From 4325835f92f3f142ebd91a3fdba4e1f1ab7f1cfb Mon Sep 17 00:00:00 2001
From: Nobuyoshi Nakada <nobu@ruby-lang.org>
Date: Thu, 16 May 2024 11:26:51 +0900
Subject: [PATCH] Read quoted attributes in chunks (#126)

---
 Gemfile                         |  1 +
 lib/rexml/parsers/baseparser.rb | 20 ++++++++++----------
 lib/rexml/source.rb             | 29 ++++++++++++++++++++++++-----
 test/test_document.rb           | 11 +++++++++++
 4 files changed, 46 insertions(+), 15 deletions(-)

diff --git a/Gemfile b/Gemfile
index 042ef8ac..f78cc861 100644
--- a/Gemfile
+++ b/Gemfile
@@ -10,4 +10,5 @@ group :development do
   gem "bundler"
   gem "rake"
   gem "test-unit"
+  gem "test-unit-ruby-core"
 end
diff --git a/lib/rexml/parsers/baseparser.rb b/lib/rexml/parsers/baseparser.rb
index 8d62391c..d09237c5 100644
--- a/lib/rexml/parsers/baseparser.rb
+++ b/lib/rexml/parsers/baseparser.rb
@@ -628,17 +628,17 @@ def parse_attributes(prefixes, curr_ns)
               message = "Missing attribute equal: <#{name}>"
               raise REXML::ParseException.new(message, @source)
             end
-            unless match = @source.match(/(['"])(.*?)\1\s*/um, true)
-              if match = @source.match(/(['"])/, true)
-                message =
-                  "Missing attribute value end quote: <#{name}>: <#{match[1]}>"
-                raise REXML::ParseException.new(message, @source)
-              else
-                message = "Missing attribute value start quote: <#{name}>"
-                raise REXML::ParseException.new(message, @source)
-              end
+            unless match = @source.match(/(['"])/, true)
+              message = "Missing attribute value start quote: <#{name}>"
+              raise REXML::ParseException.new(message, @source)
+            end
+            quote = match[1]
+            value = @source.read_until(quote)
+            unless value.chomp!(quote)
+              message = "Missing attribute value end quote: <#{name}>: <#{quote}>"
+              raise REXML::ParseException.new(message, @source)
             end
-            value = match[2]
+            @source.match(/\s*/um, true)
             if prefix == "xmlns"
               if local_part == "xml"
                 if value != "http://www.w3.org/XML/1998/namespace"
diff --git a/lib/rexml/source.rb b/lib/rexml/source.rb
index 7f47c2be..999751b4 100644
--- a/lib/rexml/source.rb
+++ b/lib/rexml/source.rb
@@ -65,7 +65,11 @@ def encoding=(enc)
       encoding_updated
     end
 
-    def read
+    def read(term = nil)
+    end
+
+    def read_until(term)
+      @scanner.scan_until(Regexp.union(term)) or @scanner.rest
     end
 
     def ensure_buffer
@@ -158,9 +162,9 @@ def initialize(arg, block_size=500, encoding=nil)
       end
     end
 
-    def read
+    def read(term = nil)
       begin
-        @scanner << readline
+        @scanner << readline(term)
         true
       rescue Exception, NameError
         @source = nil
@@ -168,6 +172,21 @@ def read
       end
     end
 
+    def read_until(term)
+      pattern = Regexp.union(term)
+      data = []
+      begin
+        until str = @scanner.scan_until(pattern)
+          @scanner << readline(term)
+        end
+      rescue EOFError
+        @scanner.rest
+      else
+        read if @scanner.eos? and !@source.eof?
+        str
+      end
+    end
+
     def ensure_buffer
       read if @scanner.eos? && @source
     end
@@ -218,8 +237,8 @@ def current_line
     end
 
     private
-    def readline
-      str = @source.readline(@line_break)
+    def readline(term = nil)
+      str = @source.readline(term || @line_break)
       if @pending_buffer
         if str.nil?
           str = @pending_buffer
diff --git a/test/test_document.rb b/test/test_document.rb
index 953656f8..f96bfd5d 100644
--- a/test/test_document.rb
+++ b/test/test_document.rb
@@ -1,8 +1,12 @@
 # -*- coding: utf-8 -*-
 # frozen_string_literal: false
 
+require 'core_assertions'
+
 module REXMLTests
   class TestDocument < Test::Unit::TestCase
+    include Test::Unit::CoreAssertions
+
     def test_version_attributes_to_s
       doc = REXML::Document.new(<<~eoxml)
         <?xml version="1.0" encoding="UTF-8" standalone="no"?>
@@ -198,6 +202,13 @@ def test_xml_declaration_standalone
       assert_equal('no', doc.stand_alone?, bug2539)
     end
 
+    def test_gt_linear_performance
+      seq = [10000, 50000, 100000, 150000, 200000]
+      assert_linear_performance(seq) do |n|
+        REXML::Document.new('<test testing="' + ">" * n + '"></test>')
+      end
+    end
+
     class WriteTest < Test::Unit::TestCase
       def setup
         @document = REXML::Document.new(<<-EOX)