# coding: binary
require_relative "rexml_test_utils"
require "rexml/document"
require "rexml/parseexception"
require "rexml/output"
require "rexml/source"
require "rexml/formatters/pretty"
require "rexml/undefinednamespaceexception"
require_relative "listener"
class Tester < Test::Unit::TestCase
include REXMLTestUtils
include REXML
def setup
@xsa_source = <<-EOL
Lars Marius Garshollarsga@garshol.priv.nohttp://www.stud.ifi.uio.no/~lmariusg/
EOL
end
def test_bad_markup
[
" foo ",
'<0/>',
'&',
'&a',
# '&a;', # FIXME
'',
'',
'',
'',
'',
'',
'',
"",
"\f",
"",
"\000",
# FIXME '',
'',
'',
# FIXME '',
# FIXME '',
].each do |src|
assert_raise( ParseException, %Q{Parse #{src.inspect} should have failed!} ) do
Document.new(src)
end
end
end
def test_attribute
# Testing constructors
#a = Attribute.new "hello", "dolly"
#b = Attribute.new a
#d = Document.new( "" )
#c = d[0].attributes.get_attribute( "hello" )
#assert_equal a, b
#for attr in [ a, b, c]
# assert_equal "hello", attr.name
# assert_equal "dolly", attr.value
#end
# This because of a reported bug in attribute handling in 1.0a8
source = 'blah'
doc = Document.new source
doc.elements.each do |a|
a.attributes['att'] << 'B'
assert_equal "AB", a.attributes['att']
a.attributes['att'] = 'C'
assert_equal "C", a.attributes['att']
end
# Bryan Murphy
text = "this is a {target[@name='test']/@value} test"
source = <<-EOL
EOL
xml = Document.new source
value = xml.root.attributes["search"]
assert_equal text, value.to_s
e = Element.new "test"
e.add_attributes({ "name1" => "test1", "name4" => "test4" })
e.add_attributes([["name3","test3"], ["name2","test2"]])
assert_equal "test1", e.attributes["name1"]
assert_equal "test2", e.attributes["name2"]
assert_equal "test3", e.attributes["name3"]
assert_equal "test4", e.attributes["name4"]
# ensure that the attributes come out in sorted order
assert_equal %w().join(' '), e.to_s
end
def test_cdata
test = "The quick brown fox jumped
& < & < \" '
over the lazy dog."
source = ""
d = REXML::Document.new( source )
# Test constructors
cdata = d[0][0]
assert_equal test, cdata.value
end
def test_comment
string = "This is a new comment!"
source = ""
comment = Comment.new string
REXML::Formatters::Default.new.write( comment, out = "" )
assert_equal(source, out)
comment2 = Comment.new comment
assert_equal(comment, comment2)
assert_raise(ParseException) {
REXML::Document.new("")
}
assert_raise(ParseException) {
REXML::Document.new("
EOF
doc = Document.new source
assert_equal "someuri", doc.root.namespace
assert_equal "default", doc.root.elements[1].namespace
assert_equal "someuri", doc.root.elements[2].namespace
assert_equal "notdefault", doc.root.elements[ 3 ].namespace
# Testing namespaces in attributes
source = <<-EOF
EOF
doc = Document.new source
b = doc.root.elements["b"]
assert_equal "x", b.attributes["b:a"]
assert_equal "y", b.attributes["a"]
doc = Document.new
doc.add_element "sean:blah"
doc.root.text = "Some text"
out = ""
doc.write(out)
assert_equal "Some text", out
end
def test_add_namespace
e = Element.new 'a'
e.add_namespace 'someuri'
e.add_namespace 'foo', 'otheruri'
e.add_namespace 'xmlns:bar', 'thirduri'
assert_equal 'someuri', e.attributes['xmlns']
assert_equal 'otheruri', e.attributes['xmlns:foo']
assert_equal 'thirduri', e.attributes['xmlns:bar']
end
def test_big_documentation
f = File.new(fixture_path("documentation.xml"))
d = Document.new f
assert_equal "Sean Russell", d.elements["documentation/head/author"].text.tr("\n\t", " ").squeeze(" ")
out = ""
d.write out
end
def test_tutorial
doc = Document.new File.new(fixture_path("tutorial.xml"))
out = ""
doc.write out
end
def test_stream
c = Listener.new
Document.parse_stream( File.new(fixture_path("documentation.xml")), c )
assert(c.ts, "Stream parsing apparantly didn't parse the whole file")
assert(c.te, "Stream parsing dropped end tag for documentation")
Document.parse_stream("", c)
Document.parse_stream("<>&", c)
assert_equal('<>&', c.normalize)
end
def test_line
Document.new File.new(fixture_path("bad.xml"))
assert_fail "There should have been an error"
rescue Exception
# We should get here
assert($!.line == 5, "Should have been an error on line 5, "+
"but was reported as being on line #{$!.line}" )
end
def test_substitution
val = "a'b\"c"
el = Element.new("a")
el.attributes["x"] = val
REXML::Formatters::Default.new.write(el, out="")
nel = Document.new( out)
assert_equal( val, nel.root.attributes["x"] )
end
def test_exception
source = SourceFactory.create_from ""
p = ParseException.new( "dummy message", source )
begin
raise "dummy"
rescue Exception
p.continued_exception = $!
end
end
def test_bad_content
in_gt = 'content>content'
in_lt = 'content'
# This is OK
tree_gt = Document.new in_gt
assert_equal "content>content", tree_gt.elements[1].text
# This isn't
begin
Document.new in_lt
assert_fail "Should have gotten a parse error"
rescue ParseException
end
end
def test_iso_8859_1_output_function
out = ""
output = Output.new( out )
koln_iso_8859_1 = "K\xF6ln"
koln_utf8 = "K\xc3\xb6ln"
source = Source.new( koln_iso_8859_1, 'iso-8859-1' )
results = source.scan(/.*/)[0]
koln_utf8.force_encoding('UTF-8') if koln_utf8.respond_to?(:force_encoding)
assert_equal koln_utf8, results
output << results
if koln_iso_8859_1.respond_to?(:force_encoding)
koln_iso_8859_1.force_encoding('ISO-8859-1')
end
assert_equal koln_iso_8859_1, out
end
def test_attributes_each
doc = Document.new("")
count = 0
doc.root.elements[1].attributes.each {|k,v| count += 1 }
assert_equal 4, count
end
def test_delete_namespace
doc = Document.new ""
doc.root.delete_namespace
doc.root.delete_namespace 'x'
assert_equal "", doc.to_s
end
def test_each_element_with_attribute
doc = Document.new ""
arry = []
block = proc { |e|
assert arry.include?(e.name)
arry.delete e.name
}
# Yields b, c, d
arry = %w{b c d}
doc.root.each_element_with_attribute( 'id', &block )
assert_equal 0, arry.size
# Yields b, d
arry = %w{b d}
doc.root.each_element_with_attribute( 'id', '1', &block )
assert_equal 0, arry.size
# Yields b
arry = ['b']
doc.root.each_element_with_attribute( 'id', '1', 1, &block )
assert_equal 0, arry.size
# Yields d
arry = ['d']
doc.root.each_element_with_attribute( 'id', '1', 0, 'd', &block )
assert_equal 0, arry.size
end
def test_each_element_with_text
doc = Document.new 'bbd'
arry = []
block = proc { |e|
assert arry.include?(e.name)
arry.delete e.name
}
# Yields b, c, d
arry = %w{b c d}
doc.root.each_element_with_text(&block)
assert_equal 0, arry.size
# Yields b, d
arry = %w{b c}
doc.root.each_element_with_text( 'b', &block )
assert_equal 0, arry.size
# Yields b
arry = ['b']
doc.root.each_element_with_text( 'b', 1, &block )
assert_equal 0, arry.size
# Yields d
arry = ['d']
doc.root.each_element_with_text( nil, 0, 'd', &block )
assert_equal 0, arry.size
end
def test_element_parse_stream
s = Source.new( "some text" )
l = Listener.new
class << l
def tag_start name, attributes
raise "Didn't find proper tag name" unless 'a'==name
end
end
Document::parse_stream(s, l)
end
def test_deep_clone
a = Document.new( 'texttexttext' )
b = a.deep_clone
assert_equal a.to_s, b.to_s
a = Document.new( 'some < text more > text > ' )
b = a.deep_clone
assert_equal a.to_s, b.to_s
c = Document.new( b.to_s )
assert_equal a.to_s, c.to_s
end
def test_whitespace_before_root
a = <
EOL
d = Document.new(a)
b = ""
d.write( b )
assert_equal a,b
end
def test_entities
a = Document.new( 'eeü' )
assert_equal('eeĆ¼'.force_encoding("UTF-8"), a.root.text)
end
def test_element_decl
element_decl = Source.new("
]>")
doc = Document.new( element_decl )
d = doc[0]
assert_equal("", d.to_s.split(/\n/)[1].strip)
end
def test_attlist_decl
doc = Document.new <<-EOL
]>
EOL
assert_equal 'gobble', doc.root.attributes['bar']
assert_equal 'xxx', doc.root.elements[2].namespace
assert_equal 'two', doc.root.elements[1].namespace
assert_equal 'foo', doc.root.namespace
doc = Document.new <<-EOL
]>
EOL
prefixes = doc.root.prefixes.sort
correct = ['svg', 'xlink', 'xml']
assert_equal correct, prefixes
end
def test_attlist_write
file=File.new(fixture_path("foo.xml"))
doc=Document.new file
out = ''
doc.write(out)
end
def test_more_namespaces
assert_raise( REXML::UndefinedNamespaceException,
%Q{Should have gotten an Undefined Namespace error} ) {
Document.new("
")
}
doc2 = Document.new("
")
es = XPath.match(doc2, '//c')
assert_equal 0, es.size
es = XPath.match(doc2, '//n:c')
assert_equal 1, es.size
doc2.root.add_namespace('m', '2')
doc2.root.add_element("m:o")
es = XPath.match(doc2, './/o')
assert_equal 0, es.size
es = XPath.match(doc2, '//n:c')
assert_equal 1, es.size
end
def test_ticket_51
doc = REXML::Document.new <<-EOL
XYZ
EOL
# The most common case. People not caring about the namespaces much.
assert_equal( "XY", XPath.match( doc, "/test/a/text()" ).join )
assert_equal( "XY", XPath.match( doc, "/test/x:a/text()" ).join )
# Surprising? I don't think so, if you believe my definition of the "common case"
assert_equal( "XYZ", XPath.match( doc, "//a/text()" ).join )
# These are the uncommon cases. Namespaces are actually important, so we define our own
# mappings, and pass them in.
assert_equal( "XY", XPath.match( doc, "/f:test/f:a/text()", { "f" => "1" } ).join )
# The namespaces are defined, and override the original mappings
assert_equal( "", XPath.match( doc, "/test/a/text()", { "f" => "1" } ).join )
assert_equal( "", XPath.match( doc, "/x:test/x:a/text()", { "f" => "1" } ).join )
assert_equal( "", XPath.match( doc, "//a/text()", { "f" => "1" } ).join )
end
def test_processing_instruction
d = Document.new("")
assert_equal 4, XPath.match(d, '//processing-instruction()' ).size
match = XPath.match(d, "//processing-instruction('foo3')" )
assert_equal 1, match.size
assert_equal 'bar3', match[0].content
end
def test_oses_with_bad_EOLs
Document.new("\n\n\n\n\n\n\n\n")
end
# Contributed (with patch to fix bug) by Kouhei
def test_ignore_whitespace
source = " abc "
context_all = {:ignore_whitespace_nodes => :all}
context_a = {:ignore_whitespace_nodes => %(a)}
context_b = {:ignore_whitespace_nodes => %(b)}
tests = [[[" abc ", "def"], context_all],
[[" abc ", "def"], context_a],
[[" ", " abc ", "def", " "], context_b]]
tests.each do |test|
assert_equal(test[0], Document.new(source, test[1]).root.texts.collect{|x|
x.to_s})
end
end
def test_0xD_in_preface
doc = "\x0D\x0D"
doc = Document.new doc
end
def test_hyphens_in_doctype
doc = REXML::Document.new <<-EOQ
EOQ
assert_equal('a-b-c', doc.doctype.name)
end
def test_accents
docs = [
%Q{
},
'
',
%Q{
},
%Q{
} ]
docs.each_with_index { |d,i|
begin
REXML::Document.new(d)
rescue
puts "#{i} => #{docs[i]}"
raise
end
}
end
def test_replace_text
e = REXML::Element.new( "a" )
e.add_text( "foo" )
assert_equal( "foo", e.to_s )
e[0].value = "bar"
assert_equal( "bar", e.to_s )
e[0].value = "<"
assert_equal( "<", e.to_s )
assert_equal( "<", e[0].value )
end
def test_write_doctype
## XML Document and Declaration
document = REXML::Document.new
xmldecl = REXML::XMLDecl.new("1.0", "UTF-8")
document.add(xmldecl)
s = ""
document.write(s)
## XML Doctype
str = ''
source = REXML::Source.new(str)
doctype = REXML::DocType.new(source)
document.add(doctype)
document.write(s)
## Element
element = REXML::Element.new("hoge")
document.add(element)
document.write(s)
end
def test_write_cdata
src = "A"
doc = REXML::Document.new( src )
out = ""
doc.write( out )
assert_equal( src, out )
src = ""
doc = REXML::Document.new( src )
out = ""
doc.write( out )
assert_equal( src, out )
end
def test_namespace_attributes
source = <<-EOL
EOL
d = Document.new( source )
assert_equal( 'foo', REXML::XPath.first(d.root, "//x:b/@x:n").value )
assert_equal( nil, REXML::XPath.first(d.root, "//x:b/@x:n", {}))
end
def test_null_element_name
a = REXML::Document.new
assert_raise( RuntimeError ) {
a.add_element( nil )
}
end
def test_text_raw
# From the REXML tutorial
# (http://www.germane-software.com/software/rexml/test/data/tutorial.html)
doc = Document.new <<-EOL
]>
EOL
a = doc.root
# This makes sure that RAW text nodes don't have their entity strings
# replaced
t = Text.new "Sean", false, nil, true
a.text = t
assert_equal( "Sean", t.to_s )
assert_equal( "Sean", t.value )
# This makes sure that they do
t = Text.new "Sean", false, nil, false
a.text = t
assert_equal( "&s;", t.to_s )
assert_equal( "Sean", t.value )
t = Text.new "&s;", false, nil, true
a.text = t
assert_equal( "&s;", t.to_s )
assert_equal( "Sean", t.value )
t = Text.new "&s;", false, nil, true
a.text = t
assert_equal( "&s;", t.to_s )
assert_equal( "Sean", t.value )
# Ticket #44
t = REXML::Text.new( "&", false, nil, true )
assert_equal( "&", t.to_s )
t = REXML::Text.new("&", false, false)
assert_equal( "&", t.to_s )
end
def test_to_xpath
doc = REXML::Document.new( %q{})
names = %w{ /tag1/tag2[1] /tag1/tag2[2] }
doc.root.elements.each_with_index {|el, i|
assert_equal( names[i], el.xpath )
}
end
def test_transitive
doc = REXML::Document.new( "")
s = ""
doc.write( s, 0, true )
end
# This is issue #40
def test_replace_with
old = 'oldold'
d = REXML::Document.new(old).root
new = REXML::Text.new('new',true,nil,true)
child = d.children[2]
child.replace_with(new)
assert_equal( new, d.children[2] )
end
def test_repeated_writes
a = IO.read(fixture_path("iso8859-1.xml"))
f = REXML::Formatters::Pretty.new
xmldoc = REXML::Document.new( a )
a_andre = xmldoc.elements['//image'].attributes['caption']
f.write(xmldoc,b="")
xmldoc = REXML::Document.new(b)
b_andre = xmldoc.elements['//image'].attributes['caption']
assert_equal( a_andre, b_andre )
f.write(xmldoc,c="")
xmldoc = REXML::Document.new(c)
c_andre = xmldoc.elements['//image'].attributes['caption']
assert_equal( b_andre, c_andre )
o = Output.new(d="","UTF-8")
f.write(xmldoc,o)
assert_not_equal( c, d )
end
def test_pretty_format_long_text_finite
n = 1_000_000
long_text = 'aaaa ' * n
xml = "#{long_text}"
formatter = REXML::Formatters::Pretty.new
document = nil
begin
document = REXML::Document.new(xml)
rescue REXML::ParseException
skip_message = "skip this test because we can't check Pretty#wrap " +
"works without # on " +
"small memory system. # " +
"will be raised on the system. See also [ruby-dev:42599]."
return skip_message
end
output = ""
formatter.write(document, output)
assert_equal("\n" +
((" " + (" aaaa" * 15) + "\n") * (n / 15)) +
" " + ("aaaa " * (n % 15)) + "\n" +
"",
output)
end
def test_pretty_format_deep_indent
n = 6
elements = ""
n.times do |i|
elements << ""
elements << "element#{i} " * 5
end
(n - 1).downto(0) do |i|
elements << ""
end
xml = "#{elements}"
document = REXML::Document.new(xml)
formatter = REXML::Formatters::Pretty.new
formatter.width = 20
output = ""
formatter.write(document, output)
assert_equal(<<-XML.strip, output)
element0
element0
element0
element0
element0\s
element1
element1
element1
element1
element1\s
element2
element2
element2
element2
element2\s
element3
element3
element3
element3
element3\s
element4
element4
element4
element4
element4
\s
element5 element5 element5 element5 element5\s
XML
end
def test_ticket_58
doc = REXML::Document.new
doc << REXML::XMLDecl.default
doc << REXML::Element.new("a")
str = ""
doc.write(str)
assert_equal("", str)
doc = REXML::Document.new
doc << REXML::XMLDecl.new("1.0", "UTF-8")
doc << REXML::Element.new("a")
str = ""
doc.write(str)
assert_equal("", str)
end
# Incomplete tags should generate an error
def test_ticket_53
assert_raise( REXML::ParseException ) {
REXML::Document.new( "" )
}
assert_raise( REXML::ParseException ) {
REXML::Document.new( "" )
}
assert_raise( REXML::ParseException ) {
REXML::Document.new( "" )
}
end
def test_ticket_52
source = ""
d = REXML::Document.new(source)
d.write(k="")
assert_equal( source, k )
source = ""
target = "\n \n"
d = REXML::Document.new(source)
REXML::Formatters::Pretty.new(4).write(d,k="")
assert_equal( target, k )
end
def test_ticket_76
src = "
at&t"
assert_raise( ParseException, %Q{"#{src}" is invalid XML} ) {
REXML::Document.new(src)
}
end
def test_ticket_21
src = ""
assert_raise( ParseException, "invalid XML should be caught" ) {
Document.new(src)
}
begin
Document.new(src)
rescue
assert_match( /missing attribute quote/, $!.message )
end
end
def test_ticket_63
Document.new(File.new(fixture_path("t63-1.xml")))
end
def test_ticket_75
d = REXML::Document.new(File.new(fixture_path("t75.xml")))
assert_equal("tree", d.root.name)
end
def test_ticket_48_part_II
f = REXML::Formatters::Pretty.new
#- rexml sanity check (bugs in ruby 1.8.4, ruby 1.8.6)
xmldoc = Document.new("")
xmldoc << XMLDecl.new(XMLDecl::DEFAULT_VERSION, "UTF-8")
content = ['61c3a927223c3e26'].pack("H*")
content.force_encoding('UTF-8') if content.respond_to?(:force_encoding)
#- is some UTF-8 text but just to make sure my editor won't magically convert..
xmldoc.root.add_attribute('attr', content)
f.write(xmldoc,out=[])
xmldoc = REXML::Document.new(out.join)
sanity1 = xmldoc.root.attributes['attr']
f.write(xmldoc,out=[])
xmldoc = REXML::Document.new(out.join)
sanity2 = xmldoc.root.attributes['attr']
f.write(xmldoc,out=[])
assert_equal( sanity1, sanity2 )
end
def test_ticket_88
doc = REXML::Document.new("")
assert_equal("", doc.to_s)
doc = REXML::Document.new("")
assert_equal("", doc.to_s)
end
def test_ticket_85
xml = <
ENDXML
yml = ""
# The pretty printer ignores all whitespace, anyway so output1 == output2
f = REXML::Formatters::Pretty.new( 2 )
d = Document.new( xml, :ignore_whitespace_nodes=>:all )
f.write( d, output1="" )
d = Document.new( xml )
f.write( d, output2="" )
# Output directives should override whitespace directives.
assert_equal( output1, output2 )
# The base case.
d = Document.new(yml)
f.write( d, output3="" )
assert_equal( output3.strip, output2.strip )
d = Document.new(yml)
f.write( d, output4="" )
assert_equal( output3.strip, output4.strip )
end
def test_ticket_91
source="great"
expected="great"
d = Document.new( source )
d.root.add_element( "bah" )
p=REXML::Formatters::Pretty.new(2)
p.compact = true # Don't add whitespace to text nodes unless necessary
p.write(d,out="")
assert_equal( expected, out )
end
def test_ticket_95
testd = REXML::Document.new ""
testd.write(out1="")
testd.elements["//c[2]"].xpath
testd.write(out2="")
assert_equal(out1,out2)
end
def test_ticket_102
doc = REXML::Document.new ''
assert_equal( "foo", doc.root.elements["item"].attribute("name","ns").to_s )
assert_equal( "item", doc.root.elements["item[@name='foo']"].name )
end
def test_ticket_14
# Per .2.5 Node Tests of XPath spec
assert_raise( REXML::UndefinedNamespaceException,
%Q{Should have gotten an Undefined Namespace error} ) {
Document.new("")
}
end
# 5.7 Text Nodes
# Character data is grouped into text nodes. As much character data as
# possible is grouped into each text node: a text node never has an
# immediately following or preceding sibling that is a text node. The
# string-value of a text node is the character data. A text node always has
# at least one character of data.
def test_ticket_105
d = Document.new("")
d.root.add_text( "a" )
d.root.add_text( "b" )
assert_equal( 1, d.root.children.size )
end
# phantom namespace same as default namespace
def test_ticket_121
doc = REXML::Document.new(
'text'
)
assert_equal 'text', doc.text( "/doc/item[@name='foo']" )
assert_equal "name='foo'",
doc.root.elements["item"].attribute("name", "ns").inspect
assert_equal "text",
doc.root.elements["item[@name='foo']"].to_s
end
def test_ticket_135
bean_element = REXML::Element.new("bean")
textToAdd = "(&(|(memberof=CN=somegroupabcdefgh,OU=OUsucks,DC=hookemhorns,DC=com)(mail=*someco.com))(acct=%u)(!(extraparameter:2.2.222.222222.2.2.222:=2)))"
bean_element.add_element("prop", {"key"=> "filter"}).add_text(textToAdd)
doc = REXML::Document.new
doc.add_element(bean_element)
REXML::Formatters::Pretty.new(3).write( doc, out = "" )
assert_equal "\n \n (&(|(memberof=CN=somegroupabcdefgh,OU=OUsucks,DC=hookemhorns,DC=com)(mail=*someco.com))(acct=%u)(!(extraparameter:2.2.222.222222.2.2.222:=2)))\n \n", out
end
def test_ticket_138
doc = REXML::Document.new(
''
)
expected = {
"inkscape" => attribute("xmlns:inkscape",
"http://www.inkscape.org/namespaces/inkscape"),
"version" => {
"inkscape" => attribute("inkscape:version", "0.44"),
"" => attribute("version", "1.0"),
},
}
assert_equal(expected, doc.root.attributes)
assert_equal(expected, REXML::Document.new(doc.root.to_s).root.attributes)
end
def test_empty_doc
assert(REXML::Document.new('').children.empty?)
end
private
def attribute(name, value)
REXML::Attribute.new(name, value)
end
end