Difference between revisions of "XmlWiki"

From Organic Design wiki
Line 1: Line 1:
−
==What is xmlWiki?==
+
<?php
−
Articles in xmlWiki work just like the normal MediaWiki articles except, when their contents is xml. In this case it will be represented as a DOM object at runtime, and the content will be validated against any referenced DTD or xml-schema. If any XSLT's are referenced, xmlWiki will attempt to apply them to the article. If the XSLT-output-method is "html" then it will be reduced to a string.
+
# xmlWiki - MediaWiki XML Hack
−
 
+
# Nad - 2005-05-18
−
==How is it Installed?==
 
−
It's three scripts (source below) which are in an "xmlwiki" directory in the main wiki folder. the hack is applied or removed by running the xmlwiki/index.php script. If xmlWiki is disabled then the security layer is also disabled, so the xmlwiki directory should be accessible only by admin.
 
  
−
==How does it tie in with MediaWiki?==
+
# INIT
−
Two pseudo-namespaces have been added "xml" and "sys". These are both XML articles, and their contents affects the article they're associated with. Any sys:article is readable and writable only by users in the "admin" group.
+
# SECURITY
−
*The sys:article holds security information: owner, read-groups, write-groups
+
# INPUT HOOK
−
*An xml:article is an XML article of docType "xmlwiki.dtd"
+
# OUTPUT HOOK
−
*The xml:article contains the transform-list, publish-list and base-class-list
+
# PARSER HOOK
 +
# SKIN HOOK
 +
# ARTICLE FUNCTIONS
 +
# UTILITY FUNCTIONS
  
−
==More Specifically==
+
# ---------------------------------------------------------------------------------------------------------------------- #
−
There are five hooks now! but it's most efficient like this :-)
 
  
−
*'''INIT:''' This hooks into ''index.php'' after the environment has been established, but before any article or input processing occurs. This does the ''require_once'' of ''xmlwiki.php''.
+
# TODO:
−
*'''INPUT:''' This hooks into ''index.php'' before input processing so it can direct any XPath inputs into the article-DOM and transform them into a standard input for xmlWiki. This also handles write-permissions.
+
# - Undomificate: validation, xslt
−
*'''PARSER:''' This hooks into the parse-function in ''includes/Parser.php''. If the article is  XML, it is validated and domificated. If it includes XSLT references, or has a transform-list in its ''xml:article'' then those transforms are also applied here. The article will be back in string form again after this.
+
# - Transform: check perms on php-execution
−
*'''SKIN:''' This is a new compulsory skin (''skins/xwskin.php'') which renders the page output as a DOM-object so xmlWiki can apply design transforms to it rather than PHP-based templates. If ''raw'' content is requested, then only the articles own transforms are applied, because skins and page-layout are part of the xmlWiki environment, not the article.
+
# Publish:
−
*'''OUTPUT:''' This replaces the output rendering in ''index.php''. It handles read permissions and transforms the xwskin-generated DOM-output with ''xmlwiki.xslt''.
+
# - xml:article - holds article publish-list
 +
# - mirror/wiki-sync
 +
# Transforms:
 +
# - xml:article - now transforms are in here (main article is now like <body>)
 +
# - so we don't do the "its one of ours" thing anymore
  
−
'''Current scripts:'''
+
# LATER:
−
*[[xmlwiki.php|xmlWiki main script]]
+
# - allow XIncludes to build large docs from others
−
*[[xwadmin.php|xmlWiki admin script]] (installs and uninstalls the hooks)
+
# - allow docBook and DSSSL content
−
*[[xwskin.php|xmlWiki skin-hook]]
+
# - Maybe only use XML transforms and use PHP via PI's
  
−
'''Templates & Transforms:'''
+
# ---------------------------------------------------------------------------------------------------------------------- #
−
*[[xmlwiki.xslt]] (xmlWiki page layout - not functional yet)
+
# INIT
−
*[[xmlwiki.css]] (xmlWiki page style)
 
  
−
'''Wiki Objects available from OUTPUT-hook:'''
+
# Exit if not included from index.php
−
*[[Wiki Objects]]
+
defined('MEDIAWIKI') or die('xmlwiki.php must be included from MediaWiki\'s index.php!');
 +
 
 +
# Allows Geshi to be installed by PARSER-HOOK if not installed th usual way already
 +
if (!@is_object($wgGeshiSyntaxSettings)) $wgGeshiSyntaxSettings = false;
 +
 
 +
# Otherwise set up xmlWiki global environment ready for input and output processing
 +
$xwMessages = array();
 +
$xwEdit = isset($_REQUEST['action']) && ($_REQUEST['action'] == 'edit');
 +
$xwView = (!isset($_REQUEST['action']) || (isset($_REQUEST['action']) && ($_REQUEST['action'] == 'view')));
 +
$xwRaw = isset($_REQUEST['action']) && ($_REQUEST['action'] == 'raw');
 +
$xwPreview = isset($_REQUEST['wpPreview']);
 +
$xwSave = isset($_REQUEST['wpSave']);
 +
$xwUserName = $wgUser->mName;
 +
$xwUserGroups = array(strtolower($xwUserName), 'anyone', 'everyone', 'nobody', 'world');
 +
$xwArticleReadableBy = array('anyone');
 +
$xwArticleWritableBy = array('anyone');
 +
$xwArticle = $xwPreview ? $wgRequest->gettext('wpTextbox1') : $wgArticle->getContent(false);
 +
$xwArticleTitle = $wgTitle->getPrefixedURL();
 +
$xwArticleCache = array();
 +
$xwArticleLinks = array();
 +
$xwUserLinks = array();
 +
$xwMetaArticle = false;
 +
$xwMsgToken = '<!--xwMsg-->';
 +
$xwTemplate = null;
 +
$xwParserHookCalled = false;
 +
$xwSkinHookCalled = false;
 +
 
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# SECURITY
 +
 
 +
# Not currently used: R/W by user, for xmlWiki prefs etc
 +
if ($xwUserXML = xwArticleContent(strtolower("xml:$xwUserName")))
 +
xwDomificateArticle($xwUserXML, "xml:$xwUserName");
 +
 
 +
# R/W by admin only, holds users groups and other system info like stats
 +
# - no sys:user means "anyone" in group "everyone" as far as perms goes
 +
if ($xwUserSYS = xwArticleContent(strtolower("sys:$xwUserName")))
 +
xwDomificateArticle($xwUserSYS, "sys:$xwUserName");
 +
 
 +
# If sys:user exists, get users' group-list
 +
if (is_object($xwUserSYS)) {
 +
$xwRoot = $xwUserSYS->document_element();
 +
if (count($xwList = $xwRoot->get_elements_by_tagname('groups')))
 +
$xwUserGroups = array_merge($xwUserGroups, split(',', strtolower($xwList[0]->get_content())));
 +
}
 +
 
 +
# - Security info is: owner, read-groups, write-groups (default values are "world")
 +
# - Users and groups are interchangeable
 +
# - Default user is "anyone", default group is "everyone"
 +
if (eregi('^xml:(.+)$', $xwArticleTitle)) {
 +
# This is an xml-pseudo-namespace
 +
# - its perms apply to its 'parent' and to itself aswell
 +
$xwXML = $xwArticle;
 +
xwDomificateArticle($xwXML, $xwArticleTitle);
 +
if (is_object($xwXML)) {
 +
$xwRoot = $xwXML->document_element();
 +
if (count($xwList = $xwRoot->get_elements_by_tagname('read')))
 +
$xwArticleReadableBy = split(',', strtolower($xwList[0]->get_content()));
 +
if (count($xwList = $xwRoot->get_elements_by_tagname('write')))
 +
$xwArticleWritableBy = split(',', strtolower($xwList[0]->get_content()));
 +
}
 +
}
 +
elseif (eregi('^sys:(.+)$', $xwArticleTitle)) {
 +
# This is a sys-pseudo-namespace
 +
# - only admin can read/write sys:
 +
$xwArticleReadableBy = $xwArticleWritableBy = array('admin');
 +
}
 +
else {
 +
# This is a normal article, get xml:article
 +
if ($xwArticleXML = xwArticleContent(strtolower("xml:$xwArticleTitle")))
 +
xwDomificateArticle($xwArticleXML, "xml:$xwArticleTitle");
 +
if (is_object($xwArticleXML)) {
 +
$xwRoot = $xwArticleXML->document_element();
 +
if (count($xwList = $xwRoot->get_elements_by_tagname('read')))
 +
$xwArticleReadableBy = split(',', $xwList = strtolower($xwList[0]->get_content()));
 +
if (count($xwList = $xwRoot->get_elements_by_tagname('write')))
 +
$xwArticleWritableBy = split(',', $xwList = strtolower($xwList[0]->get_content()));
 +
}
 +
}
 +
 
 +
# Set perms for this request
 +
$xwReadable = 0 < count(array_intersect($xwArticleReadableBy, $xwUserGroups));
 +
$xwWritable = 0 < count(array_intersect($xwArticleWritableBy, $xwUserGroups));
 +
if ($xwDebug) {
 +
xwMessage('Groups: '.join(', ', $xwUserGroups));
 +
xwMessage("Readable ($xwReadable): ".join(', ', $xwArticleReadableBy));
 +
xwMessage("Writable ($xwWritable): ".join(', ', $xwArticleWritableBy));
 +
}
 +
 
 +
# Divert to access-denied article if not readable
 +
if (!$xwReadable) {
 +
$action = 'view';
 +
$xwSave = $xwEdit = false;
 +
die('Sorry, the requested article was not readable.');
 +
}
 +
 
 +
 
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# INPUT HOOK
 +
 
 +
# Parse and process input from forms
 +
function xwInputHook() {
 +
 
 +
global $xwArticle, $xwArticleXML, $xwArticleTitle, $xwWritable, $_REQUEST;
 +
global $xwDebug, $xwSave, $xwEdit, $action, $xwUserName, $wgArticle;
 +
if ($xwDebug) xwMessage('xwInputHook()','green');
 +
 
 +
# If not writable, change action to view
 +
if (!$xwWritable && ($action != 'view')) {
 +
xwMessage('Sorry, article not writable, action changed to "view".', 'red');
 +
$action = 'view';
 +
$xwSave = $xwEdit = false;
 +
}
 +
 
 +
# Parse POST for XPath inputs
 +
# - Form-input-name is "xw:[[@]node]:XPathQuery" which directs the value into xwArticle (if DOM)
 +
# - if "node" exists, then a new element (or att if "@node") is created for the value
 +
# - hidden-inputs can direct input-handlers into the transform list
 +
foreach ($_REQUEST as $k => $v) {
 +
if (is_object($xwArticleXML) && preg_match("/^xw:(@?)(\\w*):(.+)$/", $k, $m)) {
 +
list(, $att, $node, $xpath) = $m;
 +
foreach (xwXPathQuery($xwArticleXML, $xpath) as $result)
 +
if ($node) {
 +
# create new element/attribute in result-node
 +
if ($att) $node = $xwArticleXML->create_attribute($node, $v);
 +
else {
 +
$node = $xwArticleXML->create_element($node);
 +
$node->set_content($v);
 +
}
 +
$result->append_child($node);
 +
}
 +
else $result->set_content($v); # node not set, just replace current-result-value
 +
}
 +
}
 +
 
 +
# if it's a pseudo-namespace...
 +
if (preg_match('/^(xml)|(sys):/i', $xwArticleTitle)) {
 +
 +
# wpSave should be valid XML to be saved
 +
if ($xwSave) {
 +
xwDomificate($wpSave);
 +
if (!is_object($wpSave)) {
 +
xwMessage('A meta-article must be valid XML! changing action to "edit"', 'red');
 +
$action='edit';
 +
}
 +
}
 +
 +
}
 +
 
 +
}
 +
 
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# OUTPUT HOOK
 +
 
 +
# Generate output to send back to client
 +
# - if article is not XML, then wiki's parsed content is used
 +
function xwOutputHook() {
 +
 
 +
global $wgOut, $wgUser, $xwParserHookCalled, $xwSkinHookCalled, $xwDebug;
 +
global $xwArticle, $xwArticleXML, $xwArticleTitle, $xwReadable, $xwWritable;
 +
global $xwView, $xwPreview, $xwSave, $xwRaw, $xwUserLinks, $xwArticleLinks;
 +
if ($xwDebug) xwMessage('xwOutputHook()','green');
 +
 
 +
# Generate output page
 +
$wgUser->setOption('skin', 'xwskin');
 +
$wgOut->output();
 +
 
 +
# NOTE - Why not move this into INPUT-HOOK???????? - or maybe not....?
 +
# If saving, check the publishing list
 +
if ($xwSave) {
 +
 +
# Check publish-list and process if any
 +
if (false) {
 +
xwMessage('Publishing...');
 +
# Publish each item
 +
foreach ($xwArticlePublish as $url) {
 +
if ($FH = fopen($url, 'a')) {
 +
# ftp example: ftp://user:password@example.com/file.txt
 +
fwrite($FH, $xwArticle);
 +
fclose($FH);
 +
xwMessage("Article published to $url");
 +
}
 +
else xwMessage("Failed to publish to $url", 'red');
 +
}
 +
}
 +
 +
# Check mirror-list and process if any
 +
# - allows articles in separate wiki's to be syncronised
 +
# - does rudimentary security by sending a salt and response being a hash of the password generated from the salt
 +
# - this could be done for login too using ECMA-MD5 on browser
 +
if (false) {
 +
xwMessage('Mirroring...');
 +
}
 +
}
 +
 
 +
 
 +
# Insert messages into final html and output
 +
if ($xwDebug) {
 +
if (!$xwParserHookCalled) xwMessage('PARSER-HOOK was not called','green');
 +
if (!$xwSkinHookCalled) xwMessage('SKIN-HOOK was not called','green');
 +
if (is_object($xwArticleXML)) xwMessage($xwArticleXML->dump_mem(true),'purple');
 +
}
 +
xwAddMessages();
 +
print $xwArticle;
 +
}
 +
 
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# PARSER HOOK
 +
 
 +
function xwParserHook(&$text) {
 +
 
 +
global $xwRaw, $xwParserHookCalled, $xwDebug, $xwArticleTitle, $xwArticle;
 +
if ($xwDebug) xwMessage('xwParserHook(TEXT)','green');
 +
$xwParserHookCalled = true;
 +
$transformed = false;
 +
 
 +
# If this text-fragment is our article, apply transforms
 +
if ((strncmp($text, $xwArticle, 100) == 0) || (strncmp($text, xwArticleContent($xwArticleTitle), 100) == 0)) {
 +
if ($xwDebug) xwMessage("PARSER-HOOK: \"$xwArticleTitle\" text-fragment recognised and intercepted.");
 +
$xwArticle = $text; # needed because xwTransform works on global article
 +
# If attempted xml encountered, try to domificate and transform
 +
if (preg_match("/^<\\?xml/i", $xwArticle)) {
 +
$transformed = true;
 +
xwDomificateArticle($xwArticle, 'ParserHook/text');
 +
xwTransformArticle();
 +
}
 +
# If not a 'raw' request, try some Geshi on
 +
if (!$xwRaw) {
 +
global $wgGeshiSyntaxSettings;
 +
if (!is_object($wgGeshiSyntaxSettings)) require_once('extensions/geshi/geshi.php');
 +
if ($lang = xwArticleType($xwArticleTitle, $xwArticle)) {
 +
if ($xwDebug) xwMessage("Geshi: This is \"$lang\".");
 +
# Replace tabs with spaces (better than geshi)
 +
$xwArticle = preg_replace_callback("/^(.*?)(\\t+)/m", 'xwReplaceTabs', $xwArticle);
 +
# Set up a geshi-parser
 +
$geshi = new GeSHi($xwArticle, $lang == 'xslt' ? 'xml' : $lang, 'extensions/geshi/geshi');
 +
$geshi->set_header_type(GESHI_HEADER_DIV);
 +
#$geshi->enable_line_numbers(GESHI_NORMAL_LINE_NUMBERS);
 +
#@$geshi->set_line_style('color:#aaaaaa');
 +
#@$geshi->set_code_style('color:black');
 +
@$geshi->set_keyword_group_style(1, 'color:blue', true);
 +
@$geshi->set_keyword_group_style(2, 'color:blue', true);
 +
@$geshi->set_keyword_group_style(3, 'color:blue', true);
 +
@$geshi->set_comments_style(1,'font-style: italic; color:#d00000;',true);
 +
@$geshi->set_comments_style(2,'font-style: italic; color:#d00000;',true);
 +
@$geshi->set_comments_style(3,'font-style: italic; color:#d00000;',true);
 +
@$geshi->set_escape_characters_style('color:red',true);
 +
@$geshi->set_brackets_style('color:black',true);
 +
@$geshi->set_strings_style('color:#008080',true);
 +
@$geshi->set_numbers_style('color:black',true);
 +
@$geshi->set_methods_style('color:black',true);
 +
@$geshi->set_symbols_style('color:black',true);
 +
@$geshi->set_regexps_style('color:green',true);
 +
# Do the parse
 +
$xwArticle = preg_replace("/^<div.*?>/", '<div class="xwcode">', $geshi->parse_code());
 +
$transformed = true;
 +
}
 +
elseif ($xwDebug) xwMessage('Geshi: Not my problem.');
 +
}
 +
$text = $xwArticle;
 +
}
 +
elseif ($xwDebug) xwMessage('Parsing other content');
 +
 +
# Return this back to wiki-parser so it can do wiki-markup if it wasn't one of ours
 +
return $transformed;
 +
}
 +
 +
# Converts tabs to spaces better than geshi
 +
function xwReplaceTabs($m) { return $m[1].str_repeat(' ',strlen($m[2])*4-strlen($m[1])%4); }
 +
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# SKIN HOOK
 +
 
 +
# If attempted xml encountered, try to domificate and transform
 +
# - Raw requests don't get here
 +
function xwSkinHook(&$tmpl) {
 +
 
 +
global $xwArticle, $xwTemplate, $xwDebug, $xwSkinHookCalled;
 +
if ($xwDebug) xwMessage('xwSkinHook(TEMPLATE)','green');
 +
$xwSkinHookCalled = true;
 +
$xwTemplate = $tmpl;
 +
$xwArticle = $xwTemplate->data['bodytext'];
 +
#$xslt = xwArticleContent('xmlwiki.xslt');
 +
 +
# Do xmlWiki page-layout transform
 +
 
 +
xwTransformPageLayout(); # should be: xwApplyXSLT($xwArticle, $xslt, 'xmlwiki.xslt');
 +
 
 +
}
 +
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# ARTICLE FUNCTIONS
 +
 
 +
# Retreive wiki-article as raw text
 +
function xwArticleContent($articleTitle) {
 +
# Return with results if already cached
 +
global $xwArticleCache, $xwDebug;
 +
if ($xwDebug) xwMessage("xwArticleContent(\"$articleTitle\")",'blue');
 +
if (isset($xwArticleCache[$articleTitle])) return $xwArticleCache[$articleTitle];
 +
# Get wiki article content (or use global if no article passed)
 +
# - using getContentWithoutUsingSoManyDamnGlobals() because getContent() fucks up nested-article-reads
 +
if ($article = new Article(Title::newFromText($articleTitle)))
 +
$article = $article->getContentWithoutUsingSoManyDamnGlobals();
 +
if ($article == '(There is currently no text in this page)') return false;
 +
if ($articleTitle) $xwArticleCache[$articleTitle] = $article;
 +
return $article;
 +
}
 +
 
 +
# Decide kind of article from content and title
 +
function xwArticleType($title, $article) {
 +
# Name based matches
 +
if (ereg('\\.php$', $title)) return 'php';
 +
if (ereg('\\.css$', $title)) return 'css';
 +
if (ereg('\\.as$', $title)) return 'actionscript';
 +
if (ereg('\\.py$', $title)) return 'python';
 +
if (ereg('\\.java$', $title)) return 'java';
 +
if (ereg('\\.((cpp)|(h))$', $title)) return 'cpp';
 +
if (ereg('\\.js$', $title)) return 'javascript';
 +
if (ereg('\\.css$', $title)) return 'css';
 +
# Content based matches
 +
if (preg_match("/^<\\?xml.+?\\?>\\s*<xsl:stylesheet/i", $article)) return 'xslt';
 +
if (ereg('^<\\?xml', $article)) return 'xml';
 +
if (ereg('^<\\?html', $article)) return 'html4strict';
 +
if (ereg('^#![/a-zA-Z0-9]+\\/perl', $article)) return 'perl';
 +
if (ereg('^#![/a-zA-Z0-9]+sh', $article)) return 'bash';
 +
}
 +
 
 +
# Convert passed article to a DOM object
 +
# - Article is unchanged if not valid XML
 +
function xwDomificateArticle(&$article, $id = '') {
 +
global $xwDebug;
 +
if ($xwDebug) xwMessage("xwDomificateArticle(XML,$id)",'blue');
 +
ob_start();
 +
if ($dom = domxml_open_mem($article)) $article = $dom;
 +
else {
 +
# Could not convert, extract error messages from output
 +
xwMessage("Failed :-(", 'red');
 +
xwExtractMessages();
 +
}
 +
ob_end_clean();
 +
}
 +
 
 +
# Reduce article to a string if it's a DOM object
 +
# - ie if all xslt's had xml-output-method
 +
function xwUndomificateArticle(&$article, $id = '') {
 +
if (!is_object($article)) return false;
 +
global $xwDebug;
 +
if ($xwDebug) xwMessage("xwUndomificate(DOM,$id)",'blue');
 +
# TODO:
 +
# Validate, report errors
 +
# xwApplyXSLT($article, $xslt) if xml referrs one
 +
# Convert DOM back to XML if still an object
 +
if (is_object($article)) $article = $article->dump_mem(true);
 +
return true;
 +
}
 +
 
 +
# Transform an article object
 +
# - Article is undomificated on exit
 +
# - Article remains unchanged if not DOM
 +
function xwTransformArticle() {
 +
global $xwDebug, $xwArticle, $xwArticleXML, $xwArticleTitle;
 +
# xml:article must be an object...
 +
if (is_object($xwArticleXML)) {
 +
$docType = $xwArticleXML->doctype();
 +
# ...and must be xmlwiki:* doctype
 +
if (ereg('^xmlwiki:', $docType->name)) {
 +
$root = $xwArticle->document_element();
 +
if ($xwDebug) xwMessage("xwTransformArticle()",'blue');
 +
# Loop thru transforms applying each
 +
foreach ($root->get_elements_by_tagname('transform') as $tElement) {
 +
$tName = $tElement->get_content();
 +
# Transform is not local, get transform-article and apply to current-article-object
 +
if ($tText = xwArticleContent($tName)) {
 +
# Get transform type
 +
$tType = xwArticleType($tName, $tText);
 +
if ($tType == 'xslt') xwApplyXSLT($xwArticle, $tText, $tName);
 +
elseif ($tType == 'php') {
 +
# TODO: php, check perms and execute (trapping errors for reporting - unless fatal)
 +
if ($perms_check_out_ok = true) {
 +
ob_start();
 +
eval("?>$tText<?");
 +
xwExtractMessages();
 +
ob_end_clean();
 +
}
 +
}
 +
elseif ($tType == 'css') xwMessage("CSS's from transform-list aren't applied yet :-(", 'red');
 +
}
 +
else xwMessage("No such transform-article \"$tName\"!", 'red');
 +
}
 +
}
 +
else xwMessage("Could not transform: xml:article must be doctype:xmlwiki!",'red');
 +
}
 +
else xwMessage("xwTransformArticle(NON-OBJECT)",'red');
 +
# If article is still a DOM object, stringify it once and for all
 +
if (is_object($xwArticle)) xwUndomificateArticle($xwArticle, $xwArticleTitle);
 +
return true;
 +
}
 +
 
 +
# Apply XSLT to article-dom
 +
# - if output-method is html, article will get stringified
 +
function xwApplyXSLT(&$article, &$xslt, $id = '') {
 +
global $xwDebug;
 +
if ($xwDebug) xwMessage("xwApplyXSLT(DOM,$id)",'blue');
 +
ob_start();
 +
if ($tObject = domxml_xslt_stylesheet($xslt)) {
 +
$tResult = $tObject->process($article);
 +
# If output-method is html, use XSLT's dump_mem to create an html-string
 +
if (preg_match('/<xsl:output +?method *= *"html"/i', $xslt))
 +
$tResult = $tObject->result_dump_mem($tResult);
 +
}
 +
xwExtractMessages();
 +
ob_end_clean();
 +
if (isset($tResult)) $article = $tResult; else return false;
 +
return true;
 +
}
 +
 
 +
# Return result set of an XPath query on passed DOM-object
 +
function xwXPathQuery(&$dom, $query) {
 +
global $xwDebug;
 +
if ($xwDebug) xwMessage("xwXPathQuery(DOM,\"$query\")",'blue');
 +
ob_start();
 +
$context = $dom->xpath_new_context();
 +
if ($xpath = $context->xpath_eval($query)) $result = $xpath->nodeset;
 +
else {
 +
xwMessage('Query Failed! :-(', 'red');
 +
$result = array();
 +
}
 +
xwExtractMessages();
 +
ob_end_clean();
 +
return $result;
 +
}
 +
 
 +
# Surrounds the article with HTML structural page-layout containing a ref to xmlwiki.css
 +
# - TEMP: see skin-hook
 +
function xwTransformPageLayout() {
 +
 
 +
global $xwMsgToken, $xwArticle, $xwTemplate, $xwDebug, $xwWritable;
 +
if ($xwDebug) xwMessage("xwTransformPageLayout()",'blue');
 +
$tmpl = &$xwTemplate;
 +
$data = &$tmpl->data;
 +
$doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" ';
 +
$doctype .= '"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">';
 +
$html = '<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="';
 +
$html .= htmlspecialchars($data['lang']).'" lang="'.htmlspecialchars($data['lang']).'" dir="';
 +
$html .= htmlspecialchars($data['dir']).'">';
 +
 
 +
# Head
 +
$head = '<head><meta http-equiv="Content-Type" content="'.htmlspecialchars($data['mimetype']);
 +
$head .= '; charset='.htmlspecialchars($data['charset']).'" />'.$data['headlinks'];
 +
$head .= '<link rel="stylesheet" type="text/css" href="/wiki/index.php?title=xmlwiki.css&action=raw&gen=css" />';
 +
 
 +
$head .= '<title>'.htmlspecialchars($data['pagetitle']).'</title>';
 +
$head .= '</head>';
 +
 
 +
# Body opening tag
 +
$body = '<body ';
 +
if ($data['body_ondblclick']) $body .= 'ondblclick="'.htmlspecialchars($data['body_ondblclick']).'" ';
 +
if ($data['nsclass']) $body .= 'class="'.htmlspecialchars($data['nsclass']).'"';
 +
$body .= '>';
 +
 
 +
# Content
 +
$content = '<div id="column-content">';
 +
$content .= '<div id="content"><a name="top" id="contentTop"></a>';
 +
if ($data['sitenotice']) $content .= '<div id="siteNotice">'.$data['sitenotice'].'</div>';
 +
$content .= '<h1 class="firstHeading">'.htmlspecialchars($data['title']).'</h1>';
 +
$content .= '<div id="bodyContent">';
 +
$content .= '<h3 id="siteSub">'.htmlspecialchars($tmpl->translator->translate('tagline')).'</h3>';
 +
$content .= '<div id="contentSub">'.$data['subtitle'].'</div>';
 +
if ($data['undelete']) $content .= '<div id="contentSub">'.$data['undelete'].'</div>';
 +
if ($data['newtalk']) $content .= '<div class="usermessage">'.$data['newtalk'].'</div>';
 +
$content .= "<!-- start content -->$xwArticle"; # <---- Actual article content inserted here
 +
if ($data['catlinks']) $content .= '<div id="catlinks">'.$data['catlinks'].'</div>';
 +
$content .= '<!-- end content --><div class="visualClear"></div>';
 +
$content .= '</div></div></div>';
 +
 
 +
# Action-links (remove edit link if not writable)
 +
$actions = '<div id="p-cactions" class="portlet"><h5>Views</h5><ul>';
 +
foreach ($data['content_actions'] as $key => $action) {
 +
#if ($key == '
 +
#if (!(ereg('action=edit$', $action['href'])&&!$xwWritable)) {
 +
$actions .= '<li id="ca-'.htmlspecialchars($key).'"';
 +
if ($action['class']) $actions .= ' class="'.htmlspecialchars($action['class']).'"';
 +
$href = '<a href="'.htmlspecialchars($action['href']).'">'.htmlspecialchars($action['text']).'</a>';
 +
$actions .= ">$href</li>";
 +
# }
 +
}
 +
$actions .= '</ul></div>';
 +
 
 +
# Personal-links
 +
$personal = '<div class="portlet" id="p-personal">';
 +
$personal .= '<h5>'.htmlspecialchars($tmpl->translator->translate('personaltools')).'</h5>';
 +
$personal .= '<div class="pBody"><ul>';
 +
foreach ($data['personal_urls'] as $key => $item) {
 +
$personal .= '<li id="pt-'.htmlspecialchars($key).'">';
 +
$href = '<a href="'.htmlspecialchars($item['href']).'"';
 +
if (!empty($item['class'])) $href .= 'class="'.htmlspecialchars($item['class']).'"';
 +
$href .= '>'.htmlspecialchars($item['text']).'</a>';
 +
$personal .= "$href</li>";
 +
}
 +
$personal .= "</ul></div></div>";
 +
 
 +
# Logo
 +
$logo = '<div class="portlet" id="p-logo">';
 +
$logo .= '<a style="background-image: url('.htmlspecialchars($data['logopath']).');" ';
 +
$logo .= 'href="'.htmlspecialchars($data['nav_urls']['mainpage']['href']).'" ';
 +
$logo .= 'title="'.htmlspecialchars($tmpl->translator->translate('mainpage')).'"></a></div>';
 +
$logo .= '<script type="text/javascript"> if (window.isMSIE55) fixalpha(); </script>';
 +
 
 +
# Navigation
 +
$nav = '<div class="portlet" id="p-nav">';
 +
$nav .= '<h5>'.htmlspecialchars($tmpl->translator->translate('navigation')).'</h5>';
 +
$nav .= '<div class="pBody"><ul>';
 +
foreach ($data['navigation_urls'] as $navlink) {
 +
$nav .= "<li id=\"".htmlspecialchars($navlink['id'])."\">";
 +
$nav .= '<a href="'.htmlspecialchars($navlink['href']).'">'.htmlspecialchars($navlink['text']).'</a>';
 +
$nav .= '</li>';
 +
}
 +
$nav .= '</ul></div></div>';
 +
 +
# Search
 +
$search = '<div id="p-search" class="portlet">';
 +
$search .= '<h5><label for="searchInput">'.htmlspecialchars($tmpl->translator->translate('search')).'</label></h5>';
 +
$search .= '<div class="pBody">';
 +
$search .= '<form name="searchform" action="'.htmlspecialchars($data['searchaction']).'" id="searchform">';
 +
$search .= '<input id="searchInput" name="search" type="text"';
 +
if ($tmpl->haveMsg('accesskey-search')) $search .= ' accesskey="';
 +
$search .= htmlspecialchars($tmpl->translator->translate('accesskey-search')).'"';
 +
if (isset($data['search'])) $search .= ' value="'.htmlspecialchars($data['search']).'"';
 +
$search .= ' />';
 +
$search .= '<input type="submit" name="go" class="searchButton" id="searchGoButton" value="';
 +
$search .= htmlspecialchars($tmpl->translator->translate('go')).'" />';
 +
$search .= '&nbsp;<input type="submit" name="fulltext" class="searchButton" value="';
 +
$search .= htmlspecialchars($tmpl->translator->translate('search')).'" />';
 +
$search .= '</form></div></div>';
 +
 +
# Toolbox
 +
$toolbox = '<div class="portlet" id="p-tb">';
 +
$toolbox .= '<h5>'.htmlspecialchars($tmpl->translator->translate('toolbox')).'</h5>';
 +
$toolbox .= '<div class="pBody"><ul>';
 +
if ($data['notspecialpage']) {
 +
foreach (array('whatlinkshere', 'recentchangeslinked') as $special ) {
 +
$href = '<a href="'.htmlspecialchars($data['nav_urls'][$special]['href']).'">';
 +
$href .= htmlspecialchars($tmpl->translator->translate($special)).'</a>';
 +
$toolbox .= "<li id=\"t-'$special'\">$href</li>";
 +
}
 +
}
 +
if ($data['feeds']) {
 +
$toolbox .= '<li id="feedlinks">';
 +
foreach ($data['feeds'] as $key => $feed) {
 +
$toolbox .= '<span id="feed-'.htmlspecialchars($key).'">';
 +
$toolbox .= '<a href="'.htmlspecialchars($feed['href']).'">'.htmlspecialchars($feed['text']).'</a>';
 +
$toolbox .= '&nbsp;</span>';
 +
}
 +
$toolbox .= '</li>';
 +
}
 +
foreach (array('contributions', 'emailuser', 'upload', 'specialpages') as $special ) {
 +
if ($data['nav_urls'][$special]) {
 +
$href = '<a href="'.htmlspecialchars($data['nav_urls'][$special]['href']).'">';
 +
$href .= htmlspecialchars($tmpl->translator->translate($special)).'</a>';
 +
$toolbox .= "<li id=\"t-$special\">$href</li>";
 +
}
 +
}
 +
$toolbox .= '</ul></div></div>';
 +
 
 +
# Language
 +
$language = '';
 +
if ($data['language_urls']) {
 +
$language .= '<div id="p-lang" class="portlet">';
 +
$language .= '<h5>'.htmlspecialchars($tmpl->translator->translate('otherlanguages')).'</h5>';
 +
$language .= '<div class="pBody"><ul>';
 +
foreach ($data['language_urls'] as $langlink) {
 +
$href = '<a href="'.htmlspecialchars($langlink['href']).'">'.$langlink['text'].'</a>';
 +
$language .= "<li>$href</li>";
 +
}
 +
$language .= '</ul></div></div>';
 +
}
 +
 
 +
# Footer
 +
$footer = '<div id="footer"><ul id="f-list">';
 +
if ($data['lastmod']) $footer .= '<li id="f-lastmod">'.$data['lastmod'].'</li>';
 +
if ($data['viewcount']) $footer .= '<li id="f-viewcount">'.$data['viewcount'].'</li>';
 +
if ($data['credits']) $footer .= '<li id="f-credits">'.$data['credits'].'</li>';
 +
if ($data['copyright']) $footer .= '<li id="f-copyright">'.$data['copyright'].'</li>';
 +
if ($data['about']) $footer .= '<li id="f-about">'.$data['about'].'</li>';
 +
if ($data['disclaimer']) $footer .= '<li id="f-disclaimer">'.$data['disclaimer'].'</li>';
 +
$footer .= '</ul></div>';
 +
 
 +
# Return assembled parts
 +
$xwArticle = " $doctype
 +
$html
 +
$head
 +
$body
 +
<table border=\"0\" cellpadding=\"10\" cellspacing=\"1\">
 +
<tr>
 +
<td valign=\"top\" bgcolor=\"#cccccc\">
 +
<img src=\"/wiki/xmlwiki.jpg\" />
 +
$search
 +
$nav
 +
$toolbox
 +
</td>
 +
<td width=\"100%\" valign=\"top\" bgcolor=\"#cccccc\">
 +
$personal$actions$xwMsgToken$content
 +
</td>
 +
</tr>
 +
<td colspan=\"2\" bgcolor=\"#cccccc\">$footer</td>
 +
<tr>
 +
</table>
 +
</body>
 +
</html>
 +
";
 +
}
 +
 +
# ---------------------------------------------------------------------------------------------------------------------- #
 +
# UTILITY FUNCTIONS
 +
 
 +
# Append messages to content
 +
function xwAddMessages() {
 +
global $xwArticle, $xwMessages, $xwMsgToken;
 +
if (count($xwMessages)) {
 +
$msg = '<div class="portlet" id="p-xwmessages"><h5>There are Messages</h5></div>';
 +
$msg .= '<div class="xwmessage"><ul><li>'.join("</li><li>", $xwMessages).'</li></ul></div>';
 +
$xwArticle = ereg_replace($xwMsgToken, $msg, $xwArticle);
 +
}
 +
}
 +
 
 +
# Extract error-messages from captured output and put in proper message-queue
 +
function xwExtractMessages() {
 +
$err = preg_replace("/<.+?>/", "", ob_get_contents());
 +
$err = preg_replace('/ in .+? on line.+?[0-9]+/', '', $err);
 +
$err = preg_replace('/Warning.+?\\(\\): /', '', $err);
 +
foreach (split("\n", $err) as $msg) if ($msg) xwMessage("$msg", 'red');
 +
}
 +
 +
# Add message to queue
 +
function xwMessage($msg, $col = '#000080') {
 +
global $xwMessages;
 +
$msg = htmlentities($msg);
 +
return $xwMessages[] = "<font color=\"$col\">$msg</font>";
 +
}
 +
 
 +
# Print list of passed object's methods and properties
 +
function xwCheckoutObject($obj) {
 +
if (is_object($obj)) {
 +
print strtoupper("<br>$obj properties:<br>");
 +
foreach (get_object_vars($obj) as $k=>$v) print htmlentities("$k => $v")."<br>";
 +
print strtoupper("<br>$obj methods:<br>");
 +
foreach (get_class_methods($obj) as $k=>$v) print "$v<br>";
 +
}
 +
if (is_array($obj)) {
 +
print strtoupper("<br>Array content:<br>");
 +
foreach ($obj as $k=>$v) print "$k => $v<br>";
 +
}
 +
die;
 +
}
 +
 +
?>

Revision as of 22:33, 18 June 2005

<?php

  1. xmlWiki - MediaWiki XML Hack
  2. Nad - 2005-05-18
  1. INIT
  2. SECURITY
  3. INPUT HOOK
  4. OUTPUT HOOK
  5. PARSER HOOK
  6. SKIN HOOK
  7. ARTICLE FUNCTIONS
  8. UTILITY FUNCTIONS
  1. ---------------------------------------------------------------------------------------------------------------------- #
  1. TODO:
  2. - Undomificate: validation, xslt
  3. - Transform: check perms on php-execution
  4. Publish:
  5. - xml:article - holds article publish-list
  6. - mirror/wiki-sync
  7. Transforms:
  8. - xml:article - now transforms are in here (main article is now like <body>)
  9. - so we don't do the "its one of ours" thing anymore
  1. LATER:
  2. - allow XIncludes to build large docs from others
  3. - allow docBook and DSSSL content
  4. - Maybe only use XML transforms and use PHP via PI's
  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. INIT
  1. Exit if not included from index.php

defined('MEDIAWIKI') or die('xmlwiki.php must be included from MediaWiki\'s index.php!');

  1. Allows Geshi to be installed by PARSER-HOOK if not installed th usual way already

if (!@is_object($wgGeshiSyntaxSettings)) $wgGeshiSyntaxSettings = false;

  1. Otherwise set up xmlWiki global environment ready for input and output processing

$xwMessages = array(); $xwEdit = isset($_REQUEST['action']) && ($_REQUEST['action'] == 'edit'); $xwView = (!isset($_REQUEST['action']) || (isset($_REQUEST['action']) && ($_REQUEST['action'] == 'view'))); $xwRaw = isset($_REQUEST['action']) && ($_REQUEST['action'] == 'raw'); $xwPreview = isset($_REQUEST['wpPreview']); $xwSave = isset($_REQUEST['wpSave']); $xwUserName = $wgUser->mName; $xwUserGroups = array(strtolower($xwUserName), 'anyone', 'everyone', 'nobody', 'world'); $xwArticleReadableBy = array('anyone'); $xwArticleWritableBy = array('anyone'); $xwArticle = $xwPreview ? $wgRequest->gettext('wpTextbox1') : $wgArticle->getContent(false); $xwArticleTitle = $wgTitle->getPrefixedURL(); $xwArticleCache = array(); $xwArticleLinks = array(); $xwUserLinks = array(); $xwMetaArticle = false; $xwMsgToken = ; $xwTemplate = null; $xwParserHookCalled = false; $xwSkinHookCalled = false;

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. SECURITY
  1. Not currently used: R/W by user, for xmlWiki prefs etc

if ($xwUserXML = xwArticleContent(strtolower("xml:$xwUserName"))) xwDomificateArticle($xwUserXML, "xml:$xwUserName");

  1. R/W by admin only, holds users groups and other system info like stats
  2. - no sys:user means "anyone" in group "everyone" as far as perms goes

if ($xwUserSYS = xwArticleContent(strtolower("sys:$xwUserName"))) xwDomificateArticle($xwUserSYS, "sys:$xwUserName");

  1. If sys:user exists, get users' group-list

if (is_object($xwUserSYS)) { $xwRoot = $xwUserSYS->document_element(); if (count($xwList = $xwRoot->get_elements_by_tagname('groups'))) $xwUserGroups = array_merge($xwUserGroups, split(',', strtolower($xwList[0]->get_content()))); }

  1. - Security info is: owner, read-groups, write-groups (default values are "world")
  2. - Users and groups are interchangeable
  3. - Default user is "anyone", default group is "everyone"

if (eregi('^xml:(.+)$', $xwArticleTitle)) { # This is an xml-pseudo-namespace # - its perms apply to its 'parent' and to itself aswell $xwXML = $xwArticle; xwDomificateArticle($xwXML, $xwArticleTitle); if (is_object($xwXML)) { $xwRoot = $xwXML->document_element(); if (count($xwList = $xwRoot->get_elements_by_tagname('read'))) $xwArticleReadableBy = split(',', strtolower($xwList[0]->get_content())); if (count($xwList = $xwRoot->get_elements_by_tagname('write'))) $xwArticleWritableBy = split(',', strtolower($xwList[0]->get_content())); } } elseif (eregi('^sys:(.+)$', $xwArticleTitle)) { # This is a sys-pseudo-namespace # - only admin can read/write sys: $xwArticleReadableBy = $xwArticleWritableBy = array('admin'); } else { # This is a normal article, get xml:article if ($xwArticleXML = xwArticleContent(strtolower("xml:$xwArticleTitle"))) xwDomificateArticle($xwArticleXML, "xml:$xwArticleTitle"); if (is_object($xwArticleXML)) { $xwRoot = $xwArticleXML->document_element(); if (count($xwList = $xwRoot->get_elements_by_tagname('read'))) $xwArticleReadableBy = split(',', $xwList = strtolower($xwList[0]->get_content())); if (count($xwList = $xwRoot->get_elements_by_tagname('write'))) $xwArticleWritableBy = split(',', $xwList = strtolower($xwList[0]->get_content())); } }

  1. Set perms for this request

$xwReadable = 0 < count(array_intersect($xwArticleReadableBy, $xwUserGroups)); $xwWritable = 0 < count(array_intersect($xwArticleWritableBy, $xwUserGroups)); if ($xwDebug) { xwMessage('Groups: '.join(', ', $xwUserGroups)); xwMessage("Readable ($xwReadable): ".join(', ', $xwArticleReadableBy)); xwMessage("Writable ($xwWritable): ".join(', ', $xwArticleWritableBy)); }

  1. Divert to access-denied article if not readable

if (!$xwReadable) { $action = 'view'; $xwSave = $xwEdit = false; die('Sorry, the requested article was not readable.'); }


  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. INPUT HOOK
  1. Parse and process input from forms

function xwInputHook() {

global $xwArticle, $xwArticleXML, $xwArticleTitle, $xwWritable, $_REQUEST; global $xwDebug, $xwSave, $xwEdit, $action, $xwUserName, $wgArticle; if ($xwDebug) xwMessage('xwInputHook()','green');

# If not writable, change action to view if (!$xwWritable && ($action != 'view')) { xwMessage('Sorry, article not writable, action changed to "view".', 'red'); $action = 'view'; $xwSave = $xwEdit = false; }

# Parse POST for XPath inputs # - Form-input-name is "xw:[[@]node]:XPathQuery" which directs the value into xwArticle (if DOM) # - if "node" exists, then a new element (or att if "@node") is created for the value # - hidden-inputs can direct input-handlers into the transform list foreach ($_REQUEST as $k => $v) { if (is_object($xwArticleXML) && preg_match("/^xw:(@?)(\\w*):(.+)$/", $k, $m)) { list(, $att, $node, $xpath) = $m; foreach (xwXPathQuery($xwArticleXML, $xpath) as $result) if ($node) { # create new element/attribute in result-node if ($att) $node = $xwArticleXML->create_attribute($node, $v); else { $node = $xwArticleXML->create_element($node); $node->set_content($v); } $result->append_child($node); } else $result->set_content($v); # node not set, just replace current-result-value } }

# if it's a pseudo-namespace... if (preg_match('/^(xml)|(sys):/i', $xwArticleTitle)) {

# wpSave should be valid XML to be saved if ($xwSave) { xwDomificate($wpSave); if (!is_object($wpSave)) { xwMessage('A meta-article must be valid XML! changing action to "edit"', 'red'); $action='edit'; } }

}

}

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. OUTPUT HOOK
  1. Generate output to send back to client
  2. - if article is not XML, then wiki's parsed content is used

function xwOutputHook() {

global $wgOut, $wgUser, $xwParserHookCalled, $xwSkinHookCalled, $xwDebug; global $xwArticle, $xwArticleXML, $xwArticleTitle, $xwReadable, $xwWritable; global $xwView, $xwPreview, $xwSave, $xwRaw, $xwUserLinks, $xwArticleLinks; if ($xwDebug) xwMessage('xwOutputHook()','green');

# Generate output page $wgUser->setOption('skin', 'xwskin'); $wgOut->output();

# NOTE - Why not move this into INPUT-HOOK???????? - or maybe not....? # If saving, check the publishing list if ($xwSave) {

# Check publish-list and process if any if (false) { xwMessage('Publishing...'); # Publish each item foreach ($xwArticlePublish as $url) { if ($FH = fopen($url, 'a')) { # ftp example: ftp://user:password@example.com/file.txt fwrite($FH, $xwArticle); fclose($FH); xwMessage("Article published to $url"); } else xwMessage("Failed to publish to $url", 'red'); } }

# Check mirror-list and process if any # - allows articles in separate wiki's to be syncronised # - does rudimentary security by sending a salt and response being a hash of the password generated from the salt # - this could be done for login too using ECMA-MD5 on browser if (false) { xwMessage('Mirroring...'); } }


# Insert messages into final html and output if ($xwDebug) { if (!$xwParserHookCalled) xwMessage('PARSER-HOOK was not called','green'); if (!$xwSkinHookCalled) xwMessage('SKIN-HOOK was not called','green'); if (is_object($xwArticleXML)) xwMessage($xwArticleXML->dump_mem(true),'purple'); } xwAddMessages(); print $xwArticle; }

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. PARSER HOOK

function xwParserHook(&$text) {

global $xwRaw, $xwParserHookCalled, $xwDebug, $xwArticleTitle, $xwArticle; if ($xwDebug) xwMessage('xwParserHook(TEXT)','green'); $xwParserHookCalled = true; $transformed = false;

# If this text-fragment is our article, apply transforms if ((strncmp($text, $xwArticle, 100) == 0) || (strncmp($text, xwArticleContent($xwArticleTitle), 100) == 0)) { if ($xwDebug) xwMessage("PARSER-HOOK: \"$xwArticleTitle\" text-fragment recognised and intercepted."); $xwArticle = $text; # needed because xwTransform works on global article # If attempted xml encountered, try to domificate and transform if (preg_match("/^<\\?xml/i", $xwArticle)) { $transformed = true; xwDomificateArticle($xwArticle, 'ParserHook/text'); xwTransformArticle(); } # If not a 'raw' request, try some Geshi on if (!$xwRaw) { global $wgGeshiSyntaxSettings; if (!is_object($wgGeshiSyntaxSettings)) require_once('extensions/geshi/geshi.php'); if ($lang = xwArticleType($xwArticleTitle, $xwArticle)) { if ($xwDebug) xwMessage("Geshi: This is \"$lang\"."); # Replace tabs with spaces (better than geshi) $xwArticle = preg_replace_callback("/^(.*?)(\\t+)/m", 'xwReplaceTabs', $xwArticle); # Set up a geshi-parser $geshi = new GeSHi($xwArticle, $lang == 'xslt' ? 'xml' : $lang, 'extensions/geshi/geshi'); $geshi->set_header_type(GESHI_HEADER_DIV); #$geshi->enable_line_numbers(GESHI_NORMAL_LINE_NUMBERS); #@$geshi->set_line_style('color:#aaaaaa'); #@$geshi->set_code_style('color:black'); @$geshi->set_keyword_group_style(1, 'color:blue', true); @$geshi->set_keyword_group_style(2, 'color:blue', true); @$geshi->set_keyword_group_style(3, 'color:blue', true); @$geshi->set_comments_style(1,'font-style: italic; color:#d00000;',true); @$geshi->set_comments_style(2,'font-style: italic; color:#d00000;',true); @$geshi->set_comments_style(3,'font-style: italic; color:#d00000;',true); @$geshi->set_escape_characters_style('color:red',true); @$geshi->set_brackets_style('color:black',true); @$geshi->set_strings_style('color:#008080',true); @$geshi->set_numbers_style('color:black',true); @$geshi->set_methods_style('color:black',true); @$geshi->set_symbols_style('color:black',true); @$geshi->set_regexps_style('color:green',true); # Do the parse

$xwArticle = preg_replace("/^<div.*?>/", '

', $geshi->parse_code());

$transformed = true; } elseif ($xwDebug) xwMessage('Geshi: Not my problem.'); } $text = $xwArticle; } elseif ($xwDebug) xwMessage('Parsing other content');

# Return this back to wiki-parser so it can do wiki-markup if it wasn't one of ours return $transformed; }

  1. Converts tabs to spaces better than geshi

function xwReplaceTabs($m) { return $m[1].str_repeat(' ',strlen($m[2])*4-strlen($m[1])%4); }

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. SKIN HOOK
  1. If attempted xml encountered, try to domificate and transform
  2. - Raw requests don't get here

function xwSkinHook(&$tmpl) {

global $xwArticle, $xwTemplate, $xwDebug, $xwSkinHookCalled; if ($xwDebug) xwMessage('xwSkinHook(TEMPLATE)','green'); $xwSkinHookCalled = true; $xwTemplate = $tmpl; $xwArticle = $xwTemplate->data['bodytext']; #$xslt = xwArticleContent('xmlwiki.xslt');

# Do xmlWiki page-layout transform

xwTransformPageLayout(); # should be: xwApplyXSLT($xwArticle, $xslt, 'xmlwiki.xslt');

}

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. ARTICLE FUNCTIONS
  1. Retreive wiki-article as raw text

function xwArticleContent($articleTitle) { # Return with results if already cached global $xwArticleCache, $xwDebug; if ($xwDebug) xwMessage("xwArticleContent(\"$articleTitle\")",'blue'); if (isset($xwArticleCache[$articleTitle])) return $xwArticleCache[$articleTitle]; # Get wiki article content (or use global if no article passed) # - using getContentWithoutUsingSoManyDamnGlobals() because getContent() fucks up nested-article-reads if ($article = new Article(Title::newFromText($articleTitle))) $article = $article->getContentWithoutUsingSoManyDamnGlobals(); if ($article == '(There is currently no text in this page)') return false; if ($articleTitle) $xwArticleCache[$articleTitle] = $article; return $article; }

  1. Decide kind of article from content and title

function xwArticleType($title, $article) { # Name based matches if (ereg('\\.php$', $title)) return 'php'; if (ereg('\\.css$', $title)) return 'css'; if (ereg('\\.as$', $title)) return 'actionscript'; if (ereg('\\.py$', $title)) return 'python'; if (ereg('\\.java$', $title)) return 'java'; if (ereg('\\.((cpp)|(h))$', $title)) return 'cpp'; if (ereg('\\.js$', $title)) return 'javascript'; if (ereg('\\.css$', $title)) return 'css'; # Content based matches if (preg_match("/^<\\?xml.+?\\?>\\s*<xsl:stylesheet/i", $article)) return 'xslt'; if (ereg('^<\\?xml', $article)) return 'xml'; if (ereg('^<\\?html', $article)) return 'html4strict'; if (ereg('^#![/a-zA-Z0-9]+\\/perl', $article)) return 'perl'; if (ereg('^#![/a-zA-Z0-9]+sh', $article)) return 'bash'; }

  1. Convert passed article to a DOM object
  2. - Article is unchanged if not valid XML

function xwDomificateArticle(&$article, $id = ) { global $xwDebug; if ($xwDebug) xwMessage("xwDomificateArticle(XML,$id)",'blue'); ob_start(); if ($dom = domxml_open_mem($article)) $article = $dom; else { # Could not convert, extract error messages from output xwMessage("Failed :-(", 'red'); xwExtractMessages(); } ob_end_clean(); }

  1. Reduce article to a string if it's a DOM object
  2. - ie if all xslt's had xml-output-method

function xwUndomificateArticle(&$article, $id = ) { if (!is_object($article)) return false; global $xwDebug; if ($xwDebug) xwMessage("xwUndomificate(DOM,$id)",'blue'); # TODO: # Validate, report errors # xwApplyXSLT($article, $xslt) if xml referrs one # Convert DOM back to XML if still an object if (is_object($article)) $article = $article->dump_mem(true); return true; }

  1. Transform an article object
  2. - Article is undomificated on exit
  3. - Article remains unchanged if not DOM

function xwTransformArticle() { global $xwDebug, $xwArticle, $xwArticleXML, $xwArticleTitle; # xml:article must be an object... if (is_object($xwArticleXML)) { $docType = $xwArticleXML->doctype(); # ...and must be xmlwiki:* doctype if (ereg('^xmlwiki:', $docType->name)) { $root = $xwArticle->document_element(); if ($xwDebug) xwMessage("xwTransformArticle()",'blue'); # Loop thru transforms applying each foreach ($root->get_elements_by_tagname('transform') as $tElement) { $tName = $tElement->get_content(); # Transform is not local, get transform-article and apply to current-article-object if ($tText = xwArticleContent($tName)) { # Get transform type $tType = xwArticleType($tName, $tText); if ($tType == 'xslt') xwApplyXSLT($xwArticle, $tText, $tName); elseif ($tType == 'php') { # TODO: php, check perms and execute (trapping errors for reporting - unless fatal) if ($perms_check_out_ok = true) { ob_start(); eval("?>$tText<?"); xwExtractMessages(); ob_end_clean(); } } elseif ($tType == 'css') xwMessage("CSS's from transform-list aren't applied yet :-(", 'red'); } else xwMessage("No such transform-article \"$tName\"!", 'red'); } } else xwMessage("Could not transform: xml:article must be doctype:xmlwiki!",'red'); } else xwMessage("xwTransformArticle(NON-OBJECT)",'red'); # If article is still a DOM object, stringify it once and for all if (is_object($xwArticle)) xwUndomificateArticle($xwArticle, $xwArticleTitle); return true; }

  1. Apply XSLT to article-dom
  2. - if output-method is html, article will get stringified

function xwApplyXSLT(&$article, &$xslt, $id = ) { global $xwDebug; if ($xwDebug) xwMessage("xwApplyXSLT(DOM,$id)",'blue'); ob_start(); if ($tObject = domxml_xslt_stylesheet($xslt)) { $tResult = $tObject->process($article); # If output-method is html, use XSLT's dump_mem to create an html-string if (preg_match('/<xsl:output +?method *= *"html"/i', $xslt)) $tResult = $tObject->result_dump_mem($tResult); } xwExtractMessages(); ob_end_clean(); if (isset($tResult)) $article = $tResult; else return false; return true; }

  1. Return result set of an XPath query on passed DOM-object

function xwXPathQuery(&$dom, $query) { global $xwDebug; if ($xwDebug) xwMessage("xwXPathQuery(DOM,\"$query\")",'blue'); ob_start(); $context = $dom->xpath_new_context(); if ($xpath = $context->xpath_eval($query)) $result = $xpath->nodeset; else { xwMessage('Query Failed! :-(', 'red'); $result = array(); } xwExtractMessages(); ob_end_clean(); return $result; }

  1. Surrounds the article with HTML structural page-layout containing a ref to xmlwiki.css
  2. - TEMP: see skin-hook

function xwTransformPageLayout() {

global $xwMsgToken, $xwArticle, $xwTemplate, $xwDebug, $xwWritable; if ($xwDebug) xwMessage("xwTransformPageLayout()",'blue'); $tmpl = &$xwTemplate; $data = &$tmpl->data; $doctype = '<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" '; $doctype .= '"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">'; $html = ''; # Head $head = ''.$data['headlinks']; $head .= ''; $head .= ''.htmlspecialchars($data['pagetitle']).''; $head .= ''; # Body opening tag $body = ''; $content .= '

'.htmlspecialchars($data['title']).'

'; $content .= '
'; $content .= '

'.htmlspecialchars($tmpl->translator->translate('tagline')).'

'; $content .= '
'.$data['subtitle'].'
'; if ($data['undelete']) $content .= '
'.$data['undelete'].'
'; if ($data['newtalk']) $content .= '
'.$data['newtalk'].'
'; $content .= "$xwArticle"; # <---- Actual article content inserted here if ($data['catlinks']) $content .= ''; $content .= '
'; $content .= '

'; # Action-links (remove edit link if not writable) $actions = '

Views
    '; foreach ($data['content_actions'] as $key => $action) { #if ($key == ' #if (!(ereg('action=edit$', $action['href'])&&!$xwWritable)) { $actions .= '
  • '.htmlspecialchars($action['text']).''; $actions .= ">$href
    • "; # } } $actions .= '

'; # Personal-links $personal = '

'; $personal .= '
'.htmlspecialchars($tmpl->translator->translate('personaltools')).'
'; $personal .= '

"; # Logo $logo = '

MastodonRSSCodeEmail

'; $logo .= '

'; # Navigation $nav = '

'; $nav .= '
'.htmlspecialchars($tmpl->translator->translate('navigation')).'
'; $nav .= '

'; # Search $search = '

'; # Toolbox $toolbox = '

'; $toolbox .= '
'.htmlspecialchars($tmpl->translator->translate('toolbox')).'
'; $toolbox .= '

'; # Language $language = ''; if ($data['language_urls']) { $language .= '

'; $language .= '
'.htmlspecialchars($tmpl->translator->translate('otherlanguages')).'
'; $language .= '
    '; foreach ($data['language_urls'] as $langlink) { $href = ''.$langlink['text'].''; $language .= "
  • $href
    • "; } $language .= '

'; } # Footer $footer = '

'; # Return assembled parts $xwArticle = " $doctype $html $head $body

$search $nav $toolbox $personal$actions$xwMsgToken$content
$footer

"; }

  1. ---------------------------------------------------------------------------------------------------------------------- #
  2. UTILITY FUNCTIONS
  1. Append messages to content

function xwAddMessages() { global $xwArticle, $xwMessages, $xwMsgToken; if (count($xwMessages)) {

$msg = '

There are Messages

'; $msg .= '

  • '.join("
  • ", $xwMessages).'

';

$xwArticle = ereg_replace($xwMsgToken, $msg, $xwArticle); } }

  1. Extract error-messages from captured output and put in proper message-queue

function xwExtractMessages() { $err = preg_replace("/<.+?>/", "", ob_get_contents()); $err = preg_replace('/ in .+? on line.+?[0-9]+/', , $err); $err = preg_replace('/Warning.+?\\(\\): /', , $err); foreach (split("\n", $err) as $msg) if ($msg) xwMessage("$msg", 'red'); }

  1. Add message to queue

function xwMessage($msg, $col = '#000080') { global $xwMessages; $msg = htmlentities($msg); return $xwMessages[] = "$msg"; }

  1. Print list of passed object's methods and properties

function xwCheckoutObject($obj) { if (is_object($obj)) { print strtoupper("
$obj properties:
"); foreach (get_object_vars($obj) as $k=>$v) print htmlentities("$k => $v")."
"; print strtoupper("
$obj methods:
"); foreach (get_class_methods($obj) as $k=>$v) print "$v
"; } if (is_array($obj)) { print strtoupper("
Array content:
"); foreach ($obj as $k=>$v) print "$k => $v
"; } die; }

?>

Retrieved from "https://organicdesign.nz/wiki/index.php?title=XmlWiki&oldid=589"