<?php
# XMLFile -- Initial version June 1, 2000
# May 29, 2002
# Fixed to work with later versions of PHP that have deprecated the call-time
# pass by reference syntax. This may break versions of PHP older than 3, so
# if you are using this on those versions, go with an earlier version.
# Also added some more comments on how to use it to create XML files. Some
# people (a surprising number of them, in fact) were having difficulty
# understanding the recursive nature of XML files, and that each tag has
# subtags, each of which can have subtags, each of which....
# July 12, 2001
# Fixed an oops in the find_subtags_by_name function.
# July 11, 2001
# Incorporated Kevin Howe's read_xml_string function.
# Incorporated Mike Konkov's find_subtags_by_name function (with some changes).
# Fixed an errant documentation comment (open instead of fopen -- oops).
#
# September 29, 2000
# by Chris Monson -- e408345b17be3ce90059d01d96be0599@orangatango.com This PHP module is licensed under the GNU LGPL (www.gnu.org)
# Please become familiar with this license before sending this thing all over
# the place! I would like to have any changes submitted back to me, and this
# comment must be included with any distribution of this component.
#
# The purpose of this module was to provide a simple interface for dealing with
# small to medium sized XML files. It is probably not suitable for very large
# files since it reads the entire file into a structure in memory. For very
# large files, the XML parsing functions should be used more or less directly
# so that pieces of the file can be dealt with as they are read in.
#
# The basic idea is this: Read the XML file into an internal tree structure
# that is easy to traverse. This module also allows you to create such a
# structure in memory and then write it out to disk. The output is formatted
# in a nice, readable way, using whitespace to delimit tag containment, etc.
# It makes it very easy to write nice-looking XML files.
#
# I have included some usage comments. They are almost certainly incomplete.
# If you get stumped, first use the source, then email me.
#
#------------------------------------------------------------------------------
### USAGE ###
#------------------------------------------------------------------------------
# Reading an XML file:
#------------------------------------------------------------------------------
#
# $xml = new XMLFile();
# $fh = fopen( 'myxmlfile.xml', 'r' );
# $xml->read_file_handle( $fh );
# close( $fh );
#
# Now the tags can be accessed via the root node of $xml:
#
# $root = &$xml->roottag;
# $tagname = $root->name;
# $tagdata = $root->cdata;
#
# Note that a tag is of the following form:
# <NAME attribute=value>CDATA</NAME>
# Each tag contains an attributes array, which is associative by nature. In
# other words, you would access the value of "attribute" as follows:
#
# $value = $root->attributes['attribute'];
#
# Also, each tag has a 'tags' proprerty, which is an ordered array (integer
# indices, not associative!) which has the tags that were contained within
# this tag in their order of appearance. The reason that this is not
# associative is that there can be multiple tags of the same name. There is
# nothing in the XML spec (barring a DTD) that declares the uniqueness of tag
# names. For example:
#
# <OUTER>
# <INNER>CDATA</INNER>
# <INNER name="hello"/>
# </OUTER>
#
# In the above example, the outer tag would have a tags array that has two
# entries, each of which has a tag name of "INNER". The one with CDATA wrapped
# inside would be at index 0, and the other would be at index 1.
#
# Once you have finished with the XMLFile object, you need to call the cleanup
# method. If you don't, you will get a memory leak, since PHP is reference
# counted and each element in the tree refers to its parent. 'cleanup' simply
# traverses the tree and disconnects the parent pointers. The PHP cleans up
# everything else.
#
# $xml->cleanup();
#
# Note that you can change the elements, delete tags, and do other things
# to the tree once it has been read from the file. After it has been changed,
# you can write it back out and the file will reflect your changes.
#------------------------------------------------------------------------------
# Writing a new file:
#
# $xml = new XMLFile();
# $xml->create_root(); # necessary -- no root is created until requested
# $xml->roottag->name = 'ROOT';
# $xml->roottag->add_subtag( 'INNER', array() );
# $innertag = &$xml->roottag->curtag;
# $innertag->add_subtag( 'REALLYINNER', array() );
# # Or, you can do this:
# $xml->roottag->curtag->add_subtag( 'INNER2', array() );
# # The point is that each tag can have subtags. The most recently added
# # subtag is always the curtag of its parent.
# $xml->roottag->add_subtag( 'INNER', array( 'name' => 'value' ) );
# $xml->roottag->curtag->cdata = "Hello!"; # curtag is the most recent addition
# $fh = fopen( 'myxmlfile.xml', 'w' );
# $xml->write_file_handle( $fh );
# close( $fh );
#
# The file will look like this: (no space between ? and >)
#
# <?xml version="1.0" encoding="UTF-8" ? >
# <ROOT>
# <INNER>
# <REALLYINNER/>
# <INNER2/>
# </INNER>
# <INNER name="value">Hello!</INNER>
# </ROOT>
#
#------------------------------------------------------------------------------
#------------------------------------------------------------------------------
#
if (!isset($XMLFile_Included) || !$XMLFile_Included) {
$XMLFile_Included = 1;
###############################################################################
class XMLTag
{
var $cdata;
var $attributes;
var $name;
var $tags;
var $parent;
var $curtag;
function XMLTag(&$parent)
{
if (is_object( $parent ))
{
$this->parent = &$parent;
}
$this->_init();
}
function _init()
{
$this->attributes = array();
$this->cdata = '';
$this->name = '';
$this->tags = array();
}
function add_subtag($name, $attributes=0)
{
$tag = new XMLTag( $this );
$tag->set_name( $name );
if (is_array($attributes)) {
$tag->set_attributes( $attributes );
}
$this->tags[] = &$tag;
$this->curtag = &$tag;
}
function find_subtags_by_name( $name )
{
$result = array();
$found=false;
for($i=0;$i<$this->num_subtags();$i++) {
if(strtoupper($this->tags[$i]->name)==strtoupper($name)) {
$found=true;
$array2return[]=&$this->tags[$i];
}
}
if($found) {
return $array2return;
}
else {
return false;
}
}
function clear_subtags()
{
# Traverse the structure, removing the parent pointers
$numtags = sizeof($this->tags);
$keys = array_keys( $this->tags );
foreach( $keys as $k ) {
$this->tags[$k]->clear_subtags();
unset($this->tags[$k]->parent);
}
# Clear the tags array
$this->tags = array();
unset( $this->curtag );
}
function remove_subtag($index)
{
if (is_object($this->tags[$index])) {
unset($this->tags[$index]->parent);
unset($this->tags[$index]);
}
}
function num_subtags()
{
return sizeof( $this->tags );
}
function add_attribute( $name, $val )
{
$this->attributes[strtoupper($name)] = $val;
}
function clear_attributes()
{
$this->attributes = array();
}
function set_name( $name )
{
$this->name = strtoupper($name);
}
function set_attributes( $attributes )
{
$this->attributes = (is_array($attributes)) ? $attributes : array();
}
function add_cdata( $data )
{
$this->cdata .= $data;
}
function clear_cdata()
{
$this->cdata = "";
}
function write_file_handle( $fh, $prepend_str='' )
{
# Get the attribute string
$attrs = array();
$attr_str = '';
foreach( $this->attributes as $key => $val )
{
$attrs[] = strtoupper($key) . "=\"$val\"";
}
if ($attrs) {
$attr_str = join( " ", $attrs );
}
# Write out the start element
$tagstr = "$prepend_str<{$this->name}";
if ($attr_str) {
$tagstr .= " $attr_str";
}
$keys = array_keys( $this->tags );
$numtags = sizeof( $keys );
# If there are subtags and no data (only whitespace),
# then go ahead and add a carriage
# return. Otherwise the tag should be of this form:
# <tag>val</tag>
# If there are no subtags and no data, then the tag should be
# closed: <tag attrib="val"/>
$trimmeddata = trim( $this->cdata );
if ($numtags && ($trimmeddata == "")) {
$tagstr .= ">\n";
}
elseif (!$numtags && ($trimmeddata == "")) {
$tagstr .= "/>\n";
}
else {
$tagstr .= ">";
}
fwrite( $fh, $tagstr );
# Write out the data if it is not purely whitespace
if ($trimmeddata != "") {
fwrite( $fh, $trimmeddata );
}
# Write out each subtag
foreach( $keys as $k ) {
$this->tags[$k]->write_file_handle( $fh, "$prepend_str\t" );
}
# Write out the end element if necessary
if ($numtags || ($trimmeddata != "")) {
$tagstr = "</{$this->name}>\n";
if ($numtags) {
$tagstr = "$prepend_str$tagstr";
}
fwrite( $fh, $tagstr );
}
}
}
###############################################################################
class XMLFile
{
var $parser;
var $roottag;
var $curtag;
function XMLFile()
{
$this->init();
}
# Until there is a suitable destructor mechanism, this needs to be
# called when the file is no longer needed. This calls the clear_subtags
# method of the root node, which eliminates all circular references
# in the xml tree.
function cleanup()
{
if (is_object( $this->roottag )) {
$this->roottag->clear_subtags();
}
}
function init()
{
$this->roottag = "";
$this->curtag = &$this->roottag;
}
function create_root()
{
$null = 0;
$this->roottag = new XMLTag($null);
$this->curtag = &$this->roottag;
}
# read_xml_string
# Same as read_file_handle, but you pass it a string. Note that
# depending on the size of the XML, this could be rather memory intensive.
# Contributed July 06, 2001 by Kevin Howe
function read_xml_string( $str )
{
$this->init();
$this->parser = xml_parser_create("UTF-8");
xml_set_object( $this->parser, $this );
xml_set_element_handler( $this->parser, "_tag_open", "_tag_close" );
xml_set_character_data_handler( $this->parser, "_cdata" );
xml_parse( $this->parser, $str );
xml_parser_free( $this->parser );
}
function read_file_handle( $fh )
{
$this->init();
$this->parser = xml_parser_create("UTF-8");
xml_set_object( $this->parser, $this );
xml_set_element_handler( $this->parser, "_tag_open", "_tag_close" );
xml_set_character_data_handler( $this->parser, "_cdata" );
while( $data = fread( $fh, 4096 )) {
if (!xml_parse( $this->parser, $data, feof( $fh ) )) {
die(sprintf("XML error: %s at line %d",
xml_error_string(xml_get_error_code($this->parser)),
xml_get_current_line_number($this->parser)));
}
}
xml_parser_free( $this->parser );
}
function write_file_handle( $fh, $write_header=1 )
{
if ($write_header) {
fwrite( $fh, "<?xml version='1.0' encoding='UTF-8'?>\n" );
}
# Start at the root and write out all of the tags
$this->roottag->write_file_handle( $fh );
}
###### UTIL #######
function _tag_open( $parser, $tag, $attributes )
{
#print "tag_open: $parser, $tag, $attributes\n";
# If the current tag is not set, then we are at the root
if (!is_object($this->curtag)) {
$null = 0;
$this->curtag = new XMLTag($null);
$this->curtag->set_name( $tag );
$this->curtag->set_attributes( $attributes );
}
else { # otherwise, add it to the tag list and move curtag
$this->curtag->add_subtag( $tag, $attributes );
$this->curtag = &$this->curtag->curtag;
}
}
function _tag_close( $parser, $tag )
{
# Move the current pointer up a level
$this->curtag = &$this->curtag->parent;
}
function _cdata( $parser, $data )
{
$this->curtag->add_cdata( $data );
}
}
###############################################################################
} // included
###############################################################################
|
Download
