Read generated files in binary mode

So that fixes crash with "invalid byte sequence in US-ASCII" on ci.rvm.jp and some RubyCIs.

Author: nobu

test/rdoc/test_rdoc_generator_darkfish.rb

@@ -107,14 +107,13 @@ def test_generate
 
  encoding = Regexp.escape Encoding::UTF_8.name
 
- assert_match %r%<meta charset="#{encoding}">%, File.read('index.html')
- assert_match %r%<meta charset="#{encoding}">%, File.read('Object.html')
+ assert_match %r%<meta charset="#{encoding}">%, File.binread('index.html')
+ assert_match %r%<meta charset="#{encoding}">%, File.binread('Object.html')
 
- refute_match(/Ignored/, File.read('index.html'))
- summary = File.read('index.html')[%r[<summary.*Klass\.html.*</summary>.*</details>]m]
+ refute_match(/Ignored/, File.binread('index.html'))
+ summary = File.binread('index.html')[%r[<summary.*Klass\.html.*</summary>.*</details>]m]
  assert_match(%r[Klass/Inner\.html".*>Inner<], summary)
- omit 'The following line crashes with "invalid byte sequence in US-ASCII" on ci.rvm.jp and some RubyCIs'
- klassnav = File.read('Klass.html')[%r[<div class="nav-section">.*<div id="class-metadata">]m]
+ klassnav = File.binread('Klass.html')[%r[<div class="nav-section">.*<div id="class-metadata">]m]
  assert_match(
  %r[<li>\s*<details open>\s*<summary>\s*<a href=\S+>Heading 1</a>\s*</summary>\s*<ul]m,
  klassnav
@@ -131,15 +130,15 @@ def test_generate_page
  @g.generate
  assert_file 'outer_rdoc.html'
  assert_file 'outer/inner_rdoc.html'
- index = File.read('index.html')
+ index = File.binread('index.html')
  re = %r[<summary><a href="\./outer_rdoc\.html">outer</a></summary>.*?</details>]m
  assert_match(re, index)
  summary = index[re]
  assert_match %r[<a href="\./outer/inner_rdoc.html">inner</a>], summary
  re = %r[<details open><summary><a href="\./outer_rdoc\.html">outer</a></summary>.*?</details>]m
- assert_match(re, File.read('outer_rdoc.html'))
+ assert_match(re, File.binread('outer_rdoc.html'))
  re = %r[<details open><summary><a href="\.\./outer_rdoc\.html">outer</a></summary>.*?</details>]m
- assert_match(re, File.read('outer/inner_rdoc.html'))
+ assert_match(re, File.binread('outer/inner_rdoc.html'))
  end
 
  def test_generate_dry_run
@@ -276,7 +275,7 @@ def test_generated_filename_with_html_tag
  @g.generate
 
  Dir.glob("*.html", base: @tmpdir) do |html|
- File.read(File.join(@tmpdir, html)).scan(/.*should be escaped.*/) do |line|
+ File.binread(File.join(@tmpdir, html)).scan(/.*should be escaped.*/) do |line|
  assert_not_include line, "<em>", html
  end
  end
@@ -294,23 +293,23 @@ def test_template_stylesheets
  @g.generate
 
  assert_file base
- assert_include File.read('index.html'), %Q[href="./#{base}"]
+ assert_include File.binread('index.html'), %Q[href="./#{base}"]
  end
 
  def test_title
  title = "RDoc Test".freeze
  @options.title = title
  @g.generate
 
- assert_main_title(File.read('index.html'), title)
+ assert_main_title(File.binread('index.html'), title)
  end
 
  def test_title_escape
  title = %[<script>alert("RDoc")</script>].freeze
  @options.title = title
  @g.generate
 
- assert_main_title(File.read('index.html'), title)
+ assert_main_title(File.binread('index.html'), title)
  end
 
  ##