Document sanitizer constructor

lib/AbstractSanitizer.php

@@ -0,0 +1,54 @@
+<?php
+/** @license MIT
+ * Copyright 2023 J. King
+ * See LICENSE and AUTHORS files for details */
+
+declare(strict_types=1);
+
+namespace MensBeam\HTML;
+
+/** An implementation of the W3C HTML Sanitizer API.
+ * 
+ * The class implements the following methods:
+ * 
+ * - constructor
+ * - sanitize
+ * - sanitizeFor
+ * - getConfiguration
+ * - getDefaultConfiguration
+ */
+abstract class AbstractSanitizer {
+    /** @var array $config The parsed configuration, as used for processing */
+    protected $config;
+
+    /** Initializes a sanitizer with the provided configuration, or the default configuration
+     * 
+     * The configuration array may contain any of the following keys:
+     * 
+     * - `allowElements`: an indexed array of elements to allow. Elements not in this list will be blocked as if they were included in the `blockElements` list
+     * - `blockElements`: an indexed array of elements to remove from the tree while retaining their children
+     * - `dropElements`: an indexed array of elements to remove from the tree along with their children
+     * - `allowAttributes`: an indexed array of attributes to allow on certain elements. Attributes not in this list will be dropped
+     * - `dropAttributes`: an indexed array of attributes to remove from certain elements
+     * - `allowCustomElements`: Whether to allow custom elements, false by default. For the purposes of this implementation these are elements with names containing dashed. If true, elements are still subject to the allow, block, and drop lists
+     * - `allowUnknownMarkup`: Whether to allow non-standard elements which are not custom elements, false by default. If true, elements are still subject to the allow, block, and drop lists
+     * - `allowComments`: Whether to retain comments, false by default
+     * - `allowProcessingInstructions`: Whether to retain processing instructions, false by default. Processing instructions do not normally appear in HTML documents. This option is an extension to the specification
+     * - `nullNamespaceAsHtml`: Whether to interpret elements from the tree in the null namespace as being in the HTML namespace, true by default. Per standard behaviour HTML elements have a namespace URI, but not all parsers do this. This may be set to false when sanitizing XML documents. This option is an extension to the specification
+     * 
+     * The entries in element lists may be strings, in which case these are interpreted as local names in the HTML namespace, or an array with the following keys:
+     * 
+     * - `name`: The localName of the element
+     * - `namespace`: The namespaceURI of the element, a string or null. If omitted the HTML namespace is assumed
+     * 
+     * The entries in attribute lists are arrays with the follwoing keys
+     * 
+     * - `name`: The localName of the attribute
+     * - `namespace`: The namespaceURI of the attribute. If omitted the null namespace is assumed
+     * - `elements`: An indexed array of elements on which to allow the attribute, in the same format as other element lists. The string `"*"` may also be supplied instead of an array to mean all elements
+     * 
+     * @param array $config A configuration to use instead of the default one
+     */
+    public function __construct(array $config = null) {
+    }
+}