whois-parser-next 0.4.8 → 0.4.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: dd5238487b0cc29b77e6d56f3594c009ea7712fe877fc31a41079dc1e49178be
-  data.tar.gz: 6f94667791431a36eb9b45ce8550143c89b91e9f08d88eafded9411722ae92fc
+  metadata.gz: 9e023caaec0b71465576d51e1aab35e0fa956f22fe435c9ebe91e7eae29e0d18
+  data.tar.gz: e90fe07318d7c90413f88ba7201eade76e7fc8ada882da016f1bcfb9574bc68e
 SHA512:
-  metadata.gz: 454380ba261a261d3f13e91646c54fde3adf7141b5782580323c3511f45b13c1dde9fd5f3f29b03855a090d7d48585304d660788e5a2942f943706670ae104cf
-  data.tar.gz: 21b62020ab6c094190a908d3e3e926b38928589c1727266fb691743b06003662a8ee3c22a973c6904b38f37402f6596e68de25efcf2a2f2443f9079e709f2564
+  metadata.gz: 2a55f20c2508a3a06d2a704e255532940ddf67b7ae3c9249b1a893b60c1f664c56ddd497aeac46cb1e14b4f6a9a8576db3411d318f768504027f0131916232c2
+  data.tar.gz: 3bd083076d0699b8520cd1cd232f09a8c755dcd33192198d7512974f44d73c50044201e34a919eb617c6132079710628c4fa3069e4306b4754292fbc8446f406

data/CHANGELOG.md

@@ -2,6 +2,61 @@
 
 This project uses [Semantic Versioning 2.0.0](http://semver.org/).
 
+#### Release 0.4.10 (2026-11-02)
+
+- SAFETY: A response no longer keeps a scanner looping forever. A reply cut
+  off right after a disclaimer or notice header, with no final newline (for
+  example a Verisign body of just `TERMS OF USE:`, or the CIRA, `.ie`, ATI,
+  SX, PIR and configured Afilias disclaimers), made the tokenizer report
+  success without consuming input, so parsing never finished. The scanner now
+  raises `Whois::ParserError` when tokens stop consuming input, while still
+  allowing the short in-place state changes some scanners use to close a
+  section.
+- SAFETY: The `whois.iana.org` scanner matches repeated field keys literally.
+  A key taken from the response was used as a regular expression, so a key
+  such as `()|^` matched without consuming input and looped forever. Contact
+  handles in the `.fr`-family, `.br` and `.ua` parsers are also matched
+  literally instead of as patterns.
+- TESTED: `spec/whois/scanners/progress_spec.rb` covers each truncated
+  prefix and the regex-shaped IANA keys under a deadline.
+
+#### Release 0.4.9 (2026-10-25)
+
+- ADDED: Added `registrar`, `registrant_contacts` and `technical_contacts` to
+  the `whois.aeda.net.ae` (`.ae`) parser. The registrar comes from the
+  `Registrar ID` and `Registrar Name` lines. Each contact comes from the
+  `<Role> Contact ID`, `Name` and `Organisation` lines. The registry publishes
+  no contact email over port 43; the `Contact Email` line of the legacy layout
+  only pointed to the web WHOIS, so it is not read. `admin_contacts` is not
+  supported because responses carry no admin contact. No-data responses yield
+  no registrar or contact, and blacklisted responses raise
+  `Whois::ResponseIsThrottled`.
+- EVIDENCE: On 2026-10-24 a single lookup of `google.ae` on
+  `whois.aeda.net.ae` (TCP/43) returned the registrar, registrant and technical
+  contact lines in the new `current_contacts_registered.txt` fixture
+  (unmodified, CRLF preserved; it holds organisation and role data only, no
+  personal data).
+- FIXED: The `whois.registry.qa` (`.qa`, `.com.qa` and the other `.qa`
+  second-level zones) parser raised `Whois::ParserError` on every current
+  response. The registry now opens each reply with the instructions of its web
+  lookup form and closes it with a copyright, terms of use and query limit
+  notice; the parser skips both. Current records also renamed their lines:
+  `Registrar` replaces `Registrar ID` and `Registrar Name`, `Domain Status`
+  (an EPP code followed by its ICANN URL) replaces `Status`, and
+  `Registrant Name` and `Tech Name` replace the `<Role> Contact ID/Name/Email`
+  lines. The parser reads both layouts: `status` returns the EPP codes,
+  `registered?` no longer requires the legacy `Status` line, and the registrar
+  and contacts carry a name only, because current records publish no
+  registrar ID, contact ID or email.
+- EVIDENCE: On 2026-10-24 single lookups on `whois.registry.qa` (TCP/43)
+  returned the registered `google.qa` and `ooredoo.com.qa` responses and the
+  `No Data Found` responses for one generated unregistered name in each zone,
+  stored unmodified (CRLF preserved) as the new `current_status_registered.txt`,
+  `current_com_qa_status_registered.txt`, `current_status_available.txt` and
+  `current_com_qa_status_available.txt` fixtures. The registered fixtures hold
+  role and organisation names only, no personal data. The 2026-09-19 probes
+  that received empty responses no longer reproduce.
+
 #### Release 0.4.8 (2026-10-24)
 
 - ADDED: Added `registrant_contacts` to the `whois.trabis.gov.tr` (`.tr`)

data/lib/whois/parser/version.rb

@@ -1,6 +1,6 @@
 module Whois
   class Parser
     # The current library version.
-    VERSION = "0.4.8".freeze
+    VERSION = "0.4.10".freeze
   end
 end

data/lib/whois/parsers/base_nic_fr.rb

@@ -117,7 +117,7 @@ module Whois
         return unless content_for_scanner =~ /#{element}:\s+(.+)\n/
 
         id = ::Regexp.last_match(1)
-        content_for_scanner.scan(/nic-hdl:\s+#{id}\n((.+\n)+)\n/).any? ||
+        content_for_scanner.scan(/nic-hdl:\s+#{Regexp.escape(id)}\n((.+\n)+)\n/).any? ||
             Whois::Parser.bug!(ParserError, "Unable to parse contact block for nic-hdl: #{id}")
         values = build_hash(::Regexp.last_match(1).scan(/(.+?):\s+(.+?)\n/))
 

data/lib/whois/parsers/whois.aeda.net.ae.rb

@@ -62,12 +62,50 @@ module Whois
       property_not_supported :expires_on
 
 
+      property_supported :registrar do
+        id   = field("Registrar ID")
+        name = field("Registrar Name")
+        Parser::Registrar.new(id: id, name: name) if id || name
+      end
+
+
+      property_supported :registrant_contacts do
+        build_contact("Registrant", Parser::Contact::TYPE_REGISTRANT)
+      end
+
+      property_not_supported :admin_contacts
+
+      property_supported :technical_contacts do
+        build_contact("Tech", Parser::Contact::TYPE_TECHNICAL)
+      end
+
+
       property_supported :nameservers do
         content_for_scanner.scan(/Name Server:\s+(.+)\n/).flatten.map do |name|
           Parser::Nameserver.new(:name => name)
         end
       end
 
+
+      private
+
+      def field(label)
+        content_for_scanner[/^#{label}:[ \t]+(.+)$/, 1]&.strip.presence
+      end
+
+      # A contact is the "<Role> Contact ID/Name/Organisation" lines. The
+      # registry publishes no contact email over port 43: the "Contact Email"
+      # line of the legacy layout only pointed to the web WHOIS, so it is not
+      # read.
+      def build_contact(role, type)
+        id           = field("#{role} Contact ID")
+        name         = field("#{role} Contact Name")
+        organization = field("#{role} Contact Organisation")
+        return [] unless id || name || organization
+
+        [Parser::Contact.new(type: type, id: id, name: name, organization: organization)]
+      end
+
     end
 
   end

data/lib/whois/parsers/whois.registro.br.rb

@@ -96,7 +96,7 @@ module Whois
         return unless content_for_scanner =~ /#{element}:\s+(.+)\n/
 
         id = ::Regexp.last_match(1)
-        content_for_scanner.scan(/nic-hdl-br:\s+#{id}\n((.+\n)+)\n/).any? ||
+        content_for_scanner.scan(/nic-hdl-br:\s+#{Regexp.escape(id)}\n((.+\n)+)\n/).any? ||
             Whois.bug!(ParserError, "Unable to parse contact block for nic-hdl-br: #{id}")
         values = build_hash(::Regexp.last_match(1).scan(/(.+?):\s+(.+?)\n/))
 

data/lib/whois/parsers/whois.registry.qa.rb

@@ -8,6 +8,7 @@
 
 
 require_relative 'base_shared1'
+require 'whois/scanners/whois.registry.qa'
 
 
 module Whois
@@ -15,10 +16,44 @@ module Whois
 
     # Parser for the whois.registry.qa server.
     #
+    # Current responses use the "Registrar", "Domain Status" and
+    # "<Role> Name" lines; the legacy layout used "Registrar ID/Name",
+    # "Status" and "<Role> Contact ID/Name/Email". Both are read.
+    #
     # @see Whois::Parsers::Example
     #   The Example parser for the list of all available methods.
     #
     class WhoisRegistryQa < BaseShared1
+
+      self.scanner = Scanners::WhoisRegistryQa
+
+
+      # "Domain Status" carries the EPP status code followed by its ICANN URL.
+      property_supported :status do
+        Array.wrap(node("Status") || node("Domain Status")).map { |value| value.split.first }
+      end
+
+      property_supported :registered? do
+        !available? && !status.empty? && !node("Domain Name").to_s.strip.empty?
+      end
+
+
+      property_supported :registrar do
+        name = node("Registrar Name") || node("Registrar")
+        Parser::Registrar.new(id: node("Registrar ID"), name: name, organization: name) if name.present?
+      end
+
+
+      private
+
+      def build_contact(element, type)
+        return super if node("#{element} Contact ID")
+
+        node("#{element} Name") do |name|
+          Parser::Contact.new(type: type, name: name) if name.present?
+        end
+      end
+
     end
 
   end

data/lib/whois/parsers/whois.ua.rb

@@ -65,7 +65,7 @@ module Whois
           return if contact_ids.empty?
 
           contact_ids.map do |contact_id|
-            textblock = content.slice(/contact-id:\s+#{contact_id}\n((?:.+\n)+)\n/, 1)
+            textblock = content.slice(/contact-id:\s+#{Regexp.escape(contact_id)}\n((?:.+\n)+)\n/, 1)
 
             address = textblock.scan(/address:\s+(.+)\n/).flatten
             address = address.reject { |a| a == "n/a" }
@@ -137,7 +137,7 @@ module Whois
           return if contact_ids.empty?
 
           contact_ids.map do |contact_id|
-            textblock = content.slice(/nic-handle:\s+#{contact_id}\n((?:.+\n)+)\n/, 1)
+            textblock = content.slice(/nic-handle:\s+#{Regexp.escape(contact_id)}\n((?:.+\n)+)\n/, 1)
 
             address = textblock.scan(/address:\s+(.+)\n/).flatten
             zip = nil

data/lib/whois/scanners/base.rb

@@ -6,6 +6,9 @@ module Whois
       class_attribute :tokenizers
       self.tokenizers = []
 
+      # Consecutive tokens allowed to succeed without consuming input.
+      MAX_IDLE_TOKENS = 100
+
       def self.tokenizer(name, &block)
         define_method(name, &block)
       end
@@ -26,7 +29,21 @@ module Whois
         @ast = {}
 
         @input = StringScanner.new(content)
-        tokenize until @input.eos?
+        idle = 0
+        until @input.eos?
+          position = @input.pos
+          tokenize
+          if @input.pos == position
+            # Some tokenizers succeed once without consuming input to change
+            # state (e.g. close a section) for the next one. One that keeps
+            # succeeding in place loops forever, e.g. a disclaimer header matched
+            # by lookahead in a reply truncated before its final newline.
+            idle += 1
+            unexpected_token if idle > MAX_IDLE_TOKENS
+          else
+            idle = 0
+          end
+        end
 
         @ast
       end

data/lib/whois/scanners/iana.rb

@@ -79,7 +79,9 @@ module Whois
       end
 
       def parse_section_pair_newlinevalue(key)
-        if @input.scan(/^#{key}:\s*(.+)\n/)
+        # The key comes from the response: match it literally. As a pattern, a
+        # key such as "()|^" matches without consuming input and loops forever.
+        if @input.scan(/^#{Regexp.escape(key)}:\s*(.+)\n/)
           @input[1].strip
         end
       end

data/lib/whois/scanners/whois.registry.qa.rb

@@ -0,0 +1,29 @@
+require_relative 'base_shared1'
+
+module Whois
+  module Scanners
+
+    # Scanner for the whois.registry.qa records.
+    class WhoisRegistryQa < BaseShared1
+
+      self.tokenizers = [:skip_head, :skip_end] + tokenizers
+
+
+      # Current responses open with the instructions of the registry's web
+      # lookup form.
+      tokenizer :skip_head do
+        @input.skip(/^Welcome to Qatar Domains Registry WHOIS Lookup\n(?:.*\n)*?Type the word you see into the verification box to proceed\.\n/)
+      end
+
+      # The status code reference ends the record; the copyright, terms of use
+      # and query limit notice follow it.
+      tokenizer :skip_end do
+        if @input.check(/^For more information on Whois status codes, please visit /)
+          @input.terminate
+        end
+      end
+
+    end
+
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: whois-parser-next
 version: !ruby/object:Gem::Version
-  version: 0.4.8
+  version: 0.4.10
 platform: ruby
 authors:
 - Simone Carletti
@@ -783,6 +783,7 @@ files:
 - lib/whois/scanners/whois.nic.it.rb
 - lib/whois/scanners/whois.nic.net.jo.rb
 - lib/whois/scanners/whois.pir.org.rb
+- lib/whois/scanners/whois.registry.qa.rb
 - lib/whois/scanners/whois.rnids.rs.rb
 - lib/whois/scanners/whois.smallregistry.net.rb
 - lib/whois/scanners/whois.srs.net.nz.rb