Update wizard for file links in RTE fields. Migrates links like <link fileadmin/download.pdf download >Download</link> to <link file:123 download >Download</link>

RteFileLinksUpdateWizard.php

<?php
namespace TYPO3\CMS\Install\Updates;

/***************************************************************
 *  Copyright notice
 *
 *  (c) 2013 Nico de Haen
 *  All rights reserved
 *
 *  This script is part of the TYPO3 project. The TYPO3 project is
 *  free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 3 of the License, or
 *  (at your option) any later version.
 *
 *  The GNU General Public License can be found at
 *  http://www.gnu.org/copyleft/gpl.html.
 *
 *  This script is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  This copyright notice MUST APPEAR in all copies of the script!
 ***************************************************************/
use TYPO3\CMS\Core\Utility\GeneralUtility;

/**
 * Upgrade wizard which goes through tt_content records
 * and replaces links with a path reference with the according sys_file reference
 *
 * @author Nico de Haen
 * @license http://www.gnu.org/copyleft/gpl.html
 */
class RteFileLinksUpdateWizard extends \TYPO3\CMS\Install\Updates\AbstractUpdate {

  /**
	 * @var string
	 */
	protected $title = 'Migrate file relations of tt_content "bodytext"';

	/*
	 * @var \TYPO3\CMS\Core\Html\RteHtmlParser
	 */
	protected $rteHtmlParser;

	/**
	 * should we include deleted content elements here?
	 *
	 * @var string
	 */
	protected $addWhere = 'bodytext RLIKE "<link[[:space:]]*([[:alnum:]]|_|-)*/" AND deleted = 0';

	/**
	 * @var int
	 */
	protected $convertedLinkCounter = 0;

	/**
	 *
	 * @var array
	 */
	protected $fileNotFoundErrors = array();

	/**
	 * Reference to the current record
	 * to enable meaningful error messages in recursion method
	 * @var null|array
	 */
	protected $currentRecord = NULL;

	/**
	 * The default storage
	 * @var \TYPO3\CMS\Core\Resource\ResourceStorage
	 */
	protected $storage = NULL;

	protected $fileAdminDir;

	/**
	 * Initialize all required repository and factory objects.
	 *
	 * @throws \RuntimeException
	 */
	protected function init() {
		$this->rteHtmlParser = GeneralUtility::makeInstance('TYPO3\\CMS\\Core\\Html\\RteHtmlParser');
		$storageRepository = \TYPO3\CMS\Core\Utility\GeneralUtility::makeInstance('TYPO3\\CMS\\Core\\Resource\\StorageRepository');
		$storages = $storageRepository->findAll();
		$this->storage = $storages[0];
		$this->fileAdminDir = $GLOBALS['TYPO3_CONF_VARS']['BE']['fileadminDir'];
	}

	/**
	 * Checks if an update is needed
	 *
	 * @param 	string		&$description: The description for the update
	 * @return 	boolean		TRUE if an update is needed, FALSE otherwise
	 */
	public function checkForUpdate(&$description) {
		$updateNeeded = FALSE;
		$notMigratedRowsCount = $GLOBALS['TYPO3_DB']->exec_SELECTcountRows(
			'uid',
			'tt_content',
			$this->addWhere
		);
		// Fetch records where the field contains file links
		if ($notMigratedRowsCount > 0) {
			$description = 'There are file links in RTE fields referencing files without using the file abstraction layer FAL.';
			$updateNeeded = TRUE;
		}
		return $updateNeeded;
	}

	/**
	 * Performs the update.
	 *
	 * @param 	array		&$dbQueries: queries done in this update
	 * @param 	mixed		&$customMessages: custom messages
	 * @return 	boolean		TRUE on success, FALSE on error
	 */
	public function performUpdate(array &$dbQueries, &$customMessages) {
		$this->init();
		if(!$this->storage) {
			$customMessages = 'No file resource storage found';
			return FALSE;
		}
		$records = $this->getRecordsFromTable('tt_content');
		foreach($records as $singleRecord) {
			$this->migrateRecord($singleRecord);
		}
		if(count($this->fileNotFoundErrors) > 0) {
			$customMessages .= '<strong>' . count($this->fileNotFoundErrors). ' missing files:</strong><br />';
			$customMessages .= implode('<br />', $this->fileNotFoundErrors);
			if($this->convertedLinkCounter == 0) {
					// no links converted only missing files: UPDATE was not successful
				return FALSE;
			}
		}
		if($this->convertedLinkCounter > 0) {
			$customMessages = $this->convertedLinkCounter . ' links converted.<br />' . $customMessages;
		} else {
			$customMessages .= 'No file links found';
		}
		return TRUE;
	}


	/**
	 *  Processes each record and updates the database
	 *
	 * @param array $record
	 * @return void
	 */
	protected function migrateRecord(array $record) {
		$this->currentRecord = $record;
		$convertedBodyText = $this->convertFileLink($record['bodytext']);
		$GLOBALS['TYPO3_DB']->exec_UPDATEquery(
			'tt_content',
			'uid = ' . $record['uid'],
			array('bodytext' => $convertedBodyText)
		);
	}

	/**
	 * The actual transformation of the links
	 * pretty similar to TS_links_rte in RteHtmlParser
	 *
	 * @param $value
	 * @return string
	 */
	protected function convertFileLink($value) {
		$blockSplit = $this->rteHtmlParser->splitIntoBlock('link', $value, 1);
		foreach ($blockSplit as $k => $v) {
			// Block
			if ($k % 2) {
				$tagCode = \TYPO3\CMS\Core\Utility\GeneralUtility::unQuoteFilenames(trim(substr($this->rteHtmlParser->getFirstTag($v), 0, -1)), TRUE);
				$link_param = $tagCode[1];
				if(intval(strpos($link_param, '/'))  && strpos($link_param, 'http') === FALSE){
					$fileObject = NULL;
					if(@file_exists(PATH_site . '/' . $link_param)) {
						try {
							if(strpos($link_param, $this->fileAdminDir) !== FALSE) {
								$fileObject = $this->storage->getFile('/' . str_replace($this->fileAdminDir,'',$link_param));
							}
						} catch (\TYPO3\CMS\Core\Resource\Exception\FileDoesNotExistException $notFoundException) {
							$this->fileNotFoundErrors[] = $link_param . ' not indexed (referenced in content element ' . $this->currentRecord['uid'] . ' on page ' . $this->currentRecord['pid'] . ')';
						}
						if ($fileObject instanceof \TYPO3\CMS\Core\Resource\FileInterface) {
							// It's a file
							$fileLink = 'file:' . $fileObject->getUid();
							$tagCode[1] = $fileLink;
							$bTag = implode(' ', $tagCode) . '>';
							$eTag = '</link>';
							$blockSplit[$k] = $bTag . $this->convertFileLink($this->rteHtmlParser->removeFirstAndLastTag($blockSplit[$k])) . $eTag;
							$this->convertedLinkCounter++;
						}
					} else {
							// Nothing to be done if file not found
						$this->fileNotFoundErrors[] = $link_param . ' not found (referenced in content element ' . $this->currentRecord['uid'] . ' on page ' . $this->currentRecord['pid'] . ')';
					}
				}
			}
		}

		$convertedText = implode('', $blockSplit);
		return $convertedText;
	}



	/**
	 * Retrieve every record which needs to be processe
	 *
	 * @return array
	 */
	protected function getRecordsFromTable() {
		$records = $GLOBALS['TYPO3_DB']->exec_SELECTgetRows('uid,pid,bodytext', 'tt_content', $this->addWhere);
		return $records;
	}

}


?>