(maint) Move array parser logic into a util module

3 files changed, 331 insertions, 186 deletions

diff --git a/lib/puppet/provider/augeas/augeas.rb b/lib/puppet/provider/augeas/augeas.rb
index b64b0b3..ee52ee6 100644
--- a/lib/puppet/provider/augeas/augeas.rb
+++ b/lib/puppet/provider/augeas/augeas.rb
@@ -18,11 +18,13 @@ require 'strscan'
 require 'puppet/util'
 require 'puppet/util/diff'
 require 'puppet/util/package'
+require 'puppet_x/augeas/util/parser'
 
 Puppet::Type.type(:augeas).provide(:augeas) do
   include Puppet::Util
   include Puppet::Util::Diff
   include Puppet::Util::Package
+  include PuppetX::Augeas::Util::Parser
 
   confine feature: :augeas
 
@@ -280,7 +282,7 @@ Puppet::Type.type(:augeas).provide(:augeas) do
     when '=='
       begin
         arg = clause_array.shift
-        new_array = to_array(arg)
+        new_array = parse_to_array(arg)
         return_value = (values == new_array)
       rescue
         fail(_('Invalid array in command: %{cmd}') % { cmd: cmd_array.join(' ') })
@@ -288,7 +290,7 @@ Puppet::Type.type(:augeas).provide(:augeas) do
     when '!='
       begin
         arg = clause_array.shift
-        new_array = to_array(arg)
+        new_array = parse_to_array(arg)
         return_value = (values != new_array)
       rescue
         fail(_('Invalid array in command: %{cmd}') % { cmd: cmd_array.join(' ') })
@@ -336,7 +338,7 @@ Puppet::Type.type(:augeas).provide(:augeas) do
     when '=='
       begin
         arg = clause_array.shift
-        new_array = to_array(arg)
+        new_array = parse_to_array(arg)
         return_value = (result == new_array)
       rescue
         fail(_('Invalid array in command: %{cmd}') % { cmd: cmd_array.join(' ') })
@@ -344,7 +346,7 @@ Puppet::Type.type(:augeas).provide(:augeas) do
     when '!='
       begin
         arg = clause_array.shift
-        new_array = to_array(arg)
+        new_array = parse_to_array(arg)
         return_value = (result != new_array)
       rescue
         fail(_('Invalid array in command: %{cmd}') % { cmd: cmd_array.join(' ') })
@@ -571,186 +573,4 @@ Puppet::Type.type(:augeas).provide(:augeas) do
     end
   end
   # rubocop:enable Style/GuardClause
-
-  def to_array(string)
-    s = StringScanner.new(string)
-    match = array_open(s)
-    raise "Unable to parse array. Unexpected character at: #{s.rest}" if match.nil?
-
-    array_content = array_values(s)
-
-    match = array_close(s)
-    raise "Unable to parse array. Unexpected character at: #{s.rest}" if match.nil?
-
-    array_content
-  end
-  private :to_array
-
-  def array_open(scanner)
-    scanner.scan(%r{\s*\[\s*})
-  end
-  private :array_open
-
-  def array_close(scanner)
-    scanner.scan(%r{\s*\]\s*})
-  end
-  private :array_close
-
-  def array_separator(scanner)
-    scanner.scan(%r{\s*,\s*})
-  end
-  private :array_separator
-
-  def single_quote_unescaped_char(scanner)
-    scanner.scan(%r{[^'\\]})
-  end
-  private :single_quote_unescaped_char
-
-  def single_quote_escaped_char(scanner)
-    scanner.scan(%r{\\(['\\])}) && scanner[1]
-  end
-  private :single_quote_escaped_char
-
-  def single_quote_char(scanner)
-    single_quote_escaped_char(scanner) || single_quote_unescaped_char(scanner)
-  end
-  private :single_quote_char
-
-  def double_quote_unescaped_char(scanner)
-    scanner.scan(%r{[^"\\]})
-  end
-  private :double_quote_unescaped_char
-
-  # This handles the possible Ruby escape sequences in double-quoted strings,
-  # except for \M-x, \M-\C-x, \M-\cx, \c\M-x, \c?, and \C-?. The full list of
-  # escape sequences, and their meanings is taken from:
-  # https://github.com/ruby/ruby/blob/90fdfec11a4a42653722e2ce2a672d6e87a57b8e/doc/syntax/literals.rdoc#strings
-  def double_quote_escaped_char(scanner)
-    match = scanner.scan(%r{\\(["\\abtnvfres0-7xu])})
-    return nil if match.nil?
-
-    case scanner[1]
-    when '\\' then return '\\'
-    when '"'  then return '"'
-    when 'a'  then return "\a"
-    when 'b'  then return "\b"
-    when 't'  then return "\t"
-    when 'n'  then return "\n"
-    when 'v'  then return "\v"
-    when 'f'  then return "\f"
-    when 'r'  then return "\r"
-    when 'e'  then return "\e"
-    when 's'  then return "\s"
-    when %r{[0-7]}
-      octal_character = scanner[1]
-      other_digits = scanner.scan(%r{[0-7]{1,2}})
-      octal_character << other_digits unless other_digits.nil?
-
-      return octal_character.to_i(8).chr
-    when 'x'
-      hex_character = scanner.scan(%r{[0-9a-fA-F]{1,2}})
-      return nil if hex_character.nil?
-
-      hex_character.to_i(16).chr
-    when 'u'
-      return unicode_short_hex_character(scanner) || unicode_long_hex_characters(scanner)
-    else
-      # Not a valid escape sequence as far as we're concerned.
-      return nil
-    end
-  end
-  private :double_quote_escaped_char
-
-  def unicode_short_hex_character(scanner)
-    unicode_character = scanner.scan(%r{[0-9a-fA-F]{4}})
-    return nil if unicode_character.nil?
-
-    [unicode_character.hex].pack 'U'
-  end
-  private :unicode_short_hex_character
-
-  def unicode_long_hex_characters(scanner)
-    unicode_string = ''
-    return nil unless scanner.scan(%r|{|)
-
-    loop do
-      char = scanner.scan(%r{[0-9a-fA-F]{1,6}})
-      break if char.nil?
-      unicode_string << [char.hex].pack('U')
-
-      separator = scanner.scan(%r{\s})
-      break if separator.nil?
-    end
-
-    return nil if scanner.scan(%r|}|).nil? || unicode_string.empty?
-
-    unicode_string
-  end
-  private :unicode_long_hex_characters
-
-  def single_quoted_string(scanner)
-    quoted_string = ''
-
-    match = scanner.scan(%r{'})
-    return nil if match.nil?
-
-    loop do
-      match = single_quote_char(scanner)
-      break if match.nil?
-
-      quoted_string << match
-    end
-
-    match = scanner.scan(%r{'})
-    return quoted_string if match
-
-    nil
-  end
-  private :single_quoted_string
-
-  def double_quote_char(scanner)
-    double_quote_escaped_char(scanner) || double_quote_unescaped_char(scanner)
-  end
-  private :double_quote_char
-
-  def double_quoted_string(scanner)
-    quoted_string = ''
-
-    match = scanner.scan(%r{"})
-    return nil if match.nil?
-
-    loop do
-      match = double_quote_char(scanner)
-      break if match.nil?
-
-      quoted_string << match
-    end
-
-    match = scanner.scan(%r{"})
-    return quoted_string if match
-
-    nil
-  end
-  private :double_quoted_string
-
-  def quoted_string(scanner)
-    single_quoted_string(scanner) || double_quoted_string(scanner)
-  end
-  private :quoted_string
-
-  def array_values(scanner)
-    values = []
-
-    loop do
-      match = quoted_string(scanner)
-      break if match.nil?
-      values << match
-
-      match = array_separator(scanner)
-      break if match.nil?
-    end
-
-    values
-  end
-  private :array_values
 end
diff --git a/lib/puppet_x/augeas/util/parser.rb b/lib/puppet_x/augeas/util/parser.rb
new file mode 100644
index 0000000..1abf42f
--- /dev/null
+++ b/lib/puppet_x/augeas/util/parser.rb
@@ -0,0 +1,215 @@
+# rubocop:disable Style/Documentation
+module PuppetX; end
+module PuppetX::Augeas; end
+module PuppetX::Augeas::Util; end
+# rubocop:enable Style/Documentation
+
+# Container for helpers to parse user provided data contained in manifests.
+module PuppetX::Augeas::Util::Parser
+  TOKEN_ARRAY_CLOSE                 = %r{\s*\]\s*}
+  TOKEN_ARRAY_OPEN                  = %r{\s*\[\s*}
+  TOKEN_ARRAY_SEPARATOR             = %r{\s*,\s*}
+  TOKEN_CLOSE_CURLY                 = %r|}|
+  TOKEN_DOUBLE_QUOTE                = %r{"}
+  TOKEN_DOUBLE_QUOTE_ESCAPED_CHAR   = %r{\\(["\\abtnvfres0-7xu])}
+  TOKEN_DOUBLE_QUOTE_UNESCAPED_CHAR = %r{[^"\\]}
+  TOKEN_HEX_CHAR                    = %r{[0-9a-fA-F]{1,2}}
+  TOKEN_OCTAL_CHAR                  = %r{[0-7]{1,3}}
+  TOKEN_OPEN_CURLY                  = %r|{|
+  TOKEN_SINGLE_QUOTE                = %r{'}
+  TOKEN_SINGLE_QUOTE_ESCAPED_CHAR   = %r{\\(['\\])}
+  TOKEN_SINGLE_QUOTE_UNESCAPED_CHAR = %r{[^'\\]}
+  TOKEN_SPACE                       = %r{\s}
+  TOKEN_UNICODE_LONG_HEX_CHAR       = %r{[0-9a-fA-F]{1,6}}
+  TOKEN_UNICODE_SHORT_HEX_CHAR      = %r{[0-9a-fA-F]{4}}
+
+  # Parse a string into the (nearly) equivalent Ruby array. This only handles
+  # arrays with string members (double-, or single-quoted), and does not
+  # support the full quite of escape sequences that Ruby allows in
+  # double-quoted strings.
+  #
+  # @param [String] The string to be parsed.
+  # @return [Array<String>] The parsed array elements, including handling any
+  #   escape sequences.
+  def parse_to_array(string)
+    s = StringScanner.new(string)
+    match = array_open(s)
+    raise "Unexpected character in array at: #{s.rest}" if match.nil?
+
+    array_content = array_values(s)
+
+    match = array_close(s)
+    raise "Unexpected character in array at: #{s.rest}" if match.nil? || !s.empty?
+
+    array_content
+  end
+
+  def array_open(scanner)
+    scanner.scan(TOKEN_ARRAY_OPEN)
+  end
+  private :array_open
+
+  def array_close(scanner)
+    scanner.scan(TOKEN_ARRAY_CLOSE)
+  end
+  private :array_close
+
+  def array_separator(scanner)
+    scanner.scan(TOKEN_ARRAY_SEPARATOR)
+  end
+  private :array_separator
+
+  def single_quote_unescaped_char(scanner)
+    scanner.scan(TOKEN_SINGLE_QUOTE_UNESCAPED_CHAR)
+  end
+  private :single_quote_unescaped_char
+
+  def single_quote_escaped_char(scanner)
+    scanner.scan(TOKEN_SINGLE_QUOTE_ESCAPED_CHAR) && scanner[1]
+  end
+  private :single_quote_escaped_char
+
+  def single_quote_char(scanner)
+    single_quote_escaped_char(scanner) || single_quote_unescaped_char(scanner)
+  end
+  private :single_quote_char
+
+  def double_quote_unescaped_char(scanner)
+    scanner.scan(TOKEN_DOUBLE_QUOTE_UNESCAPED_CHAR)
+  end
+  private :double_quote_unescaped_char
+
+  # This handles the possible Ruby escape sequences in double-quoted strings,
+  # except for \M-x, \M-\C-x, \M-\cx, \c\M-x, \c?, and \C-?. The full list of
+  # escape sequences, and their meanings is taken from:
+  # https://github.com/ruby/ruby/blob/90fdfec11a4a42653722e2ce2a672d6e87a57b8e/doc/syntax/literals.rdoc#strings
+  def double_quote_escaped_char(scanner)
+    match = scanner.scan(TOKEN_DOUBLE_QUOTE_ESCAPED_CHAR)
+    return nil if match.nil?
+
+    case scanner[1]
+    when '\\' then return '\\'
+    when '"'  then return '"'
+    when 'a'  then return "\a"
+    when 'b'  then return "\b"
+    when 't'  then return "\t"
+    when 'n'  then return "\n"
+    when 'v'  then return "\v"
+    when 'f'  then return "\f"
+    when 'r'  then return "\r"
+    when 'e'  then return "\e"
+    when 's'  then return "\s"
+    when %r{[0-7]}
+      # Back the scanner up by one byte so we can grab all of the potential
+      # octal digits at the same time.
+      scanner.pos = scanner.pos - 1
+      octal_character = scanner.scan(TOKEN_OCTAL_CHAR)
+
+      return octal_character.to_i(8).chr
+    when 'x'
+      hex_character = scanner.scan(TOKEN_HEX_CHAR)
+      return nil if hex_character.nil?
+
+      hex_character.to_i(16).chr
+    when 'u'
+      return unicode_short_hex_character(scanner) || unicode_long_hex_characters(scanner)
+    else
+      # Not a valid escape sequence as far as we're concerned.
+      return nil
+    end
+  end
+  private :double_quote_escaped_char
+
+  def unicode_short_hex_character(scanner)
+    unicode_character = scanner.scan(TOKEN_UNICODE_SHORT_HEX_CHAR)
+    return nil if unicode_character.nil?
+
+    [unicode_character.hex].pack 'U'
+  end
+  private :unicode_short_hex_character
+
+  def unicode_long_hex_characters(scanner)
+    unicode_string = ''
+    return nil unless scanner.scan(TOKEN_OPEN_CURLY)
+
+    loop do
+      char = scanner.scan(TOKEN_UNICODE_LONG_HEX_CHAR)
+      break if char.nil?
+      unicode_string << [char.hex].pack('U')
+
+      separator = scanner.scan(TOKEN_SPACE)
+      break if separator.nil?
+    end
+
+    return nil if scanner.scan(TOKEN_CLOSE_CURLY).nil? || unicode_string.empty?
+
+    unicode_string
+  end
+  private :unicode_long_hex_characters
+
+  def single_quoted_string(scanner)
+    quoted_string = ''
+
+    match = scanner.scan(TOKEN_SINGLE_QUOTE)
+    return nil if match.nil?
+
+    loop do
+      match = single_quote_char(scanner)
+      break if match.nil?
+
+      quoted_string << match
+    end
+
+    match = scanner.scan(TOKEN_SINGLE_QUOTE)
+    return quoted_string if match
+
+    nil
+  end
+  private :single_quoted_string
+
+  def double_quote_char(scanner)
+    double_quote_escaped_char(scanner) || double_quote_unescaped_char(scanner)
+  end
+  private :double_quote_char
+
+  def double_quoted_string(scanner)
+    quoted_string = ''
+
+    match = scanner.scan(TOKEN_DOUBLE_QUOTE)
+    return nil if match.nil?
+
+    loop do
+      match = double_quote_char(scanner)
+      break if match.nil?
+
+      quoted_string << match
+    end
+
+    match = scanner.scan(TOKEN_DOUBLE_QUOTE)
+    return quoted_string if match
+
+    nil
+  end
+  private :double_quoted_string
+
+  def quoted_string(scanner)
+    single_quoted_string(scanner) || double_quoted_string(scanner)
+  end
+  private :quoted_string
+
+  def array_values(scanner)
+    values = []
+
+    loop do
+      match = quoted_string(scanner)
+      break if match.nil?
+      values << match
+
+      match = array_separator(scanner)
+      break if match.nil?
+    end
+
+    values
+  end
+  private :array_values
+end
diff --git a/spec/unit/puppet_x/augeas/util/parser_spec.rb b/spec/unit/puppet_x/augeas/util/parser_spec.rb
new file mode 100644
index 0000000..f8b5b2b
--- /dev/null
+++ b/spec/unit/puppet_x/augeas/util/parser_spec.rb
@@ -0,0 +1,110 @@
+require 'spec_helper'
+require 'puppet_x/augeas/util/parser'
+
+describe PuppetX::Augeas::Util::Parser do
+  include described_class
+
+  it 'handles an empty array' do
+    expect(parse_to_array('[]')).to eq([])
+  end
+
+  it 'handles an array with a simple single-quoted entry' do
+    expect(parse_to_array("['entry']")).to eq(['entry'])
+  end
+
+  it 'handles an array with a simple double-quoted entry' do
+    expect(parse_to_array('["entry"]')).to eq(['entry'])
+  end
+
+  it 'handles an array with both single- and double-quoted entries' do
+    expect(parse_to_array(%q(['first', "second"]))).to eq(['first', 'second'])
+  end
+
+  context 'inside single-quoted strings' do
+    it 'allows a literal backslash' do
+      expect(parse_to_array("['entry\\\\here']")).to eq(['entry\\here'])
+    end
+
+    it 'allows an internal single-quote' do
+      expect(parse_to_array("['entry\\'here']")).to eq(['entry\'here'])
+    end
+  end
+
+  context 'inside double-quoted strings' do
+    it 'allows a literal backslash' do
+      expect(parse_to_array('["entry\\\\here"]')).to eq(['entry\\here'])
+    end
+
+    it 'allows an internal double-quote' do
+      expect(parse_to_array('["entry\\"here"]')).to eq(['entry"here'])
+    end
+
+    it 'does not require escaping a single-quote' do
+      expect(parse_to_array('["entry\'here"]')).to eq(["entry'here"])
+    end
+
+    it 'allows a bell character escape' do
+      expect(parse_to_array('["entry\\ahere"]')).to eq(["entry\ahere"])
+    end
+
+    it 'allows a backspace character escape' do
+      expect(parse_to_array('["entry\\bhere"]')).to eq(["entry\bhere"])
+    end
+
+    it 'allows a horizontal tab character escape' do
+      expect(parse_to_array('["entry\\there"]')).to eq(["entry\there"])
+    end
+
+    it 'allows a line feed character escape' do
+      expect(parse_to_array('["entry\\nhere"]')).to eq(["entry\nhere"])
+    end
+
+    it 'allows a vertical tab character escape' do
+      expect(parse_to_array('["entry\\vhere"]')).to eq(["entry\vhere"])
+    end
+
+    it 'allows a form feed character escape' do
+      expect(parse_to_array('["entry\\fhere"]')).to eq(["entry\fhere"])
+    end
+
+    it 'allows a carriage return character escape' do
+      expect(parse_to_array('["entry\\rhere"]')).to eq(["entry\rhere"])
+    end
+
+    it 'allows an escape character escape' do
+      expect(parse_to_array('["entry\\ehere"]')).to eq(["entry\ehere"])
+    end
+
+    it 'allows a space character escape' do
+      expect(parse_to_array('["entry\\shere"]')).to eq(['entry here'])
+    end
+
+    it 'allows octal character escapes' do
+      expect(parse_to_array('["\7", "\41", "\101", "\1411"]')).to eq(["\a", '!', 'A', 'a1'])
+    end
+
+    it 'allows hexadecimal character escapes with \\x' do
+      expect(parse_to_array('["\x7", "\x21", "\x211"]')).to eq(["\a", '!', '!1'])
+    end
+
+    it 'allows single-character unicode hexadecimal character escapes with \\u' do
+      expect(parse_to_array('["\u2015", "\u20222"]')).to eq(["\u2015", "\u2022" << '2'])
+    end
+
+    it 'allows multi-character unicode hexadecimal character escapes with \\u{...}' do
+      expect(parse_to_array('["\u{7}", "\u{20}", "\u{100}", "\u{2026}", "\u{1F464}", "\u{100000}", "\u{53 74 72 69 6E 67}"]')).to eq(["\a", ' ', "\u{100}", "\u{2026}", "\u{1F464}", "\u{100000}", 'String'])
+    end
+  end
+
+  it 'fails with garbage in front of the array' do
+    expect { parse_to_array("junk ['good', 'array', 'here']") }.to raise_error(RuntimeError, %r{^Unexpected character in array at: junk \['good})
+  end
+
+  it 'fails with garbage in the middle of the array' do
+    expect { parse_to_array("['got', 'some', junk 'here']") }.to raise_error(RuntimeError, %r{^Unexpected character in array at: junk 'here'})
+  end
+
+  it 'fails with garbage after the array' do
+    expect { parse_to_array("['good', 'array', 'here'] junk after") }.to raise_error(RuntimeError, %r{^Unexpected character in array at: junk after})
+  end
+end