
+        # *exclude* the subdirectory and generates links that *include* the
+        # subdirectory (so if the app is mounted at "/foo" and serving "/bar"
+        # then the full path would be "/foo/bar" and a generated link would
+        # include the mount path like "/foo/link").
+        #
+        # Anyway, this mount path prefix must be trimmed from link paths so that
+        # correct app routes are created.
+        baseless_path = path.delete_prefix(config.base.path)
+        new_route = Route.new(baseless_path, crawl: true)
+
+        next if visited?(new_route)
+
+        routes << new_route
+      end
     end
 
-    def start
-      @routes = route_set.to_a
-      @visited = Set.new
-
-      while (route = @routes.shift)
-        processed = process_route(route)
-        config.reporter.print('.') if processed
+    def get(path)
+      headers = if (etag = cache&.etag(path))
+        { 'HTTP_IF_NONE_MATCH' => etag }
+      else
+        nil
       end
 
-      config.reporter.puts
+      browser.get(path, headers: headers)
     end
 
-    private
-      def process_route(route)
-        already_processed = if route.crawl
-          # No need to re-process an already-crawled route (but do re-process
-          # a route that has been visited but not crawled).
-          @visited.include?(route)
-        else
-          # This route isn't being crawled so there's no need to re-process
-          # it if it has already been visited or crawled.
-          crawled_route = Route.new(route.path, crawl: true)
-          @visited.include?(route) || @visited.include?(crawled_route)
-        end
-
-        return false if already_processed
+    def responder_for_status(status)
+      @responder_for_status[status] ||= RESPONDERS
+        .fetch(status, Responder::Unknown)
+        .new(self)
+    end
 
+    def start
+      while (route = routes.shift)
+        next if visited?(route)
         response = get(route.path)
-
-        case response.status
-        when 200
-          # Continue processing the route.
-        when 301, 302
-          raise HTTPRedirectError.new(
-            response.status,
-            browser.uri_for(route.path),
-            response.headers['location']
-          )
-        when 404
-          case config.on_404
-          when :warn
-            $stderr.puts HTTPError.new(404, route.path).message
-          when :skip
-            return false
-          else
-            raise HTTPError.new(404, route.path)
-          end
-        else
-          raise HTTPError.new(response.status, route.path)
-        end
-
-        Utils.save_page(route.path, response.body, config)
-
         @visited << route
+        config.reporter.visit(route, response)
+        responder_for_status(response.status).call(route, response)
+      end
 
-        if route.crawl
-          Utils.scan_for_links(response.body) do |path|
-            # When an app is mounted at a path it responds to URLs that must
-            # exclude the mount path but it generates links that include it (if
-            # it is correctly configured). This prefix must therefore be
-            # stripped from links discovered via crawling.
-            baseless_path = path.delete_prefix(config.base.path)
-
-            route = Route.new(baseless_path, crawl: true)
-
-            # Don't revisit the route if it has already been visited with
-            # crawl=true but do revisit if it wasn't crawled.
-            next if @visited.include?(route)
-
-            @routes << route
-          end
-        end
+      config.reporter.finish
+    ensure
+      build.write_meta unless config.skip_build_meta
+    end
 
-        true
+    def visited?(route)
+      if route.crawl
+        # A crawl=true route is only counted as visited when it has already been
+        # crawled, if it's been visited by a non-crawl route then it must be
+        # visited again so it can be crawled.
+        @visited.include?(route)
+      else
+        # A crawl=false route is counted as visited whether it was previously
+        # visited with either a crawl or non-crawl route.
+        @visited.include?(route) || @visited.include?(route.with_crawl)
       end
+    end
   end
 end

data/lib/parklife/errors.rb

@@ -2,7 +2,6 @@
 
 module Parklife
   Error = Class.new(StandardError)
-  BuildDirNotDefinedError = Class.new(Error)
   RackAppNotDefinedError = Class.new(Error)
 
   class HTTPError < Error

data/lib/parklife/logger.rb

@@ -0,0 +1,29 @@
+# frozen_string_literal: true
+require 'forwardable'
+require 'thor/shell/color'
+
+module Parklife
+  class Logger
+    extend Forwardable
+
+    attr_accessor :no_colour
+    attr_reader :stderr, :stdout
+
+    def initialize(stdout = $stdout, stderr = $stderr, no_colour: false)
+      @stdout = stdout
+      @stderr = stderr
+      @no_colour = no_colour
+      @thor_color = Thor::Shell::Color.new
+    end
+
+    def_delegators :@stdout, :print, :puts
+
+    def colour(string, *colours)
+      no_colour ? string : @thor_color.set_color(string, *colours)
+    end
+
+    def warn(*message)
+      stderr.puts(colour(*message, :on_red))
+    end
+  end
+end

data/lib/parklife/reporter/base.rb

@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+require 'forwardable'
+
+module Parklife
+  module Reporter
+    class Base
+      STATUS_COLOUR = {
+        200 => :green,
+        304 => :blue,
+        404 => :yellow,
+      }
+
+      extend Forwardable
+
+      attr_reader :logger
+
+      def initialize(logger)
+        @logger = logger
+      end
+
+      def_delegators :@logger, :colour, :print, :puts
+
+      def finish
+      end
+
+      def visit(_route, _response)
+      end
+    end
+  end
+end

data/lib/parklife/reporter/log.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Reporter
+    class Log < Base
+      def visit(route, response)
+        status = response.status
+        puts "#{colour(status, *STATUS_COLOUR[status])} #{route.path}"
+      end
+    end
+  end
+end

data/lib/parklife/reporter/null.rb

@@ -0,0 +1,9 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Reporter
+    class Null < Base
+    end
+  end
+end

data/lib/parklife/reporter/progress.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Reporter
+    class Progress < Base
+      def finish
+        puts
+      end
+
+      def visit(_route, response)
+        print colour('.', *STATUS_COLOUR[response.status])
+      end
+    end
+  end
+end

data/lib/parklife/responder/base.rb

@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+module Parklife
+  module Responder
+    class Base
+      attr_reader :crawler
+
+      def initialize(crawler)
+        @crawler = crawler
+      end
+
+      def call(route, response)
+      end
+    end
+  end
+end

data/lib/parklife/responder/not_found.rb

@@ -0,0 +1,21 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Responder
+    class NotFound < Base
+      def call(route, response)
+        case crawler.config.on_404
+        when :skip
+          # No-op.
+        when :warn
+          crawler.config.logger.warn(
+            HTTPError.new(response.status, route.path).message
+          )
+        else
+          raise HTTPError.new(response.status, route.path)
+        end
+      end
+    end
+  end
+end

data/lib/parklife/responder/not_modified.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Responder
+    class NotModified < Base
+      def call(route, response)
+        pathname = crawler.cache&.get(route, response)
+
+        return unless pathname&.exist?
+
+        crawler.build.copy(pathname, route, response)
+        crawler.crawl(pathname.read) if route.crawl
+      end
+    end
+  end
+end

data/lib/parklife/responder/ok.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Responder
+    class Ok < Base
+      def call(route, response)
+        crawler.build.add(route, response)
+        crawler.crawl(response.body) if route.crawl
+      end
+    end
+  end
+end

data/lib/parklife/responder/redirect.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Responder
+    class Redirect < Base
+      def call(route, response)
+        raise HTTPRedirectError.new(
+          response.status,
+          crawler.browser.uri_for(route.path),
+          response.headers['location']
+        )
+      end
+    end
+  end
+end

data/lib/parklife/responder/unknown.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+require_relative 'base'
+
+module Parklife
+  module Responder
+    class Unknown < Base
+      def call(route, response)
+        raise HTTPError.new(response.status, route.path)
+      end
+    end
+  end
+end

data/lib/parklife/route.rb

@@ -21,5 +21,9 @@ module Parklife
     def inspect
       %(<#{self.class.name} path="#{path}" crawl="#{crawl}">)
     end
+
+    def with_crawl
+      crawl ? self : self.class.new(path, crawl: true)
+    end
   end
 end

data/lib/parklife/utils.rb

@@ -7,36 +7,11 @@ module Parklife
   module Utils
     extend self
 
-    def build_path_for(path, index: true)
-      path = path.gsub(/^\/|\/$/, '')
-
-      if File.extname(path).empty?
-        if path.empty?
-          'index.html'
-        elsif index
-          File.join(path, 'index.html')
-        else
-          "#{path}.html"
-        end
-      else
-        path
-      end
-    end
-
     def host_with_port(uri)
       default_port = uri.scheme == 'https' ? 443 : 80
       uri.port == default_port ? uri.host : "#{uri.host}:#{uri.port}"
     end
 
-    def save_page(path, content, config)
-      build_path = File.join(
-        config.build_dir,
-        build_path_for(path, index: config.nested_index)
-      )
-      FileUtils.mkdir_p(File.dirname(build_path))
-      File.write(build_path, content, mode: 'wb')
-    end
-
     def scan_for_links(html)
       doc = Nokogiri::HTML.parse(html)
       doc.css('a[href]').each do |a|

data/lib/parklife/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Parklife
-  VERSION = '0.8.1'
+  VERSION = '0.9.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: parklife
 version: !ruby/object:Gem::Version
-  version: 0.8.1
+  version: 0.9.0
 platform: ruby
 authors:
 - Ben Pickles
@@ -59,6 +59,7 @@ extensions: []
 extra_rdoc_files: []
 files:
 - ".github/FUNDING.yml"
+- ".github/dependabot.yml"
 - ".github/workflows/examples.yml"
 - ".github/workflows/tests.yml"
 - ".gitignore"
@@ -76,10 +77,22 @@ files:
 - lib/parklife.rb
 - lib/parklife/application.rb
 - lib/parklife/browser.rb
+- lib/parklife/build.rb
 - lib/parklife/cli.rb
 - lib/parklife/config.rb
 - lib/parklife/crawler.rb
 - lib/parklife/errors.rb
+- lib/parklife/logger.rb
+- lib/parklife/reporter/base.rb
+- lib/parklife/reporter/log.rb
+- lib/parklife/reporter/null.rb
+- lib/parklife/reporter/progress.rb
+- lib/parklife/responder/base.rb
+- lib/parklife/responder/not_found.rb
+- lib/parklife/responder/not_modified.rb
+- lib/parklife/responder/ok.rb
+- lib/parklife/responder/redirect.rb
+- lib/parklife/responder/unknown.rb
 - lib/parklife/route.rb
 - lib/parklife/route_set.rb
 - lib/parklife/templates/Parkfile.erb