The Perl Advent Calendar needs more articles for 2022. Submit your idea today!

3.52 - 2016-11-23  - minor maintenance release

- fixed: the previous fix was buggy...

3.51 - 2016-11-23  - minor maintenance release

- fixed: failing tests when XML::XPathEngine and XML::XPath not available

3.50 - 2016-11-22  - minor maintenance release

- added: the no_xxe option to XML::Twig::new, which causes the parse 
  to fail if external entities are used (to prevent malicious XML to
  access the filesystem). 
  See RT#118097

- fixed: warning (and soon error) due to unescaped literal left braces
  in regular expressions in the code generating
  reported by trwyant

- fixed: (partial fix) implement getNamespaces in XML::Twig::XPath::Elt
  the expression doesn't crash the code, but doesn't return anything
  interesting (yet)
  reported by Nathan Glenn

- fixed: various spelling mistakes
  thanks to James McCoy for the patch

- git repo cleanup, thanks to mjg17

3.49 - 2015-04-12  - minor maintenance release

- added: the DTD_base option to XML::Twig new, that forces XML::Twig to look
  for the DTD in a given directory
  thanks to Arun lakhana for the idea

- Prevent PAUSE from trying to index packages that are only used for monkey
  patching (to re-use XML::XPath as the XPath engine for XML::Twig::XPath).
  Will also prevent UNAUTHORIZED flag on metacpan.
  patch sent by Graham Knop

- fixed: RT # 96009
  keep_atts_order => 0 behaviour. Spotted by Dolmen

- fixed: bug RT #97461
  wrong error message was returned calling parse on an invalid filehandle
  Thanks to Slaven Rezic for the bug report and test case  

  fixed: bug RT #
  inconsistency between simplify and XML::Simple for empty elements (including
  elements with start and end tags but no contents)
  the XML::Simple behaviour is to map them to an empty hash, not an 
  empty/undef scalar (depending of whether the element is a PCDATA or not)
  as was the case in previous versions of the module. 
  This has the potential to break some existing code, but simplify should be
  strictly the same as XML::Simple's XMLin
  Thanks to Vangelis Katsikaros for the bug report and test case

3.48 - 2014-03-30 - minor maintenance release

- fixed: tests

3.47 - 2014-03-27 - minor maintenance release

- fixed: missing entities when parsing HTML
  RT #93604

- fixed: tests failed when using a version of HTML::TreeBuilder with a non-numeric version

- fixed in twig_handlers, '=' in regexps on attributes are turned into 'eq'
  RT #94295

3.46 - 2014-03-05 - minor maintenance release

- fixed: test failed on Windows

3.45 - 2014-02-27 - minor maintenance release

- fixed: link to idented_a format description
  RT #85400
  fixed by Martin McGrath

- fixed: code that gave a warning in 5.19.9

- fixed: RT #86651
  xml_pp, quote not escaped in attribute values

- fixed: various typos in docs RT#87660
  thanks to David Steinbrunner

- fixed: RT #86773 
  CDATA sections in HTML were not properly escaped when using the
  (default) HTML::TreeBuilder conversion
  spotted by Marco Pessotto

- fixed: RT #85933
  quotes in attributes were not properly escaped
  spotted by Arun Lakhana

- added: docs for tools and safe_print_to_file

- added: support for XPath variables
  thanks to Nathan Glenn for the initial implementation

- updated: Changes to conform to CPAN::Changes + test

3.44 - 2013-05-13 - minor maintenance release

- added: XML::Twig::Elt new method now acccepts literal content, eg
  my $e= XML::Twig::Elt->new( '<div><p>foo</p><p>bar</p></div>');

- added: twig handler triggers now accept the syntax <tag>#<id>
  use *#<id> if you don't want to specify the tag name (#<id> would not
  work since this is the syntax for "private" elements, this makes it
  ugly, but is due to the fact that when I started working on XML::Twig
  CSS wasn't really around)

- fixed: merge had some problems dealing with embedded comments

- improved: more tests

- improved: make Changes conform to the CPAN::Changes spec

3.43 - 2013-02-31 - minor maintenance release

- improved: docs for parse, see RT #78877

- fixed: xml_pp -i now preserves the permissions of the
  original file, see RT #81165 
  reported by Alberto Simoes

- fixed: RT #80503 Newlines in attribute values
  reported (and explained) by Ambrus Zsban: \r, \n
  and \n explicitely set in attribute values should
  be escaped (with &#x<nb>;) when output 

3.42 - 2012-11-06 - minor maintenance release

- fixed: bug, elements created with XML::Twig::Elt->parse were 
  garbage collected too early, 

- added: some tests

3.41 - 2012-08-08 - minor maintenance release

- fixed: META.json generation

3.40 - 2012-05-10 - minor maintenance release

- added: support for alternations ('|') at the top level of handler
  triggers and navigation
  you can now have twig_handlers => { 't1|t2' => \&handler }
  and $elt->children( 't1|t2')

- added: the discard_all_spaces option that discards more aggressively
  non-significant white spaces
  see RT #71164 

- fixed: used $[ instead of $] in 3 tests,
  see RT#72765
  reported by Kevin Ryde

- fixed: did not use Text::Wrap with indented_c
  see RT #71375
  reported and patch provided by Martin Str?mberg 

- fixed: doc change for XML::Twig::Elt flush, see RT #72279

- fixed: replaced HTML::TreeBuilder::as_XML with am XML::Twig specific
  version, to avoid bugs in the original version and improve
  stability of the output

3.39 - 2011-09-22 - minor maintenance release

- fixed: xml_pp -i would blank all files after the first one 
  thanks to dvercande for spotting this 

- added: findvalues method (XML::Twig and XML::Twig::Elt)
  same as findvalue except that it returns an array of value

- added: the output_html_doctype option to XML::Twig::new, that
  outputs the DOCTYPE declaration for HTML docs converted
  by HTML::TreeBuilder (fixing it if necessary)
  see RT #71009:

- fixed: t/test_autoencoding_conversion.t failed with $PERL_UNICODE
  set to SA* (which prevents autoconversion)
  reported by Martin J Evans, RT #71084 

3.38 - 2011-02-27 - minor maintenance release

- fixed: RT 65865: _ should be allowed at the start on an XML name
  reported by Steve Prokopowich

3.37 - 2010-10-08 - minor maintenance release

- fixed: more tests fixed for HTML::TreeBuilder, hopefully 
  will pass now

- changed: making att and class lvalues created problems: in certain
  context they made regular calls to the method create empty 
  attributes. I could find no satisfactory fix,they were either
  incompletes, or to complex for often used methods. So att and
  class are back to being regular, non l-value methods. 
  latt and lclass are the l-value versions.

- added: documented the -html option for xml_grep, that allows processing
  HTML input

- added: the -Tidy option to xml_grep, that uses HTML::Tidy to convert

3.36 - 2010-10-07 - minor maintenance release

- added: the use_tidy option to XML::Twig->new, which uses
  HTML::Tidy to convert HTML to well-formed XHTML, as an
  alternative to the default conversion which uses

- added: XML::Twig::Elt method siblings which returns the
  siblings of the element

- added: methods att_accessors, elt_accessors and field_accessor
  as well as the similarly named options when creating an 

- added: set_outer_xml XML::Twig::Elt method

- added: print_to_file on an XML::Twig::Elt

- added: can use the tag[nested] form in twig handlers that
  triggers on elements 'tag' that include a child 'nested'

- added: aliased the add_to_class XML::Twig::Elt method to add_class,
  which seems more natural

- added: the remove_class method

- added: made att and class lvalues (in perl 5.6 and up)

- fixed: copy did not copy the empty status of an element
  RT#31664 spotted by Roland Minner

- fixed: cut_children would always set the empty status of an element,
  even if it had children left

- fixed: tests did not pass with HTML::TreeBuilder 3.23_1 due to a
  change in an error message

3.35 - 2010-05-15 - minor maintenance release

- added: the by_file option to xml_grep that limits
  the number of hits per file

- added: allowed the text of ignored elements to be buffered
  in a string

- fixed: comments need to be escaped (you can't have 2 '-' in a
  row), RT#57389 spotted by Konstantin Tchernov

- fixed: after $elt->cut_children, $elt->empty is false RT#54570
  spotted and patched by Andrew Pimlott

- fixed: documented the fact that latin1 is ISO-8859-15, see RT#37431

3.34 - 2010-01-18 - minor maintenance release, test suite fixes

- fixed: tests failed when XML::XPath was used as the XPath engine

3.33 - 2010-01-15 - minor maintenance release, bug fixes

- added: XML::Twig::Elt method att_exists which returns true if the attribute
  exists in the XML

- added: XML::Twig::Elt method lc_attnames which lower cases the names
  of all the attribute of the element

- added: better error message if find_nodes or get_xpath are called instead
  of findnodes when using XML::Twig::XPath (suggested by Zed Pobre)

- added: indented_close_tag pretty_print option (suggested by H.Merijn Brand)

- added: RT #49692 xml_split test on win 32 systems. Patch sent through RT

- added: using position selector (eg foo[2]) in handler triggers now raises 
  an error, spotted by Selvakumar

- added: you can use css like selectors for class in navigation: 'p.title' will
  select p elements with a class that contains title. 
  In order to preserve backward compatibility and to allow the use of
  elements with a dot in their name, if there are already parsed elements
  with a tag name of 'p.title' then they will be selected instead

- added: you can also use css class selectors in trigger handlers. 

- fixed:  avoids expat (and XML::Parser) "Ran out of memory for input buffer"
  error, and instead reports an "empty file" error (and does not attempt
  to parse the file).

- fixed:  RT #51432 attributes containing quote character don't escape properly
  found, and patch provided by Jeremy Kahn

- fixed:  RT #48616 handler condition of foo/* crashed the module
  reported by Osfameron

- fixed:  xml_grep bug: warning when --count is used and no match is found 
  found by Hermann Peifer

- fixed:  xml_split bug when using an XML declaration and a utf8 encoding. Spotted
  by Chris Price.

- fixed:  xml_pp bug, pod2text command to display help was not properly quoted.
  Spotted by Chris Price. 

- fixed:  failing tests when LWP::UserAgent is not available

- fixed:  RT #41147: use of uninitialized value in eval when attribute isn't found
  reported by Zed Pobre

- fixed:  memory leak when the XML included id's

- fixed:  XML::Twig::Elt->set_content fails when argument is 'XML::Twig::Elt'
  (or the name of a subclass of XML::Twig::Elt)

- fixed:  bug RT #39849, set_output_encoding( 'utf-8') did not work quite right
  on filehandles that were already open in >:utf-8 mode
  spotted by Zed Pobre

- fixed:  xml_pp now accepts all formating options available in XML::Twig

- fixed:  RT #31664, element attributes are not preserving their order when
  using elt->copy spotted, and fix provided by jbubbabrown

- fixed:  RT #31832, wrapped link to in L< > tag in the doc
  spotted by Slaven Srezic

- fixed:  RT #31833 doc fix, spotted by Slaven Srezic

- fixed:  Makefile.PL doesn't nag the poor tester anymore when running with

- fixed:  bug calling set_text when using XML::Twig::XPath, spotted by Ted Sung

- fixed:  improved speed when parsing big elements, RT#35672, reported by Seth 
  Viebrock (fi is to explicitely return null from the character handler,
  instead of the text already parsed... a few hundred thousand times)

- fixed:  RT #47257, minor doc bug, spotted by David Steinbrunner

- fixed:  bug in navigation conditions of the form elt[text()=~ /text with 'or' or 'and'/] 

- improved: speed, somewhat

-improved: put the project on github:

3.32 - 2007-11-13 - minor maintenance release with a bug fix

- fixed:  change to the regexp that parses XPath-like conditions so
  it can accept leading non-ascii letters ([^\W\d] does not 
  work), not used in perl 5.005

- fixed:  set use utf8 (except in 5.005), which gets rid of the dreaded
  "SWASHNEW" error in 5.6.*, fixed things that then broke in 5.6.

3.31 - 2007-11-07 - minor maintenance release, fixing some tests

- fixed: fixes to stop tests from failing in various configurations

3.30 - 2007-11-06 

- fixed: a couple of bugs in namespace handling, spotted by
  Shlomo Yonas (see

- added: the XML::Twig::Elt fields method which returns a list of

- added: the normalize method in XML::Twig and XML::Twig::Elt,
  which merge together consecutive pcdata elements. As much as 
  possible (so far after a cut, delete or erase), the twig is 
  kept normalized, eg there are no consecutive #PCDATA elements
  in it. Suggestion of someone whose name (and emails) I can't
  find at the moment.

- added: the indented_a / cvs format for pretty_print, that makes the
  output friendly to line-oriented version control tools, as described
  in (RT #24954). Thanks to Sjur Moshagen
  for a patch that I adapted to the current version.

- fixed: bug RT #25113: system entities were not properly resolved
  if the XML file was not in the current directory. Thanks to
  Dave Charness for the patch.

- added: the XML::Twig method finish_now that terminates parsing
  immediately, without checking the rest of the XML. This feature was
  half suggested by Nick Clayton

- added: the -s option to xml_split, which splits when the given 
  size is reached for a file, suggested by Radek Saturka.

- added: the -g option to xml_split, which groups elements to be
  split, suggested and tested by Dhirendra Singh Kholia.

- added: the safe_parsefile_html and safe_parseurl_html methods,
  and a --html option to xml_grep. Suggested by Bill Ricker.

- improved: by default xml_grep now skips non well-formed files, the 
  --strict option makes it die when it finds one
- fixed: a bunch of bugs in xml_grep
- fixed: a warning when using optional modules with a version
  number that includes an _, spotted and fix suggested by 
  Bill Ricker.

- fixed: test failure on cygwin, thanks to Erik Rantapaa for the

- fixed: a bunch of typos in docs, RT #25836, spotted and fixed by David

- improved: re-use of XML::Twig objects for repetitive parsing. It 
  looks like it should be OK now , but I am sure I haven't tested
  all cases yet (especially when DTDs and entities are involved).

- improved: HTML parsing; XML::Twig now tries to find the proper
  encoding for the document (that's not done by HTML::TreeBuilder
  at the moment).

-fixed: XML::Twig::Elt purge and flush methods now only purge/flush up to
  the element, not up to the current element in the twig (duh!)

- fixed: bug in handlers of the form elt[string(subelt)="foo"] and
  elt[string(subelt)=1] which did not work at all

- fixed: bug in parameter entity output, spotted by BenHopkins on
  perlmonks (see

- fixed: bug in xml_string: options were not used
- improved error reporting for missing SYSTEM entities, including
  the option to set twig_expand_external_ents to -1, which makes
  missing SYSTEM entities not fatal, but reports them in 
  $t->{twig_missing_system_entities} Thanks to Frank Wegmann for 
  his suggestions and for testing the various versions of the feature

 - fixed: internals so new versions of Pod::Coverage won't barf
3.29 - 2007-01-22 

 - fixed: a bug in the handling of handlers after an ignore (RT #24392, 
  reported by Robert Eden).

3.28 - 2007-01-05 

 - now builds on Windows and OS2

 - improved: refactored the code that triggers handlers, 
  more complex expressions can now be handled,
  such as '/doc/section[@def="1"]/title'

  Up to version 3.26, you could change the attribute
  of a parent of a node on which you had a handler,
  and be able to trigger a handler on that parent node
  based on the new attribute value:

      XML::Twig->new( twig_handlers =>
        { 'sect/title'          => sub { $_->parent->set_att( has_title => 1)},
          'sect[@has_title="1"]'=> sub { ... }, # called for any sect that has
        }                                       # a title

  This won't work now. The trigger expression ('sect[@has_title="1"]')
  is evaluated strictly against the input XML. This is more logical and
  consistent (if you changed the element name, the new name was never
  used in the evaluation of the trigger).

  The only exception to that rule is if you use "private attributes": 
  attributes which name starts with a '#'. By definition this in an invalid
  XML name, so it can't be in the input, and has to have been created . In 
  that case the code that evaluates the trigger looks at the attribute in 
  the element in the tree in memory (if it exists). 

  So in the example above, if you replace 'has_title' by '#has_title', 
  everything will work fine. Note that private attributes are not output 
  when using the print/sprint/xml_string... methods.    
- fixed: xml_pp so it does not leave a tempfile
  and a broken original file all when the original
  file is not well-formed.
- added: the nparse_pp method that does an nparse
  with pretty_print set to 'indented', nparse_e
  that sets error_context, and nparse_ppe that 
  does both
- added: XML::Twig::Elt tag_to_span and tag_to_div
  methods (turn an element into a span/div and 
  set its class to the old tag name)

- added: the quote option for XML::Twig new, which 
  sets the output quote character for attributes
  ('single' or 'double')

- added: the text_only and xml_text_only methods
  that return the text of the element, but not of 
  the sub-elements. 

- added: outer_xml method (synonym for sprint)

- fixed: bug where entity names were not matched
  properly (RT #22854, spotted by Bob Faist)

- fixed: bug  on some DOCTYPE config with 
- fixed: bug in set_keep_encoding (the method, 
  not the option).

- fixed: bug in simplify: the code attempted to
  replace variables in attribute values even if no
  option required it, spotted by Klaus Rush

- improved: clean-up and fixed bugs in ignore: the method
  can now be called from a regular handler (it
  always could but the docs did not say so, 
  thanks to kudra for noticing this). It can
  also be called to ignore a parent of the current 
  element. There were bugs there, and the tree 
  was not built properly

- added: error message when an XPath query with
  a leading / is used on a node that does not
  belong to a whole twig (because it's been cut
  or because the twig itself went out of scope)

- improved: when parsing HTML with error_context set, the
  HTML is indented, in order to give better error

3.26 - 2006-07-01 

- added: argument to -i in the Makefile to prevent
  problem in win32

- added: XML::Twig::Elt former_next_sibling, 
  former_prev_sibling and former_parent methods

- squashed a memory leak when parsing html
  (forgot to call delete on the HTML::Tree object)

 - fixed: bug that caused XML::Twig to hang if
  there was a syntax error in a predicate
  (RT#19499, reported by Dan Dascalescu)

 -improved: made start_tag and end_tag more consistent: they
  now both return the empty string for comments,
  PIs... (reported by Dan Dascalescu)

- added: parsefile_inplace and parsefile_html_inplace
  methods (thanks to GrandFather on perlmonks)

- added: support to add css stylesheet in the
  add_stylesheet method (thanks to Georgi Sotirov)

- patched tests to work on Win32

 - added: set_inner_xml inner_xml and set_inner_html

3.25 - 2006-05-10 

 - patched to work with perl 5.005!

- fixed: a bug in xml_pp when pretty printing a 
  file in place in a different file system

3.24 - 2006-05-09 

- added: loading the text of entities stored in 
  separate files (using SYSTEM) when the (awfully
  named!) expand_external_ents option is used.
  Thanks to jhx for spotting this.

- changed: set_cdata, set_pi and set_comment so that
  if you call them on an element of the wrong kind,
  everything works as expected, instead of swallowing
  silently the data. Bug spotted by cmccutcheon

- fixed: a whole bunch of things to make the module 
  run and the tests pass on VMS, thanks to Peter 
  (Stig) Edwards who reported bug RT #18655 and 
  provided a patch.

- fixed: bug on get_xpath( '/root[1]') expressions,
  RT #18789 spotted by memfrob.

- added: the add_stylesheet method, that... adds a 
  stylesheet (xsl type is supported, let me know if
  other types are needed) to a document.

- improved: allowed pasting PI/Comment elements before or after
  the root of a document (see discussion at 
  Thanks to rogue90 for noticing the problem, and to
  Tanktalus for finding the best way to solve it.

- added: aliased unwrap to erase (ie added the unwrap method
  to XML::Twig::Elt, identical to the existing erase)
  suggested by Chris Burbridge.

- fixed: bug RT #17522: flushing twice at the end of 
  the the parse would output the last fragment twice. 
  Spotted by Harco de Hilster.

- fixed: bug RT #17500: parsing a pipe when using
  the UTF8 perlIO layer (through PERL_UNICODE or -C)
  now raises an error, found by Nikolaus Rath.

cwimproved: made the tests pass when the UTF8 perlIO layer is
  used. At this point potential problems when parsing 
  non-UTF8 XML in this configuration are not trapped.

3.23 - 2006-01-23 

- added: autoflush: there is no more need for the
  last $twig->flush after the parsing, it is done
  automatically at the end of the parsing, with the
  same arguments as the first flush on the twig.
  This can be turned of by setting $twig->{twig_autoflush}
  to 0.
  WARNING: if you finished the output with a direct
  print instead of a flush, then this change will 
  cause a bug. Hopefully this should not be the case
  and is easily fixable. 

- fixed: bug RT #17145 where get_xpath('//root/elt[1]/child')
  would produce a fatal error if there were no elt 
  element under root. Spotted by Dan Dascalescu.

- fixed: bug RT #17064 (comments and PIs after the
  root element were not properly processed), spotted
  by Dan Dascalescu.
- fixed: bug RT #17044: the SYSTEM value was not
  output in UpdateDTD mode, thanks to Michal
  Lewandowski for pointing this out.

- changed: the way empty tags are expanded with the
  'html' style: only tags that are allowed to be 
  empty in XHTML are output as '<tag />', thanks
  to Tom Rathborne for proding me to look into this.

- added: a 'wrapped' pretty_print option, that is
  a bit dodgy I think but that might please some.

- fixed: bug RT #16540 (tags with specific names 
  (like 'level'), tripped XML::Twig, spotted by
- added: comparison with XML::LibXML in the SEE ALSO
  section (and in the FAQ), following a question
  from surf on c.l.p.m
- added: XML::Twig now rejects string/regexp condition
  in twig_roots
- added: better error checking in xml_grep
- fixed: string/regexp condition in xml_grep 

- added: support for ! @att (or not @att) in get_xpath

- added: support for several predicates in get_xpath
  (not nested predicates though).

- fixed: bug RT #15671 (wrong condition interpretation
  for attribute value 0)

- added: XML::Twig print_to_file method

- added: XML::Twig::Elt methods: following_elt,
  following_elts, preceding_elt, preceding_elts
  (needed to support the corresponding axis in

3.22 - 2005-10-14 

- added: the XML::Twig xparse method, which parses
  whatever is thrown at it (filehandle, string, 
  HTML file, HTML URL, URL or file).

- added: the XML::Twig nparse method, which creates
  a twig and then calls xparse on the last parameter.

- added: the parse_html and parsefile_html methods,
  which parse HTML strings (or fh) and files 
  respectively, with the help of HTML::TreeBuilder.
  the implementation may still change. Note that
  at the moment there seems to be encoding problems
  with it (if the input is not UTF8).

- added: info to t/zz_dump_config.t

- fixed: a bug that caused subs_text to leave empty
  #PCDATA elements if the regexp matched at the beginning
  or at the end of the text of an element.

- fixed: RT #15014: in a few methods objects were 
  created as XML::Twig::Elt, instead of in the class?!F
  of the calling object.

- fixed: RT #14959: problem with wrap_children when
  an attribute of one of the child element includes
  a '>'

- improved: the docs for wrap_children

- added: a better error message when re-using an
  existing twig during the parse

- fixed: (partially) a bug with windows line-endings in
  CDATA sections with keep_encoding set (RT #14815)

- added: Test::Pod::Coverage test to please the kwalitee 
  police ;--)

3.21 - 2005-08-12 

- fixed: a test that failed if Tie::IxHash was not

- added: link to Atom feed for the CPAN testers 
  results at

3.20 - 2005-08-11 

- fixed: the pod (which caused the tests to fail)   

3.19 - 2005-08-10 

- fixed: the fix to RT # 14008, this one should be ok
  restructured tests

- added: the _dump method (probably not finished)

3.18 - 2005-08-08 

- added: a fix to deal with a bug in XML::Parser in the
  original_string method when used in CDATA sections
  longer than 1024 chars (RT # 14008) thanks to Dan 
  Dascalescu for spotting the bug and providing a test 

- added: better error diagnostics when the wrong arguments
  are used in paste

- fixed: a bug in subs_text when the text of an element
  included \n (RT #13665) spotted by Dan Dascalescu

- improved: cleaned up the behaviour of erase when the element
  being erased has extra_data (comments or pis) attached

- fixed: a bug in subs_text that sometimes messed up text
  after the matching text

- fixed: the erase/group_tags option of simplify to make
  it exactly similar to XML::Simple's

- fixed: a bug that caused XML::Twig to crash when ignore
  was used with twig_roots (RT #13382) spotted by Larry 

- fixed: bug in xml_split with default entities (they
  ended up being doubly escaped)

- fixed: various bugs when dealing with ids (changing
  existing ids, setting the attribute directly...)

- improved mark and split, both methods  now accepts several
  tags/ as arguments, so you can write for example:
   $elt->mark( qr/^(\w+): (.*)$/, 'dt', 'dd');
 - added: XML::Twig::Elt children_trimmed_text method,
  patch sent by ambrus (RT #12510)

 - changed: children_text and children_trimmed_text to
  have them return the entire text in scalar context

 - fixed: bug that caused XML::Twig not to play nice with
  XML::Xerces (due to improper import of UNIVERSAL::isa)
  spotted and patched by Colin Robertson.

- changed: most references to 'gi' in the docs, replaced 
  them by tag. I guess Robin Berjon's relentless teasing
  is to be credited with this one.

- added: tag_regexp condition on handlers (a regexp instead
  of a regular condition will trigger the handler if the
  tag matches), suggested by Franck Porcher, implementation
  helped by a few Perl Monks 

- fixed: typos in xml_split (RT #11911 and #11911),
  reported by Alexey Tourbin

- added: tests for xml_split and xml_merge and fixed
  a few bugs in the process

- added: the -i option to xml_split and xml_merge,
  that use XInclude instead of PIs (preliminary
  support, the XInclude namespace is not declared
  for example).

- added the XML::Twig and XML::Twig::Elt trim method
  that trims an element in-place

-added the XML::Twig last_elt method and the XML::Twig::Elt
  last_descendant method

- added: more tests

3.17 - 2005-03-16 

- improved: documentation, mostly to point better to 
  the resources at
-fixed: a few tests that would fail under perl 5.6.*
  and Solaris (t/test_safe_encode.t and t/test_bug_3.15.t),
  see RT bug # 11844, thanks to Sven Neuhaus

- changed: the licensing terms in the README to match the
  ones in the main module (same as Perl), see RT bug #11725

- added: a test on XML::SAX::Writer version number to 
  avoid failing tests with old versions (<0.39)

- improved: xml_split

3.16 - 2005-02-11 

- added: the xml_split/xml_merge tools

- fixed: PI handler behaviour when used in twig_roots mode

- fixed: bug  that prevented the DTD to be output
  when update_DTD mode is on, no DTD is present but
  entities have been created

- added: level(<n>) trigger for handlers

- fixed: bug that prevented the output_filter to be
  called when printing an element. Spotted thanks to
  Louis Strous.
- fixed: bug in the nsgmls pretty printer that output
  invalid XML (an extra \n was added in the end tag)
  found by Lee Goddard

- fixed: test 284 in test_additional to make it pass 
  in RedHat's version of perl 5.8.0, thanks to
  rdhayes for debugging and fixing that test.

- improved: first shot at getting Pis and comments back in the
  proper place, even in 'keep' mode. At the moment
  using set_pcdata  (or set_cdata) removes all
  embedded comments/pis

- fixed: a bug with pi's in keep mode (pi's would not
  be copied if they were within an element) found by
  Pascal Sternis

- added: a fix to get rid of spurious warnings, sent
  by Anthony Persaud 

- added: the remove_cdata option to the XML::Twig new
  method, that will output CDATA sections as regular
  (escaped) PCDATA

- added: the index option to the XML::Twig new method,
  and the associated XML::Twig index method, which 
  generates a list of element matching a condition 
  during parsing

- added: the XML::Twig::Elt first_descendant method

- fixed: bug with the keep_encoding option where
  attributes were not parsed when the element name was
  followed by more than one space (spotted by Gerald

- fixed: a bug where whitespace at the beginning of an
  element could be dropped (if followed by an element
  before any other character). Now whitespace is
  dropped only if it includes a \n

- added: feature: when load_DTD is used, default
  attributes are now filled

- fixed: bug on xmlns in path expression trigger 
  (would not replace prefixes in path expressions),
  spotted by amonroy on perlmonks, see

- optimized: XML::Twig text, thanks to Nick Lassonde
  for the patch

- fixed: bug that generated an empty line before some
  comments (pointed out by Tanya Huang)

- fixed: tests to check XML::Filter::BufferText version
  (1.00 has a bug in the CDATA handling that makes XML::Twig
  tests fail).

- added: new options --nowrap and --exclude (-v) to xml_grep

- fixed: warning in tests under 5.8.0 (spotted by Ed Avis)

- improved: skipped HTML::Entities tests in 5.8.0 (make test for this
  module seem to fail on my system, it might be the same 

- fixed: bug RT #6067 (problems with non-standard versions of
  Scalar::Utils which do not include weaken)

- fixed: bug RT #6092 (error when using safe output filter)

- fixed: bug when using map_xmlns, tags in default namespace 
  were not output

3.15 - 2004-04-05 

- fixed: tests now pass on more systems (thanks to Ed Avis for his testing)

- added: normalize_space option for simplify (suggestion of Lambert Lum) 

- improved: removed usage of $&

- improved: the doc for paste, as it was a bit short (suggestion of Richard Jolly)

3.14 - 2004-03-17 

- improved: namespace processing , it should work fine now,
  as long as twig_roots is not used.

  Potentially uncompatible change: the behaviour of simplify has
  been changed to mimic as exactly as possible XML::Simple's XMLin

- improved: the pod to cover the entire API
- improved: tests, now pass with perl 5.005_04-RC1 (fail with 5.005 reported
  by David Claughton), added more tests and a config summary at the
  end of the tests

- added: methods on the class attribute, convenient for dealing with
  XHTML or preparing display with CSS:
  class set_class add_to_class att_to_class add_att_to_class 
  move_att_to_class tag_to_class add_tag_to_class set_tag_class in_class
  navigation functions can use '.<class>' expressions

- fixed: (yet another!) bug in the way DTDs were output

- fixed: bug for pi => 'drop' option

- changed: the names of lots on internal (undocumented) methods, prefixed
  them with _

3.13 - 2004-03-16  - maintenance release to get the tests to pass on various platforms

- improved: the README file

- fixed: problem with encoding conversions (using safe_encode and
  safe_encode_hex) under perl 5.8.0, see RT ticket #5111

- fixed: tests to pass when trying to use an unsupported iconv filter    

3.12 - 2004-01-29 - new features and greatly increased test coverage

- added: lots of tests (>900), thanks to David Rigaudiere, Forrest 
  Cahoon, Sebastien Aperghis-Tramoni, Henrik Tougaard and Sam Tregar
  for testing this release on various OSs, Perl, XML::Parser and 
  expat versions.

- added: XML::Twig::XPath that uses XML::XPath as the XPath engine
  for findnodes, findnodes_as_string, findvalue, exists, find and
  matches. Just use XML::Twig::XPath instead of use XML::Twig;
  (see the tests in t/xmlxpath_*).

- added: special case to output some HTML tags ('script' to start with)
  as not empty.

- fixed: XML::Twig::Elt->new now properly flags empty elements (spotted by
  Dave Roe)

- added: XML::Twig::Elt contains_a_single method

- added: #ENT twig_handlers (not necessarily complete, so not yet 
  documented, needs more tests)

- added: doc for XML::Twig and XML::Twig::Elt subs_text methods
  tags starting with # are now "invisible" (they are not output),
  useful for example for pretty_printing

- added: new options --wrap '' and --date to xml_grep
  improved XPath support (added [nb] support)

- added: xpath method, which generates a unique XPath for an element

- added: has_child and has_children as synonyms of first_child

- added: XML::Twig::set_id_seed to control how generated id's are

- improved: when using ignore on an element, end_tag_handlers are now tested
  at the end of the element (so you can for example get the byte 
  offset in the document), suggestion of Philippe Verdret

- added: XML::Twig::Elt change_att_name 

- fixed: XML::Twig::Elt new now properly works when called as an object
  (and not a class) method

- fixed: namespace processing somewhat

- fixed: SAX output methods

- fixed: bug when keep_atts_order on and using set_att on an element
  with no existing attribute (spotted by scharloi) 

  WARNING - potentially incompatible changes -
  when using finish_print, the document used to be flushed. This is no
  longer the case, you will have to do it before calling finish_print. 
  This way you have the choice of doing it or not.

- improved: removed XML::Twig::Elt::unescape function (was no longer used)

3.11 - 2003-08-28 

- added: --text_only option to xml_grep (outputs the text of the
  result, without tags)

- fixed: bug where "Comments [was] always dropped after a twig 
  object set 'comments' to 'drop'" (RT#3711), bug report and first
  patch by Simon Flack

- added: option "keep_atts_order" that keeps the
  original attribute order in the output. This option needs the
  Tie::IxHash module to work.

3.10 - 2003-06-09 

- added: xml_pp xml_grep and xml_spellcheck to the distribution

- improved: the print method now calls 'print $elt->sprint' instead of printing
  content as it converts them to text, in order to reduce the number
  of calls to Perl's print (which should increase performance)

- changed: XML::Twig::Elt erase to allow erasing the root element 
  if it has only 1 child

- added: findvalue method to XML::Twig and XML::Twig::Elt

- added: aliased findnodes to get_xpath in XML::Twig and XML::Twig::Elt

- added: the elt_class option to XML::Twig::new

- added: the do_not_chain_handlers option to XML::Twig::new

- added: the XML::Twig::Elt is_first_child and is_last_child methods

- improved: set_gi,set_text, prefix, suffix, set_att, set_atts, del_atts, del_att
  now return the element for easier chains

- fixed: bug in pretty printing comments before elements (RT #2315)

- added: the XML::Twig::Elt children_copy method which returns a list
  of elements that are copies of the children of the element

- fixed: a bug in wrap_in when the element wrapped is not attached to a tree

- fixed: bug with get_xpath: regexp modifiers were not taken into account
  spotted by Eric Safern (RT #2284)

- fixed: bug in methods inherited from XML::Parser::Expat (arguments
  were not properly passed)

- improved: installed local empty SIG handlers to trap error messages triggered
  by require for optional modules, so that user signal handlers would
  not have to deal with them (suggestion from Philippe Verdret)

- fixed: bug in the navigation XPath engine: text() was used instead of 
  string(). Both are now allowed.

- added: XML::Twig::Elt sort_children, sort_children_on_value, 
  sort_children_on_att and sort_children_on_field methods that sort the
  children of an element in place

- added:XML::Twig::Elt field_to_att and att_to_field methods

- fixed:a memory leak due to ids not being weak references

- added: the XML::Twig::Elt wrap_children method that wraps children
  of an element that satisfy a regexp in a new element

- added: the XML::Twig::Elt add_id method that adds an id to an element
- added: the XML::Twig::Elt strip_add method that deletes an attribute
  from an element and its descendants

  fixed:a quasi-bug in set_att where the hash passed in reference was
  used directly, which makes it a problem when the same reference is
  passed several times: all the elements share the same attributes.
  This is a potentially incompatible change for code that relied on 
  this feature. Please report problems to the author.

- fixed: bug in set_id

- fixed: bug spotted by Bill Gunter: allowed _ as the initial character 
  for XML names. Also now allow ':' as the first element

- added: the simplify methods, which load a twig into an XML::Simple like 
  data structure

- fixed: bug in get_type and is_elt, spotted and fixed by Paul Stodghill

- added: the XML::Twig::Elt ancestors_or_self method

- fixed: bug when doc root is also a twig_root (twig was not built)

- improved: the README (fleshed out examples, added OS X to the list of 
  tested platforms)

- fixed: bug when using the no_dtd_output option

- added: doc for the XML::Twig::Elt children_count method

- added: the XML::Twig::Elt children_text method

- improved: updated the doc so it can be properly formatted by my custom pod2html,
  the generated doc (with a bigger ToC and better links) is available
  from the XML::Twig page at
3.09 - 2002-11-10 

- added: XML::Twig::Elt xml_text method

- fixed: several bugs in the split method under 5.8.0 when matching a utf8
  character (thanks to Dominic Mitchell who spotted them)

- improved: cleaned-up the pod (still in progress)

- added: the XML::Twig::Elt pos method that gives the position of
  an element in its parent's child list

- fixed: re-introduced parseurl (thanks to Denis Kolokol for spotting its 
  absence in this version)

- fixed: ent_tag_handlers were not called on the root (thanks
  to Philippe Verdret

- improved: #PI (also declared as '?') and #COMMENT handler support 

- added: check on reference type (must be XML::Twig::Elt) in 
  XML::Twig::Elt::paste (patch by Forrest Cahoon)

3.08 - 2002-09-17 

- fixed: the previous fix wasn't enough :--( 

3.07 - 2002-09-17 

- fixed:the way weaken is imported from Scalar::Util

3.06 - 2002-09-17 

- added: XML::Twig::Elt trimmed_text and related methods (trimmed_field,
  first_child_trimmed_text, last_child_trimmed_text...)

- added: XML::Twig::Elt replace_with method

- added: XML::Twig::Elt cut_children method

- added: XML::Twig contains_only method

- added: *[att=~ /regexp/] condition type (suggested by Nikola Janceski)

- fixed: bug in the way handlers for gi, path and subpath were chained
  (Thanks to Tommy Wareing)

- fixed: bug where entities caused an error on other handlers (Thanks
  to Tommy Wareing)

- fixed: bug with string(sub_elt)=~ /regexp/ (thanks to Tommy Wareing)

- fixed: bug with output_filter used with expand_external_entities
  (thanks to Tommy Wareing)

- fixed: (yet another!) bug with whitespace handling (whitespace, then an
  entity made the whitespace move after the entity) (spotted by the usual
  Tommy Wareing)

- added: an error message when pasting on an undef reference (suggestion
  of Tommy Wareing)

- fixed: bug in in_context (found by Tommy Wareing)

- fixed: bug when loading the DTD (local undef $/ did not stay local,
  bug found and patch sent by Steve Pomeroy and Henry Cipolla)

- fixed: bug in setting output filter

- fixed: bug in using a filehandle with twig_print_outside_roots

- added: safe_encode_hex filter

- fixed: bug in set_indent, $INDENT not set properly (thanks
  to Eric Jain)

- fixed: dependencies (no check with 5.8.0, added Scalar::Util
  as a possible source for weaken)

- added: no_prolog option to XML:Twig::new

- improved: tested build on Windows (thanks to Cory Trese and Josh Hawkins)

- changed:in 3.05

- added: _ALPHA_ SAX export methods: 
   XML::Twig toSAX1, toSAX2, flush_toSAX1, flush_toSAX2 XML::Twig::Elt toSAX1, toSAX2
   The following gotchas apply:
   + these methods work only for documents that are completely
     loaded by XML::Twig (ie if you use twig_roots the data
     outside of the roots will not be output as SAX). 
   + SAX1 support is a bit dodgy: the encoding is not preserved
     (it is always set to 'UTF-8'),
   + locator is not supported (and probably will not, what's the 
     location of a newly created element?)
  Also when exporting SAX you should consider setting Twig to a 
  mode where all aspects of the XML are treated as nodes by XML::Twig,
  by setting the following options when you create the twig:
  comments => 'process', pi => 'process', keep_spaces => 1
- improved: twig_print_outside_roots now supports a file handle ref as argument:
  the untouched part of the tree will be output to the filehandle:

- added: the 'indented_c' style that gives a slightly more compact pretty
  print than 'indented': the end tags are on the same line as the 
  preceding text (suggestion of Hugh Myers)
- added: option in get_xpath (aka find_nodes) to apply the query to
  a list of elements 

- added: processing of conditions on the current node in get_xpath:
  my @result= get_xpath( q{.[@att="val"]});
  This is of course mostly useful with the previous option.
  The idea stemmed from a post from Liam Quin to the perl-xml list

- added: XML::Twig xml_version, set_xml_version, standalone, set_standalone
  methods on the XML declaration

- fixed: bug in change_gi (which simply did not work at all), found
  by Ron Hayden.

- fixed: bug in space handling with CDATA (spaces before the CDATA section
  were moved to within the section), comments and PI's

- fixed: bug in parse_url (exit was not called at the end of the child),
  found by David Kulp

- improved: cleanup a bit the code that parses xpath expressions (still some work 
  to be done on this though), fixed a bug with last, found by Roel de Cock

- fixed: the SYNOPSIS (parsefile is used to parse files, spotted by e.sammer)

- fixed: bug in pretty printing (reported by Zhu Zhou)

- fixed: bugin the install: the Makefile now uses the same perl used 
  to perl Makefile.PL to run speedup and check_optional_modules 
  (reported by Ralf Santos)

- fixed: bugs in pretty printing when using flush, trying to figure out
  as well as possible if an element contains other elements or text 
  (there is still a gotcha, see the BUGS section in the docs) 

- fixed: bug that caused the XML declaration and the DTD not to be reset
  between parses

- improved: the conversion functions (errors are now reported when the
  function is created and not when it is first used)

- added: the output_encoding option to XML::Twig->new, which allows 
  specifying an encoding for the output: the conversion filter is
  created using Encode (perl 5.8.0) Text::Iconv or Unicode::* The
  XML declaration is also updated

- added: #CDATA and #ENT can now be used in handler expressions

- added: XML::Twig::Elt remove_cdata method, which turns CDATA sections
  into regular PCDATA elements

- improved: set_asis can now be used to output CDATA sections un-escaped (and without
  the CDATA section markers) 

3.04 - 2002-04-01

- fixed: handlers for XML::Parser 2.27 so the module can pass the tests

3.03 - 2002-03-26

- fixed: bugs in entity handling in twig_roots mode

- added: the ignore_elts option, to skip completely elements

- improved: enhanced the XPath-like syntax in navigation and get_xpath
  methods: added operators (>, < ...)

- fixed: [RT 168]: setTwigHandler failed when no handler was already set
  (thanks to Jerry)
- improved: turned %valid_option into a package global so AnyData can access it
- fixed: bug in sprint that prevented it from working with filters

- fixed: bug in erase when erasing an empty element that was the
  last child of its parent ([RT390], thanks to Julian Arnold)
- fixed: copy now correctly copies the asis status of elements

- fixed:typos on the docs (thanks to Shlomo Yona)

- added: tests (for erase and entities in twig_roots mode)

3.02 - 2002-01-16

- fixed: tweaked speedup to replace constructs that did not work in
  perl 5.005003

3.01 - 2002-01-09
- fixed: the directory name in the tar file 

3.00 - 2002-01-09


  But it is The Right Thing To Do

  In normal mode (when KeepEncoding is not used) the XML data is
  now stored as parsed by XML::Parser, ie the base entities are
  expanded. The "print" methods (print, sprint and flush, plus the
  new xml_string, pcdata_xml_string and att_xml_string) return the
  data in XML-escaped form: & and < are escaped in PCDATA and
  &, < and the quote (" by default) are turned to &amp; &lt; and
  &quot; (or &apos; if the quote is '). The "text" methods (text,
  att and pcdata) return the stored text as is.
  So if you want to output XML you should use the "print" methods
  and if you want to output text you should use the "text" methods.

  Note that this breaks the trick consisting in adding tags to the
  content of an element: $elt->prefix( "<b>") no longer adds a <b>
  tag before an element. $elt->print will now output "&lt;b>...".
  (but you can still use it by marking those elements as 'asis').
  It also fixes the annoying &apos; thingie that used to replace '
  in the data. 

  When the KeepEncoding option is used this is not true, the data
  is stored asis, base entities are kept un-escaped.

  Note that KeepEncoding is a global setting, if you use several twigs,
  some with KeepEncoding and some without then you will have to manually
  set the option using the set_keep_encoding method, otherwise the last
  XML::Twig::new call will have set it

  In addition when the KeepEncoding option is used the start tag is
  parsed using a custom function parse_start_tag, which works only
  for 1-byte encodings (it is regexp-based). This method can be 
  overridden using the ParseStartTag (or parse_start_tag) option
  when creating the twig. This function takes the original string as
  input and returns the gi and the attributes (in a hash).
  If you write a function that works for multi-byte encodings I would
  very much appreciate if you could send it back to me so I can add it
  to the module, so other users can benefit from it.

  An additional option ExpansExternalEnts will expand external entity
  references to their text (in the output, the text stored is &ent;).

- added: when handlers (twig_handlers or start_tag_handlers) are called
  $_ is set to the element node, so quick hacks look better:
      my $t= new XML::Twig( twig_handlers => 
            { elt => sub { print $_->att( 'id'), ": ", $_->text, "\n"; } }

- added: XML::Twig dispose method which properly reclaims all the memory
  used by the object (useful if you don't have WeakRef installed)

- added: XML::Twig and XML::Twig::Elt ignore methods, which can be called
  from a start_tag_handlers handler and cause the element (or the
  current element if called on a twig) to be ignored by the 

- added: XML::Twig parse_start_tag option that overrides the default function
  used to parse start tags when KeepEncoding is used

- added: XML::Twig::Elt xml_string, pcdata_xml_string and att_xml_string
  all return an XML-escaped string for an element (including 
  sub-elements and their tags but not the enclosing tags for the 
  element), a #PCDATA element and an attribute

- added: XML::Twig::Elt methods tag and set_tag, equivalent respectively
  to gi and set_gi
- added: XML::Twig and XML::Twig::Elt set_keep_encoding methods can be used
  to set the keep_encoding value if you use several twigs with 
  different keep_encoding options

- improved: option names for XML::Twig::new are now checked (a warning is output
  if the option is not a valid one);

- improved: when using pretty_print nice or indented keep_spaces_in is now checked
  so the elements within an element listed in keep_spaces_in are not

- added: XML::Twig::Elt insert_new_elt method that does a new and a paste

- added: XML::Twig::Elt split_at method splits a #PCDATA element in 2

- added: XML::Twig::Elt split method splits all the text descendants of an 
   element, on a regep, wrapping text captured in brackets in the
   regexp in a specified element, all elements are returned 

- added: XML::Twig::Elt mark method is similar to the split method, except
  that only newly created elements (matched by the regexp) are

- added: XML::Twig::Elt get_type method returns #ELT for elements and the gi
  (#PCDATA, #CDATA...) otherwise

- added: XML::Twig::Elt is_elt returns the gi if the element is a real element
  and 0 if it is #PCDATA, #CDATA...
- added: XML::Twig::Elt contains_only_text returns 1 if the element contains no
  "real" element (is_field is another name for it)

- added: First implementation of the output_filter option which filters the
  text before it is output by the print, sprint, flush and text methods
  (only works for print at the moment, and still under test with various
  versions of XML::Parser). Standard filters are also available

      #!/bin/perl -w
      use strict;
      use XML::Twig;

      my $t = new XML::Twig(output_filter => 'latin1');
      $t->parse( \*DATA);
      <?xml version="1.0" encoding="ISO-8859-1"?>
         <docù atté="valuè">Un homme soupçonné d'être impliqué dans 
           la mort d'un motard de la police, renversé

  The 'latin1', 'html' and 'safe' filters are predefined, you can also
  build additional filters using Iconv (requires text::Iconv) and 
  Unicode::String (requires Unicode::String and Unicode::Map8):

      my $conv = XML::Twig::iconv_convert( 'latin1');
      my $t = new XML::Twig(output_filter => $conv);

      my $conv = XML::Twig::unicode_convert( 'latin1');
      my $t = new XML::Twig(output_filter => $conv);

  warning: conversions work fine with XML::Parser 2.27 but sometimes fail
  with XML::Parser 2.30 (on Perl 5.6.1, Linux 2.4 on a PC) when using 
  'latin1' without Text::Iconv or Unicode::String and Unicode::Map8

  The input_filter option works the same way, except the text is 
  converted before it is stored in the twig (so you can use regexp in 
  your native encoding for example) 

- added: the XML::Twig::Elt set_asis method sets a property of an element that
  causes it to be output asis (without XML-escaping < " and &) so you
  can still create tagged text

- added: the XML::Twig::Elt prefix and suffix methods accept an optional
  'asis' argument that causes the prefix or suffix to get the asis
  property (so you can do $elt->prefix( '<b>foo</b>', 'asis') for

- added: the XML::Twig and XML::Twig::Elt find_nodes methods are aliases
  to the get_xpath method (this is the name used in XML::XPath)

- added: the XML::Twig parseurl and safe_parseurl methods parse a document
  whose url is given

- added: XML::Twig::Elt extra_data, set_extra_data and append_extra_data to
  access the... extra data (PI's and comments) attached to an element

- added: XML::Twig method parser returns the XML::Parser::Expat object used
  by the twig 

- improved: Most XML::Parser::Expat methods are now inherited by XML::Twig

- added: XML::Twig::Elt descendant_or_self method that returns the element
  and its descendants

- fixed: element (and attribute) names can now include '.'

- fixed: get_xpath now works for root based XPath expressions ('/doc/elt') 

- fixed: get_xpath now works for regexps (including regexps on attribute values)

- fixed: you can now properly restore pretty_print and empty_tag_style values

- fixed: speedup (at install) now checks the Perl version and uses qr or ""
   so XML::Twig works in 5.004 

- fixed: XML::Twig::Elt wrap_in now allows wrapping the root element

- fixed: various bugs in the DOCTYPE and DTD output with XML::Parser 2.30
- fixed: the tests to fix a bug when working with XML::Parser 2.27  

- fixed: the tests to fix a bug preventing test2 to pass under windows

- fixed: _default_  handlers now work (thanks Zoogie)

- fixed: the text method now returns the XML base entities (<>&'") un-escaped
  (thanks to Hakan Kallberg's persistence to ask for it ;--)

- fixed: pretty_print works better for elements without content

- fixed: end_tag_handlers now work properly (thanks to Phil Glanville for the

- improved: attributes which name starts with # are not output by the print
  methods, and thus can be used to store private data on elements

- improved: WeakRef is used if installed, so no more memory leaks
- improved: sped-up print and flush by creating the _print and _flush methods 
  which do not check for file handle and pretty print options

- improved: the doc has been enhanced and somewhat restructured. All options are
  now written as this_is_an_option although the legacy form thisIsAnOption
  can still be used. Links now display properly in the text form (thanks to
  Dominic Mitchell for spotting this and sending a patch)

- improved: navigation functions (including descendants) now allow not only a gi
  to be used as filter, but also the '#ELT' token, to filter only "real"
  elements (as opposed to #PCDATA, #CDATA, #PI, #COMMENT, #ENT), the 
  '#TEXT' token, to filter only text (PCDATA and CDATA elements), 
  regular expressions (built with qr//) applied on the elements gi's,
  code references, the code is passed the element as argument, and a
  subset of XPath.
  Functions that can use this token are: children, first_child, last_child,
  prev_sibling, last_sibling, next_elt, last_elt, descendants, get_xpath,
  child, sibling, sibling_text, prev_siblings, next_siblings field, 

- improved: the paste method now accepts a 'within' position, which inserts the
  element at the $offset argument (a 3rd, required, argument) in the
  reference element or in its first text child 

- improved: the XML::Twig::Elt insert method now accepts attributes (hashrefs)
  applied to the element(s) being inserted:
  $elt->insert( e1 => { a => 'v'}, e2 => e3 => { a1 =>'v1', a2 => 'v2'}); 

- improved: the XML::Twig::erase method now outputs a meaningful error message if
  applied to the root (or a cut element)

- improved: optimizations for better performances (in the end performances are about
  the same or a little worse than XML::Twig 2.02 but the module is much
  more powerful)

[Known bugs]

- the DTD interface is completely broken, and I have little hope of
  fixing it considering I have to deal with 2 incompatible versions of
  XML::Parser. Plus no one seems to be using it...

- some XPath/Navigation expressions using " or ' in the text()="" part
  of the expression will cause a fatal error

- note that this version works better (but doesn't necessarily require)
  with WeakRef (Perl version 5.6.0 and above) and Text::Iconv for all
  its encoding conversions.