whois-parser-next 0.4.6 → 0.4.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 66e7767b738ae0cf7df9736dc9a6a7685f9ddf7d344506a3a4233accf905f0d2
-  data.tar.gz: c15b010e832b1e52cebdb02fd473e8c75176a0f232c777ba385e39b17676b327
+  metadata.gz: bf39e6c3afbb1b6a0cb49c48cb3fb11cf6db7d0162aaba77c9027ca4d2fadc3f
+  data.tar.gz: 1134d08ab3b1bcef0551f8008c366234805614e86cbe02ffe90501f5203dae84
 SHA512:
-  metadata.gz: 587e231a29331459c3d14a9cd655ef2fa0230b819cbbcde397be5465500fd4a8eef4b5742407f402ed203d7104a9dc235553795427ec1b2daa1d08a34a24e82e
-  data.tar.gz: d4e01866484c46fcf88ef68889d677459ae34db5c0920dcabfc4bb2eaf390f16af8d5524436056a543154320742d51cc445dfb3553de4eb4386adc5c0ef511f1
+  metadata.gz: addd71819acd60b21142b60e419625ac8ebbed5b160b33e6cfe5562bcd064727e04b98c820f53bb98c1e1649158829b10b10d3c9ae7b900b47b25f57c1023043
+  data.tar.gz: 56a6dde832f4f0a9990e7753ea7f709a2bcba56d461d4eac0aa299bdd69b5896fd3b584647b6766267e621c4f6d52e42961a1fca1521c619b086f96f2dc80fcf

data/CHANGELOG.md

@@ -2,6 +2,26 @@
 
 This project uses [Semantic Versioning 2.0.0](http://semver.org/).
 
+#### Release 0.4.7 (2026-10-19)
+
+- ADDED: Added strict parsing for the official `.gt` and `.com.gt` web WHOIS
+  response. Registration requires one result heading with a single valid domain
+  and `ACTIVO`; absence requires one `alert alert-danger` result with the same
+  domain and the registry's exact `no está registrado` marker. Duplicate,
+  contradictory, malformed, mismatched, incomplete, denied, throttled, or
+  challenge responses remain non-positive. The parser ignores scripts, styles,
+  comments, and unrelated page content; contacts and dates are not parsed.
+- EVIDENCE: The official UVG registry lookup is
+  [`whois.php?dn=...`](https://www.gt/sitio/whois.php?dn={domain}). Bounded
+  2026-10-19 observations returned registered results for `google.gt` and
+  `google.com.gt`, and exact absence results for generated eight-letter names
+  under both suffixes. The registry's
+  [policy](https://www.gt/sitio/registration_policy.php?lang=en) limits lookup
+  to individual names rather than lists of registered names. Regression
+  examples use only minimal synthetic HTML. This is a web WHOIS source; no
+  TCP/43 or RDAP service is claimed. See the
+  [dated evidence report](docs/audits/guatemala-web-whois-2026-10-19.md).
+
 #### Release 0.4.6 (2026-10-19)
 
 - ADDED: Added `.ga` parsing for ANINF's current French WHOIS response, requiring

data/lib/whois/parser/version.rb

@@ -1,6 +1,6 @@
 module Whois
   class Parser
     # The current library version.
-    VERSION = "0.4.6".freeze
+    VERSION = "0.4.7".freeze
   end
 end

data/lib/whois/parsers/www.gt.rb

@@ -0,0 +1,107 @@
+# frozen_string_literal: true
+
+require 'cgi'
+require 'strscan'
+require_relative 'base'
+require_relative 'registry_response_safety'
+
+module Whois
+  class Parsers
+    # The registry's per-domain public web WHOIS, observed 2026-10-19.
+    # Inspect only its result heading/alert, never navigation or script text.
+    class WwwGt < Base
+      include RegistryResponseSafety
+
+      property_supported(:status) { web_result[:status] }
+      property_supported(:domain) { web_result[:domain] }
+      property_supported(:registered?) { status == :registered }
+      property_supported(:available?) { status == :available }
+
+      def response_unavailable?
+        super || content_for_scanner.match?(/<cap-widget\b|\bclass=["']g-recaptcha["']|cf-chl-/i)
+      end
+
+      private
+
+      def web_result
+        @web_result ||= classify_web_result
+      end
+
+      def classify_web_result
+        blocks = result_blocks
+        return { status: :unknown } unless blocks
+
+        candidates = blocks.filter_map do |kind, text|
+          text = CGI.unescapeHTML(text.gsub('&aacute;', 'á').gsub('&Aacute;', 'Á')).gsub(/[ \t\r\n]+/, ' ').strip
+          if kind == 'h3' && (match = /\A(\S+)\. ACTIVO\z/i.match(text))
+            [match[1].downcase, :registered]
+          elsif kind == 'div' && (match = /\A(\S+)\. no está registrado \.\z/i.match(text))
+            [match[1].downcase, :available]
+          end
+        end
+        return { status: :unknown } unless candidates.size == 1
+
+        name, state = candidates.first
+        return { status: :unknown } unless name.bytesize <= 253 &&
+          name.match?(/\A(?:[a-z0-9](?:[a-z0-9-]{0,61}[a-z0-9])?\.)+gt\z/)
+
+        { domain: name, status: state }
+      end
+
+      # One forward scan, including malformed tags. Limit each tag before
+      # inspecting attributes and skip scripts, styles and comments entirely.
+      def result_blocks
+        input = StringScanner.new(content_for_scanner)
+        blocks = []
+        active = nil
+        depth = 0
+        text = +''
+        until input.eos?
+          chunk = input.scan_until(/</)
+          unless chunk
+            text << input.rest if active
+            input.terminate
+            break
+          end
+          text << chunk[0...-1] if active
+          if input.peek(3) == '!--'
+            return nil unless input.scan_until(/-->/)
+            next
+          end
+          tag = input.scan_until(/>/)
+          return nil unless tag && tag.bytesize <= 16_384
+          match = /\A[ \t]*(\/?)[ \t]*([a-z0-9]+)/i.match(tag)
+          next unless match
+          closing = match[1] == '/'
+          name = match[2].downcase
+          if !closing && %w[script style].include?(name)
+            return nil if active
+            return nil unless input.scan_until(%r{</#{name}[ \t\r\n]*>}i)
+            next
+          end
+          if active
+            depth += 1 if !closing && name == active
+            depth -= 1 if closing && name == active
+            if depth.zero?
+              blocks << [active, text]
+              active = nil
+              text = +''
+            else
+              text << ' '
+            end
+          elsif !closing && (name == 'h3' || (name == 'div' && alert_classes?(tag)))
+            active = name
+            depth = 1
+          end
+        end
+        active ? nil : blocks
+      end
+
+      def alert_classes?(tag)
+        value = tag[/\bclass[ \t]*=[ \t]*["']([^"']*)["']/i, 1]
+        classes = value.to_s.split
+        classes.include?('alert') && classes.include?('alert-danger')
+      end
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: whois-parser-next
 version: !ruby/object:Gem::Version
-  version: 0.4.6
+  version: 0.4.7
 platform: ruby
 authors:
 - Simone Carletti
@@ -753,6 +753,7 @@ files:
 - lib/whois/parsers/whois.za.org.rb
 - lib/whois/parsers/whois.zicta.zm.rb
 - lib/whois/parsers/whois1.nic.bi.rb
+- lib/whois/parsers/www.gt.rb
 - lib/whois/parsers/za_central_registry.rb
 - lib/whois/safe_record.rb
 - lib/whois/scanners/base.rb