TYPO3 v15 dev-main snapshot ()

Classes/Resource/TextExtraction/PlainTextExtractor.php

@@ -0,0 +1,64 @@
+<?php
+
+/*
+ * This file is part of the TYPO3 CMS project.
+ *
+ * It is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License, either version 2
+ * of the License, or any later version.
+ *
+ * For the full copyright and license information, please read the
+ * LICENSE.txt file that was distributed with this source code.
+ *
+ * The TYPO3 project - inspiring people to share!
+ */
+
+namespace TYPO3\CMS\Core\Resource\TextExtraction;
+
+use TYPO3\CMS\Core\Resource\FileInterface;
+use TYPO3\CMS\Core\Utility\PathUtility;
+
+/**
+ * A simple text extractor to extract text from plain text files.
+ */
+class PlainTextExtractor implements TextExtractorInterface
+{
+    /**
+     * Checks if the given file can be read by this extractor
+     *
+     * @return bool
+     */
+    public function canExtractText(FileInterface $file)
+    {
+        $canExtract = false;
+
+        if ($file->getMimeType() === 'text/plain') {
+            $canExtract = true;
+        }
+
+        return $canExtract;
+    }
+
+    /**
+     * The actual text extraction.
+     *
+     * @return string
+     */
+    public function extractText(FileInterface $file)
+    {
+        $localTempFile = $file->getForLocalProcessing(false);
+
+        // extract text
+        $content = (string)file_get_contents($localTempFile);
+
+        // In case of remote storage, the temporary copy of the
+        // original file in typo3temp must be removed
+        // Simply compare the filenames, because the filename is so unique that
+        // it is nearly impossible to have a file with this name in a storage
+        if (PathUtility::basename($localTempFile) !== $file->getName()) {
+            unlink($localTempFile);
+        }
+
+        return $content;
+    }
+}

Classes/Resource/TextExtraction/TextExtractorInterface.php

@@ -0,0 +1,41 @@
+<?php
+
+/*
+ * This file is part of the TYPO3 CMS project.
+ *
+ * It is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License, either version 2
+ * of the License, or any later version.
+ *
+ * For the full copyright and license information, please read the
+ * LICENSE.txt file that was distributed with this source code.
+ *
+ * The TYPO3 project - inspiring people to share!
+ */
+
+namespace TYPO3\CMS\Core\Resource\TextExtraction;
+
+use TYPO3\CMS\Core\Resource\FileInterface;
+
+/**
+ * An interface for text extractors
+ */
+interface TextExtractorInterface
+{
+    /**
+     * Checks if the given file can be read by this extractor
+     *
+     * @return bool
+     */
+    public function canExtractText(FileInterface $file);
+
+    /**
+     * The actual text extraction.
+     *
+     * Should return a string of the file's content
+     *
+     * @param FileInterface $file
+     * @return string
+     */
+    public function extractText(FileInterface $file);
+}

Classes/Resource/TextExtraction/TextExtractorRegistry.php

@@ -0,0 +1,101 @@
+<?php
+
+/*
+ * This file is part of the TYPO3 CMS project.
+ *
+ * It is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License, either version 2
+ * of the License, or any later version.
+ *
+ * For the full copyright and license information, please read the
+ * LICENSE.txt file that was distributed with this source code.
+ *
+ * The TYPO3 project - inspiring people to share!
+ */
+
+namespace TYPO3\CMS\Core\Resource\TextExtraction;
+
+use TYPO3\CMS\Core\Resource\FileInterface;
+use TYPO3\CMS\Core\SingletonInterface;
+use TYPO3\CMS\Core\Utility\GeneralUtility;
+
+class TextExtractorRegistry implements SingletonInterface
+{
+    /**
+     * Registered text extractor class names
+     *
+     * @var array
+     */
+    protected $textExtractorClasses = [];
+
+    /**
+     * Instance cache for text extractor classes
+     *
+     * @var TextExtractorInterface[]
+     */
+    protected $instances = [];
+
+    /**
+     * Allows to register a text extractor class
+     *
+     * @param string $className
+     * @throws \InvalidArgumentException
+     */
+    public function registerTextExtractor($className)
+    {
+        if (!class_exists($className)) {
+            throw new \InvalidArgumentException('The class "' . $className . '" you are trying to register is not available', 1422906893);
+        }
+
+        if (!in_array(TextExtractorInterface::class, class_implements($className) ?: [], true)) {
+            throw new \InvalidArgumentException($className . ' must implement interface' . TextExtractorInterface::class, 1422771427);
+        }
+
+        $this->textExtractorClasses[] = $className;
+    }
+
+    /**
+     * Get all registered text extractor instances
+     *
+     * @return TextExtractorInterface[]
+     */
+    public function getTextExtractorInstances()
+    {
+        if (empty($this->instances) && !empty($this->textExtractorClasses)) {
+            foreach ($this->textExtractorClasses as $className) {
+                $object = $this->createTextExtractorInstance($className);
+                $this->instances[] = $object;
+            }
+        }
+
+        return $this->instances;
+    }
+
+    /**
+     * Create an instance of a certain text extractor class
+     *
+     * @param string $className
+     * @return TextExtractorInterface
+     */
+    protected function createTextExtractorInstance($className)
+    {
+        return GeneralUtility::makeInstance($className);
+    }
+
+    /**
+     * Checks whether any registered text extractor can deal with a given file
+     * and returns it.
+     *
+     * @return TextExtractorInterface|null
+     */
+    public function getTextExtractor(FileInterface $file)
+    {
+        foreach ($this->getTextExtractorInstances() as $textExtractor) {
+            if ($textExtractor->canExtractText($file)) {
+                return $textExtractor;
+            }
+        }
+
+        return null;
+    }
+}