=====Checklinks Action=====
>>==See also:==
	- Documentation: ChecklinksActionInfo.
==works with:==
	- Wikka 1.1.6.2 & 1.1.6.3 & 1.1.6.4
>>//NOT included in any Wikka version//{{lastedit show="3"}}
This is the development page for the checklinks action.

===Installation===
	- Save the code below as ##plugins/actions/checklinks/checklinks.php## 
	- Give it the same file permissions as the other php files in that directory.

=== Code ===
%%(php)<?php
#
# Checks hyperlinks appearing on wiki pages
# (intra-wiki, intra-wiki, actions and external links to web pages or shared files)
#
# @package      Actions
# @name         checklinks
#
# @authors      DomBonj
# @authors      Credit to steven@haryan.to, carinridge@hotmail.com (part of external link check code)
#
# @version      0.94
#
# @input        Parameters =  [scope=('page'|'user'|'all')] [sort=('tag'|'type'|'cnt'|)] [i][t][w]
#               default values: scope='user', sort='tag'
#               options: i=check internal links only ; t=terse report ; w=check that page tags are valid WikiWords
#
# @uses         Wakka::Format()
# @uses         Wakka::LoadAll()
# @uses         Wakka::LoadPage()
# @uses         Wakka::GetUserName()
# @uses         Wakka::GetPageTag()
# @uses         Wakka::Href()
# @uses         Wakka::SetPage()
# @uses         Wakka::GetInterWikiUrl()
#

// i18n strings
if (!defined('CL_SERVER_OK')) define('CL_SERVER_OK', 'Server OK');
if (!defined('CL_HOST_NOT_FOUND')) define('CL_HOST_NOT_FOUND', 'Server not found');
if (!defined('CL_HOST_UNREACH')) define('CL_HOST_UNREACH', 'Server unreachable');
if (!defined('CL_HOST_TIMEOUT')) define('CL_HOST_TIMEOUT', 'Server timeout');
if (!defined('CL_HOST_REJECT')) define('CL_HOST_REJECT', 'Server reject');
if (!defined('CL_FILE_NOHTTP')) define('CL_FILE_NOHTTP', 'No HTTP reply');
if (!defined('CL_NOSUCH_FILE')) define('CL_NOSUCH_FILE', 'File not found');
if (!defined('CL_MISSING_PAGE')) define('CL_MISSING_PAGE', 'Page non-existent');
if (!defined('CL_MISSING_INTERIWIKI')) define('CL_MISSING_INTERIWIKI', 'Wiki not defined');
if (!defined('CL_NON_WIKINAME')) define('CL_NON_WIKINAME', 'Page tag not a WikiName');
if (!defined('CL_COLNAME_SUMMARY')) define('CL_COLNAME_SUMMARY', 'Link type,Valid links #,Broken links #,Total');
if (!defined('CL_COLNAME_DETAILED')) define('CL_COLNAME_DETAILED', 'Link type,Page name,Link value,Error message,Occurences');
if (!defined('CL_NAME_AC')) define('CL_NAME_AC', 'Action');
if (!defined('CL_NAME_EX')) define('CL_NAME_EX', 'External');
if (!defined('CL_NAME_FI')) define('CL_NAME_FI', 'File');
if (!defined('CL_NAME_IW')) define('CL_NAME_IW', 'Inter-wiki');
if (!defined('CL_NAME_WN')) define('CL_NAME_WN', 'Wiki page');
if (!defined('CL_SUMMARY')) define('CL_SUMMARY', 'Link check summary');
if (!defined('CL_DETAILED')) define('CL_DETAILED', 'Detailed report');
// internal constants
if (!defined('CL_MAX_TRY')) define('CL_MAX_TRY', 3); // page fetch attempts
if (!defined('CL_MAX_REDIRECTS')) define('CL_MAX_REDIRECTS', 3); // allowed redirections
if (!defined('CL_CX_TIMEOUT ')) define('CL_CX_TIMEOUT', 8); // connection time allowance
if (!defined('CL_CACHE_LIFETIME')) define('CL_CACHE_LIFETIME', 120); // caching for 2 minutes
if (!defined('CL_MAX_LINK_LENGTH')) define('CL_MAX_LINK_LENGTH', 60); // max displayed length for an hyperlink

if (!function_exists('CLerror'))
{
	function CLerror ($msg)
	{
		return ('<em class="error">'. $msg .'</em><br />');
	}

	function CLsummary ($goodlinks, $badlinks)
	{
		$linknames = array('ac'=>CL_NAME_AC, 'ex'=>CL_NAME_EX, 'iw'=>CL_NAME_IW, 'wn'=>CL_NAME_WN, 'fi'=>CL_NAME_FI);
		$table_css = "class='data' cellpadding='2' cellspacing='1' border='2'";
		$colheads = split(',', CL_COLNAME_SUMMARY);
		$sumgood = 0; $sumbad = 0;
		$output = sprintf ("<table $table_css><tr class='comment'><th>%s</th><th>%s</th><th>%s</th></tr>", $colheads[0], $colheads[1], $colheads[2]);
		foreach ($goodlinks as $linktype => $cnt)
		{
			$output .= "<tr><td>{$linknames[$linktype]}</td><td align='right'>{$cnt}</td><td align='right'>".(empty($badlinks[$linktype])?0:$badlinks[$linktype])."</td></tr>";
			$sumgood += $cnt;
			$sumbad += (empty($badlinks[$linktype])? 0 : $badlinks[$linktype]);
		}
		$output .= "<tr class='comment'><td>{$colheads[3]}</td><td align='right'>$sumgood</td><td align='right'>$sumbad</td></tr></table>";
		return $output;
	}

	function CLreport ($thisone, $badlinks, $keyorder)
	{
		$linknames = array('ac'=>CL_NAME_AC, 'ex'=>CL_NAME_EX, 'iw'=>CL_NAME_IW, 'wn'=>CL_NAME_WN, 'fi'=>CL_NAME_FI);
		$table_css = "class='wikka' cellpadding='2' cellspacing='1' border='2'";
		$colheads = split(',', CL_COLNAME_DETAILED);
		$output = sprintf ("<table $table_css><tr class='comment'><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th></tr>", $colheads[0], $colheads[1], $colheads[2], $colheads[3], $colheads[4]);
		foreach ($keyorder as $tag => $val)
		{
			$cnt = $badlinks[$tag];
			preg_match('/^(.+) (\w\w)\/(.+?)\*(.+)$/', $tag, $matches);
			$pagelink = "<a href='". $thisone->Href('', $matches[1]) ."'>{$matches[1]}</a>";
			$link = ($matches[2] == 'ex') ? ("<a href='{$matches[3]}'>".substr($matches[3], 0, CL_MAX_LINK_LENGTH).((strlen($matches[3])>CL_MAX_LINK_LENGTH)?'...':'').'</a>')
				: (($matches[2] == 'fi') ? ("<a href='file://{$matches[3]}'>".substr($matches[3], 0, CL_MAX_LINK_LENGTH).((strlen($matches[3])>CL_MAX_LINK_LENGTH)?'...':'').'</a>') : $matches[3]);
			$output .= "<tr><td>{$linknames[$matches[2]]}</td><td>$pagelink</td><td>$link</td><td>{$matches[4]}</td><td align='right'>$cnt</td></tr>";
		}
		$output .= '</table>';
		return $output;
	}

	function CLcheck_page($fp, $page, $hostname, $firstcall)
	{
		$filestatus = CL_FILE_NOHTTP;
		$tmp = fputs ($fp, sprintf( "HEAD %s HTTP/1.0\r\nHost: %s\r\nUser-Agent: WikkaCheckLinks/1.0\r\n\r\n", $page, $hostname));
		for ($try = 1; ($try <= CL_MAX_TRY) && ($filestatus == CL_FILE_NOHTTP); $try++)
		{
			if (($http_reply = fgets($fp, 256)) == NULL)
			{
				break;
			}
			if (preg_match('/^HTTP\/(\d)\.(\d)\s+(\d+)\s+(.*)$/', $http_reply, $matches ))
			{
				$filestatus = ($matches[4]) ? trim($matches[4]) : trim($matches[3]);
				if ($firstcall && ($matches[3] == '100'))
				{ // in HTTP/1.1, '100' means Continue
					$filestatus = CLcheck_page($fp, $page, $hostname, false);
				}
				else if (substr($matches[3], 0, 1) == '3')
				{ // redirection: let's find the new location
					while (!feof($fp))
					{
						$reply .= fgets($fp, 256);
					}
					if (preg_match('/^Location:\s+(\S+)\s*$/m', $reply, $matches1))
					{
						$filestatus = 'MOV '. $matches1[1];
					}
				}
			}
		}
		return $filestatus;
	}

	function CLcheck_link($url, $p='')
	{
		static $statuses = array();
		static $hostnames = array();
		$now = time();
		$purl = parse_url($url);
		$proto = isset($purl['scheme']) ? $purl['scheme'] : 'http';
		$port = isset($purl['port']) ? $purl['port'] : '';
		$path = isset($purl['path']) ? $purl['path'] : '/';
		$suffix = isset($purl['query']) ? $purl['query'] : '';
		if ((empty($purl['host'])) || ($proto!='http')&&($proto!='https')&&($proto!='ftp'))
		{
			$serverstatus = CL_HOST_NOT_FOUND;
		}
		else
		{
			$hostname = strtolower($purl['host']);
			if (preg_match('/^d+.d+.d+.d+/', $hostname))
			{ // the host is an IP address
				$ip = $hostname;
			}
			else
			{ // the host is a domain name, so we have to resolve it first
				$from_cache = false;
				if (isset($hostnames[$hostname]))
				{ // have we tried to resolve it not so long ago?
					if (($now - $hostnames[$hostname][1]) <= CL_CACHE_LIFETIME)
					{
						$ip = $hostnames[$hostname][0];
						$from_cache = true;
					} 
				} 
				if (!$from_cache)
				{
					$ip = gethostbyname($hostname);
					// if hostname not resolvable, gethostbyname returns its argument unchanged
					if ($ip === $hostname)
					{
						$ip = '';
					}	
					// cache this resolve 
					else
					{
						$hostnames[$hostname] = array($ip, $now);
					}
				}
			}
			if (!$ip)
			{ // was the hostname unresolvable?
				$serverstatus = CL_HOST_NOT_FOUND;
			}
			else
			{
				if (!$port)
				{
					if ($proto == 'http')
					{
						$port = 80;
					} 
					elseif ($proto == 'https')
					{
						$port = 443;
					} 
					elseif ($proto == 'ftp')
					{
						$port = 21;
					
					} 
				} 
				$key = "$ip:$port";
				// have we checked the server not so long ago?
				$from_cache = false;
				if (isset($statuses[$key]))
				{ 
					if (($now - $statuses[$key][2]) <= CL_CACHE_LIFETIME)
					{
						$serverstatus = $statuses[$key][0];
						if ($serverstatus == CL_SERVER_OK)
						{
							$from_cache = true;
						}
					} 
				} 
				if (!$from_cache || ($from_cache && ($serverstatus == CL_SERVER_OK)))
				{ // we have to check the server, or the host is ok so check the file
					$errno = 0;
					$errstr = '';
					if ($fp = fsockopen($ip, $port, $errno, $errstr, CL_CX_TIMEOUT))
					{
						$serverstatus = CL_SERVER_OK;
						$filestatus = CL_FILE_NOHTTP;
						$page = ($suffix) ? $path .'?'. $suffix : $path;
						$filestatus = CLcheck_page($fp, $page, $hostname, false);
					}
					else
					{ // could not connect to server
						if (preg_match('/timed?[- ]?out/i', $errstr))
						{ 
							$serverstatus = CL_HOST_TIMEOUT;
						}
						elseif (preg_match('/refused/i', $errstr))
						{ 
							$serverstatus = CL_HOST_REJECT;
						}
						else
						{ 
							$serverstatus = CL_HOST_UNREACH;
						} 
					} 
					// cache this (server, file) pair
					$statuses[$key] = array($serverstatus, $filestatus, $now);
				}
			} 
		} 
		if ($filestatus == '200')
		{
			$filestatus = 'OK';
		}
		else if ($filestatus == '302')
		{
			$filestatus = 'OK';
		}
		$output = ($serverstatus != CL_SERVER_OK) ? $serverstatus : $filestatus;

		return ($output);
	}
}

$output = '';
if ( (isset($vars['opts']) && (!preg_match("/^[tiw]{1,3}$/i", $vars['opts'])))
	|| (isset($vars['scope']) && (!preg_match("/^(page|user|all)$/i", $vars['scope'])))
	|| (isset($vars['sort']) && (!preg_match("/^(tag|type|cnt)$/i", $vars['sort']))) )
{
	$output .= CLerror("Usage: checklinks [scope=\"user|page|all\"] [sort=\"tag|type|cnt\"] [opts=\"[i][t][w]\"]");
}
else
{
	$save_page = $this->page;
	$save_tag = $this->tag;
	$goodcnts = array('ac'=>0, 'ex'=>0, 'iw'=>0, 'wn'=>0, 'fi'=>0);
	$show_external = (!isset($vars['opts']) || !preg_match('/i/', $vars['opts']));
	$show_badwnames = (isset($vars['opts']) && preg_match('/w/', $vars['opts']));
	$base_url = preg_quote($this->config['base_url'], '/');
	$badlinks = array();

	// first pass to load all page tags
	$query = "SELECT tag, time FROM ".$this->config['table_prefix']."pages WHERE ((latest = 'Y'))";
	$rows = $this->LoadAll($query);
	foreach ($rows as $row)
	{
		$exist[$row['tag']] = $row['time'];
	}
	// start checking requested pages
	$query = "SELECT * FROM ".$this->config['table_prefix']."pages WHERE ((latest = 'Y') AND (user <> 'WikkaInstaller')";
	if ('all' == strtolower($vars['scope']))
	{
		$query .= '';
	}
	else if ('page' == strtolower($vars['scope']))
	{
		$query .= " AND (tag = '". $this->GetPageTag() ."')";
	}
	else 
	{ // default value
		$query .= " AND (owner = '". $this->GetUserName() ."')";
	}
	$query .= ')';
	$rows = $this->LoadAll($query);
	foreach ($rows as $row)
	{
		if ($this->HasAccess('read', $row['tag']))
		{ // this page is to be scanned: pretend your are this page
			$this->SetPage($this->LoadPage($row['tag']));
			$tmppage = $this->page['body'];
			// get rid of raw HTML and code blocks
			$tmppage = preg_replace("/\"\"(.*?)\"\"/s", '', $tmppage);
			$tmppage = preg_replace("/\%\%(.*?)\%\%/s", '', $tmppage);

			// 1. is page tag formatted as a valid WikiName ?
			if (!preg_match("/^([A-ZÄÖÜ]+[a-zßäöü]+[A-Z0-9ÄÖÜ][A-Za-z0-9ÄÖÜßäöü]*)$/", $row['tag'], $matches) && $show_badwnames)
			{
				$badlinks[$row['tag']." wn/&nbsp;*".CL_NON_WIKINAME] = (empty($badlinks[$row['tag']." wn/&nbsp;*".CL_NON_WIKINAME])) ? 1 : $badlinks[$row['tag']." wn/&nbsp;*".CL_NON_WIKINAME]+1;
				$badcnts['wn'] += 1;
			}
			// 2. check actions
			preg_match_all("/\{\{(.*?)\}\}/", $tmppage, $matches);
			foreach ($matches[1] as $actionname)
			{
				if (preg_match("/^([A-Za-z0-9]+)/", trim($actionname), $matches1))
				{
					if (!file_exists($this->config['action_path']."/".$matches1[1].".php"))
					{
						$badlinks[$row['tag']." ac/{$matches1[1]}*". CL_NOSUCH_FILE] = (empty($badlinks[$row['tag']." ac/{$matches1[1]}*". CL_NOSUCH_FILE])) ? 1 : $badlinks[$row['tag']." ac/{$matches1[1]}*". CL_NOSUCH_FILE]+1;
						$badcnts['ac'] += 1;
					}
					else
					{
						$goodcnts['ac'] +=1;
					}
				}
			}
			// now get rid of actions to avoid confusion
			$tmppage = preg_replace("/\{\{(.*?)\}\}/", '', $tmppage);
			// 3. check interwiki links
			preg_match_all("/([A-ZÄÖÜ][A-Za-zÄÖÜßäöü]+)[:](\S*)\b/", $tmppage, $matches);
			foreach ($matches[1] as $interwikiname)
			{
				if (!$this->GetInterWikiUrl(trim($interwikiname), ''))
				{
					$badlinks[$row['tag']." iw/{$interwikiname}*". CL_MISSING_INTERIWIKI] = (empty($badlinks[$row['tag']." iw/{$interwikiname}*". CL_MISSING_INTERIWIKI])) ? 1 : $badlinks[$row['tag']." iw/{$interwikiname}*". CL_MISSING_INTERIWIKI]+1;
					$badcnts['iw'] += 1;
				}
				else
				{
					$goodcnts['iw'] +=1;
				}
			}
			// now get rid of interwiki links to avoid confusion
			$tmppage = preg_replace("/([A-ZÄÖÜ][A-Za-zÄÖÜßäöü]+[:]\S*)\b/", '', $tmppage);

			// now check hyperlinks; first, prevent recursive calling
			$page = preg_replace('/\{\{\s*checklinks\b.*?\}\}/i', '', $this->page['body']);
			// do not count twice non-existent links
			$page = preg_replace('/\{\{\s*wantedpages\s*\}\}/i', '', $page);
			// render the page
			$html = $this->Format($page, 'wakka');
			if (preg_match_all("/href\=[\"|\']((http|https|ftp):\/\/[^\\s\"\'<>]+)/", $html, $matches))
			{
				foreach ($matches[1] as $url)
				{ // 4. check intra-wiki links
					if (preg_match('/'.$base_url.'([A-Za-zÄÖÜßäöü][A-Za-z0-9ÄÖÜßäöü]*)/', $url, $matches1))
					{
						$wikiname = $matches1[1];
						if (!$exist[trim($wikiname)])
						{
							$badlinks[$row['tag']." wn/{$wikiname}*". CL_MISSING_PAGE] = (empty($badlinks[$row['tag']." wn/{$wikiname}*". CL_MISSING_PAGE])) ? 1 : $badlinks[$row['tag']." wn/{$wikiname}*". CL_MISSING_PAGE]+1;
							$badcnts['wn'] += 1;
						}
						else
						{
							$goodcnts['wn'] += 1;
						}
					}
					else if ($show_external)
					{ // 5. check external hyperlinks
						$OK = (strtoupper($tmp = CLcheck_link($url)) == 'OK');
						// allow at most 3 successive redirections
						for ($i=1; !$OK && preg_match("/^MOV (.+)$/", $tmp, $matches1) && ($i<=CL_MAX_REDIRECTS); $i++)
						{
							$OK = (strtoupper($tmp = CLcheck_link(trim($matches1[1]))) == "OK");
						}
						if (!$OK)
						{
							$badlinks[$row['tag']." ex/{$url}*$tmp"] = (empty($badlinks[$row['tag']." ex/{$url}*$tmp"])) ? 1 : $badlinks[$row['tag']." ex/{$url}*$tmp"]+1;
							$badcnts['ex'] += 1;
						}
						else
						{
							$goodcnts['ex'] += 1;
						}
					}
				} // foreach $matches
			} // if preg_match_all http
			// 5. check href-ed files
			if (preg_match_all("/href\=[\"|\']file:\/{2,}([^\\\"\'<>]+)[\'\"]/", $html, $matches))
			{
				foreach ($matches[1] as $rawfname)
				{
					if (!file_exists(rawurldecode($rawfname)))
					{
						$fname = str_replace(' ', '&nbsp;', $rawfname);
						$badlinks[$row['tag']." fi/$fname*".CL_NOSUCH_FILE] = (empty($badlinks[$row['tag']." fi/$fname*".CL_NOSUCH_FILE])) ? 1 : $badlinks[$row['tag']." fi/$fname*".CL_NOSUCH_FILE]+1;
						$badcnts['fi'] += 1;
					}
					else
					{
						$goodcnts['fi'] += 1;
					}
				} // foreach $matches
			} // if preg_match_all file
		} // if $this->HasAcess
	} // foreach $rows

	// restore original values
	$this->tag = $save_tag;
	$this->page = $save_page;

	// now, sort the associative array's keys
	if (isset($vars['sort']))
	{
		if ($vars['sort'] == 'type')
		{
			$field = 2;
			$fn = 'asort';
		}
		else if ($vars['sort'] == 'cnt')
		{
			$field = 'cnt';
			$fn = 'arsort';
		}
	}
	// default case: sort on tag
	if (!$field)
	{
		$field = 1;
		$fn = 'asort';
	}
	$keyorder = array();
	foreach ($badlinks as $index => $cnt)
	{
		preg_match("/^(.+) (\w\w)\/(.+)\*(.+)$/", $index, $matches);
		$keyorder[$index] = ($field == 'cnt') ? $cnt : strtolower($matches[$field]);
	}
	// do the actual sorting
	$fn($keyorder);

	$output .= '<h3>'. CL_SUMMARY .'</h3>'. CLsummary($goodcnts, $badcnts).'<br />';
	if (!isset($vars['opts']) || !preg_match('/t/', $vars['opts'], $tmp))
	{
		$output .= '<h3>'. CL_DETAILED .'</h3>'. CLreport($this, $badlinks, $keyorder);
	}
}
echo $output;
// avoid side-effect if there were footnotes on checked pages
if (function_exists('FNprint'))
{
	FNprint($this, 'purge', '', $this->Href());
}
?>
%%

----
CategoryUserContributions