Module: RightDevelop::CI::Util

Defined in:
lib/right_develop/ci/util.rb

Constant Summary collapse

JAVA_CLASS_NAME =

Regular expression used to determine which characters of a string are allowed in Java class names.

/[A-Za-z0-9_]/
JAVA_PACKAGE_SEPARATOR =

The dot character gets special treatment: even if we XML-escape it, Jenkins will assume that it functions as a package separator. So, we’ll replace it with an equivalent Unicode character. Hooray homographic character attacks!

'.'
JAVE_PACKAGE_SEPARATOR_HOMOGLYPH =

Replacement codepoint that looks a bit like a period

'·'
INVALID_CDATA_CHARACTER =

Regular expression that matches characters that need to be escaped inside CDATA c.f. www.w3.org/TR/xml11/#charsets RestrictedChar ::= [#x1-#x8] | [#xB-#xC] | [#xE-#x1F] | [#x7F-#x84] | [#x86-#x9F]

Regexp.new '[\x01-\x08\x0b-\x0c\x0e-\x1f\x7f-\x84\x86-\x9f]', nil, 'n'

Class Method Summary collapse

Class Method Details

.pseudo_java_class_name(name) ⇒ String

Make a string suitable for parsing by Jenkins JUnit display plugin by escaping any non-valid Java class name characters as an XML entity. This prevents Jenkins from interpreting “hi1.2” as a package-and-class name.



32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
 
# File 'lib/right_develop/ci/util.rb', line 32

def pseudo_java_class_name(name)
  result = ''

  name.each_char do |chr|
    if chr =~ JAVA_CLASS_NAME
      result << chr
    elsif chr == JAVA_PACKAGE_SEPARATOR
      result << JAVE_PACKAGE_SEPARATOR_HOMOGLYPH
    else
      chr = chr.unpack('U')[0].to_s(16)
      result << "&#x#{chr};"
    end
  end

  result
end

.purify(untrusted) ⇒ String

Strip invalid UTF-8 sequences from a string and entity-escape any character that can’t legally appear inside XML CDATA. If test output contains weird data, we could end up generating invalid JUnit XML which will choke Java. Preserve the purity of essence of our precious XML fluids!



56
57
58
59
60
61
62
63
64
65
66
67
68
69
 
# File 'lib/right_develop/ci/util.rb', line 56

def purify(untrusted)
  # First pass: strip bad UTF-8 characters
  if RUBY_VERSION =~ /^1\.8/
    iconv = Iconv.new('UTF-8//IGNORE', 'UTF-8')
    result = iconv.iconv(untrusted)
  else
    result = untrusted.force_encoding(Encoding::BINARY).encode('UTF-8', :undef=>:replace, :replace=>'')
  end

  # Second pass: entity escape characters that can't appear in XML CDATA.
  result.gsub(INVALID_CDATA_CHARACTER) do |ch|
    "&#x%s;" % [ch.unpack('H*').first]
  end
end