mirror of
https://git.proxmox.com/git/pve-docs
synced 2025-04-30 22:47:23 +00:00
3e5c017486
Some (new) mediawiki versions can convert '<' and '>' to '<' or '>', this resulted in a situation where our encoded tag did not get detected by the decode logic, and thus a visitor only saw something like the following: > <--- @PVEDOCHACK@ > PCEtLSBhc2NpaWRvYyBIRUFERVIgLS0+DQo8c2NyaXB0IHR5..... > @PVEDOCHACK@ --> The word "HACK" was a bit unfortunate too, as it could be interpreted as in the docs got hacked, by unknowing people. So change that too to 'PVEDOCS_BASE64' Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
76 lines
2.3 KiB
PHP
76 lines
2.3 KiB
PHP
<?php
|
|
|
|
# see http://www.mediawiki.org/wiki/Manual:Parser_functions
|
|
|
|
$wgExtensionCredits['parserhook'][] = array(
|
|
'name' => "PVE Documentation Pages",
|
|
'description' => "Display PVE Documentation Pages",
|
|
'author' => "Dietmar Maurer",
|
|
);
|
|
|
|
# Define a setup function
|
|
$wgHooks['ParserFirstCallInit'][] = 'efPvedocsParserFunction_Setup';
|
|
$wgHooks['ParserAfterTidy'][] = 'efPvedocsPostProcessFunction';
|
|
|
|
# Add a hook to initialise the magic word
|
|
$wgHooks['LanguageGetMagic'][] = 'efPvedocsParserFunction_Magic';
|
|
|
|
function efPvedocsParserFunction_Setup(&$parser) {
|
|
# Set a function hook associating the "pvedocs" magic
|
|
# word with our function
|
|
$parser->setFunctionHook('pvedocs', 'efPvedocsParserFunction_Render');
|
|
|
|
$parser->setHook('pvehide', 'renderTagPveHideContent');
|
|
|
|
return true;
|
|
}
|
|
|
|
# similar code as in <htmlet> tag...
|
|
function efPvedocsPostProcessFunction($parser, &$text) {
|
|
$text = preg_replace_callback(
|
|
'/-_- @PVEDOCS_BASE64@ ([0-9a-zA-Z\\+\\/]+=*) @PVEDOCS_BASE64@ -_-/sm',
|
|
function ($m) { return base64_decode("$m[1]"); },
|
|
$text);
|
|
|
|
return true;
|
|
}
|
|
|
|
// Render <pvehide>
|
|
function renderTagPveHideContent($input, array $args, Parser $parser,
|
|
PPFrame $frame ) {
|
|
// simply return nothing
|
|
return '';
|
|
}
|
|
|
|
|
|
function efPvedocsParserFunction_Magic(&$magicWords, $langCode) {
|
|
# Add the magic word
|
|
# The first array element is whether to be case sensitive,
|
|
# in this case (0) it is not case sensitive, 1 would be sensitive
|
|
# All remaining elements are synonyms for our parser function
|
|
$magicWords['pvedocs'] = array( 0, 'pvedocs' );
|
|
|
|
# unless we return true, other parser functions extensions won't get loaded.
|
|
return true;
|
|
}
|
|
|
|
function efPvedocsParserFunction_Render($parser, $param1 = '', $param2 = '') {
|
|
|
|
$parser->disableCache();
|
|
|
|
# only allow simply names, so that users can only include
|
|
# files from within "/usr/share/pve-docs/"
|
|
if (!preg_match("/[a-z0-9.-]+\.html/i", $param1)) {
|
|
die("no such manual page");
|
|
}
|
|
|
|
$content = file_get_contents("/usr/share/pve-docs/$param1");
|
|
|
|
# do not use '<' or '>', it seems newer mediawiki converts it to '<' and '>'
|
|
# and then the regex for the decode in efPvedocsPostProcessFunction does not matches..
|
|
$output = '-_- @PVEDOCS_BASE64@ '.base64_encode($content).' @PVEDOCS_BASE64@ -_-';
|
|
return array($output, 'noparse' => true, 'isHTML' => true);
|
|
}
|
|
|
|
?>
|