ask-web-fetch 0.5.7 → 0.5.9

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '08275e5b901ca9aaba313cac4cd3f776efb2c16b93aa709645c0ad367a1ff720'
-  data.tar.gz: 0eaba632b36081a0bc3fbf7dd1ff2b463cadb613e44d8a280d0546aeeda40676
+  metadata.gz: d8f3dc79c573c21dc567f86edd621ad84da98437ef512faea0e13b36219670d2
+  data.tar.gz: 694519194d63c9f4066f943ed2fe472f029ea10afd6e423a41440ff7d8a04964
 SHA512:
-  metadata.gz: 3c82f4b9abfb154fb06b9a5feaaf2254ddb6caa1cac99892a034311b4113ab1775114ad7c5aa126db88682c3ecf4cdfbee13786b61a06a00f2f30d6b838a2cdf
-  data.tar.gz: f3d14810e581699b5d2f8717fa329052ad8b81f03611d637d99fde979f1b324d2939ee79217ee812018327596f07ae0659c987b2e42f197df60af4bff30c8a7b
+  metadata.gz: 270301527344d9b49c982e1d7c27dd83117f23c56de3e321f67e15aae7696998d423bd7cec72852a188412ee54b4183736917a5ec6c7030eab888b621af611f1
+  data.tar.gz: 66933232dbb0f106362e2a9552f19fc152022c5f9ed9b7ca4bda6fe1d2b3935ce3a3b7288bc2d0130aa8d9bddd6a4e5861ff5ed675e11b763b0e0d171e5749ef

data/lib/ask/web_fetch/backend.rb

@@ -16,6 +16,13 @@ module Ask
     # A backend that fetched the page but found nothing usable in it.
     class EmptyContentError < Error; end
 
+    # A backend that fetched a REGISTRAR PARKING PAGE — an ad for a
+    # parked (for-sale) domain, not the site's content. Deterministic and
+    # terminal: retrying will never turn a parking ad into content, so
+    # the pipeline must classify (not retry) it. A subclass of
+    # EmptyContentError so existing empty-content handling still applies.
+    class ParkedDomainError < EmptyContentError; end
+
     # Network-level failure — timeout, connection refused/reset, bad
     # socket. Transient: the same URL may succeed on retry.
     class TimeoutError < Error; end

data/lib/ask/web_fetch/backends/attached_browser.rb

@@ -27,7 +27,11 @@ module Ask
 
         # Creates a page as a fresh tab in the attached browser.
         def create_page
-          Page.new(client, timeout: @timeout)
+          # Capture the frontmost app once per browser instance, not per
+          # page: a batch of fetches steals focus once and restores to the
+          # same app, and consecutive fetches don't fight each other.
+          @frontmost_app ||= FocusRestorer.capture_frontmost
+          Page.new(client, timeout: @timeout, restore_focus_to: @frontmost_app)
         end
 
         private
@@ -43,9 +47,10 @@ module Ask
         # enough CDP for the Browser backend: navigate, read title/URL/HTML,
         # and report the main-document HTTP status.
         class Page
-          def initialize(client, timeout:)
+          def initialize(client, timeout:, restore_focus_to: nil)
             @client = client
             @timeout = timeout
+            @restore_focus_to = restore_focus_to
             @target_id = @client.command('Target.createTarget', url: 'about:blank')['targetId']
             @session = @client.session(
               @client.command('Target.attachToTarget', targetId: @target_id, flatten: true)['sessionId']
@@ -91,6 +96,12 @@ module Ask
             @client.command('Target.closeTarget', targetId: @target_id)
           rescue Ferrum::Error
             nil
+          ensure
+            # The attached dev Chrome window steals focus when a tab is
+            # created; give it back once the tab is gone — but only to an
+            # app that ISN'T Chrome, so a fetch while the user is already
+            # looking at Chrome doesn't ping-pong.
+            FocusRestorer.restore_frontmost(@restore_focus_to)
           end
 
           def evaluate(expression)
@@ -189,6 +200,48 @@ module Ask
           end
         end
       end
+
+      # macOS dev convenience: the attached Chrome window steals focus when
+      # a tab is created, and a developer working in another app wants it
+      # back once the tab is closed. Captures the frontmost app before the
+      # page is created and re-activates it after close. Deliberately
+      # cheap and quiet: one osascript each way, rescued to nil everywhere
+      # else (prod is headless Linux — this module is a no-op there, and
+      # even on macOS a failure must never fail a fetch).
+      module FocusRestorer
+        # The command runner, injectable for tests (minitest 6 ships no
+        # mock support, so tests swap this instead of stubbing).
+        class << self
+          attr_writer :runner
+
+          def runner
+            @runner ||= method(:system)
+          end
+        end
+
+        def self.capture_frontmost
+          return nil unless RUBY_PLATFORM.include?('darwin')
+
+          out = `osascript -e 'tell application "System Events" to get name of first application process whose frontmost is true' 2>/dev/null`.strip
+          out.empty? ? nil : out
+        rescue StandardError
+          nil
+        end
+
+        def self.restore_frontmost(app)
+          return if app.nil? || app.empty?
+
+          # Never ping-pong: if the user is already looking at Chrome (or
+          # headless/CI has no frontmost app at all), there's nothing to
+          # give back.
+          return if app == 'Google Chrome'
+
+          runner.call("osascript -e 'tell application \"#{app}\" to activate' 2>/dev/null")
+          nil
+        rescue StandardError
+          nil
+        end
+      end
     end
   end
 end

data/lib/ask/web_fetch/backends/browser.rb

@@ -201,8 +201,9 @@ module Ask
           end
           # Browser renders the parked page a JS redirect lands on (the
           # server shell hands /lander to JS) — Local never sees it. The
-          # shared parked-domain detector catches it here.
-          raise EmptyContentError, "parked domain at #{url} — registrar parking page, not site content" if parked_domain?(body)
+          # shared parked-domain detector catches it here; the distinct
+          # ParkedDomainError lets the pipeline classify (never retry) it.
+          raise ParkedDomainError, "parked domain at #{url} — registrar parking page, not site content" if parked_domain?(body)
 
           result = Markdown.generate(body, base_url: url, filter: self.class.content_filter)
           result[:outlinks] = outlink_urls(body, url)

data/lib/ask/web_fetch/backends/local.rb

@@ -55,7 +55,7 @@ module Ask
           # parked pages are fully server-rendered, so both Local and
           # Browser see the same ad.
           if parked_domain?(body)
-            raise EmptyContentError, "parked domain at #{url} — registrar parking page, not site content"
+            raise ParkedDomainError, "parked domain at #{url} — registrar parking page, not site content"
           end
           raise EmptyContentError, "no readable content at #{url}" unless usable_content?(page[:content])
           # The completeness signal: a JS-app shell whose server HTML

data/lib/ask/web_fetch/version.rb

@@ -2,6 +2,6 @@
 
 module Ask
   module WebFetch
-    VERSION = '0.5.7'
+    VERSION = '0.5.9'
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ask-web-fetch
 version: !ruby/object:Gem::Version
-  version: 0.5.7
+  version: 0.5.9
 platform: ruby
 authors:
 - Kaka Ruto