From e83fd277178cd1cab4a50bdc7a33c16d8ee877b5 Mon Sep 17 00:00:00 2001
From: Victor Deniz Falcon <victor@moodle.com>
Date: Fri, 27 Sep 2019 11:01:12 +0100
Subject: [PATCH] MDL-66397 filter_h5p: converts H5P URLs to embed code

---
 filter/h5p/classes/privacy/provider.php |  46 ++++++++++++
 filter/h5p/filter.php                   | 120 ++++++++++++++++++++++++++++++++
 filter/h5p/lang/en/filter_h5p.php       |  34 +++++++++
 filter/h5p/settings.php                 |  31 +++++++++
 filter/h5p/tests/filter_test.php        |  87 +++++++++++++++++++++++
 filter/h5p/version.php                  |  29 ++++++++
 lib/classes/plugin_manager.php          |   2 +-
 7 files changed, 348 insertions(+), 1 deletion(-)
 create mode 100644 filter/h5p/classes/privacy/provider.php
 create mode 100644 filter/h5p/filter.php
 create mode 100644 filter/h5p/lang/en/filter_h5p.php
 create mode 100644 filter/h5p/settings.php
 create mode 100644 filter/h5p/tests/filter_test.php
 create mode 100644 filter/h5p/version.php

diff --git a/filter/h5p/classes/privacy/provider.php b/filter/h5p/classes/privacy/provider.php
new file mode 100644
index 00000000000..215faccf3aa
--- /dev/null
+++ b/filter/h5p/classes/privacy/provider.php
@@ -0,0 +1,46 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle 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.
+//
+// Moodle 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.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * Privacy Subsystem implementation for filter_h5p.
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+namespace filter_h5p\privacy;
+
+defined('MOODLE_INTERNAL') || die;
+
+/**
+ * Privacy Subsystem for filter_h5p implementing null_provider.
+ *
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+class provider implements \core_privacy\local\metadata\null_provider {
+
+    /**
+     * Get the language string identifier with the component's language
+     * file to explain why this plugin stores no data.
+     *
+     * @return  string
+     */
+    public static function get_reason(): string {
+        return 'privacy:metadata';
+    }
+}
diff --git a/filter/h5p/filter.php b/filter/h5p/filter.php
new file mode 100644
index 00000000000..fcb35ec4a88
--- /dev/null
+++ b/filter/h5p/filter.php
@@ -0,0 +1,120 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle 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.
+//
+// Moodle 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.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+/**
+ * H5P filter
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+defined('MOODLE_INTERNAL') || die;
+
+/**
+ * H5P filter
+ *
+ * This filter will replace any occurrence of H5P URLs with the corresponding H5P content embed code
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+class filter_h5p extends moodle_text_filter {
+
+    /**
+     * @var boolean $loadresizerjs This is whether to request the resize.js script.
+     */
+    private static $loadresizerjs = true;
+
+    /**
+     * Function filter replaces any h5p-sources.
+     *
+     * @param  string $text    HTML content to process
+     * @param  array  $options options passed to the filters
+     * @return string
+     */
+    public function filter($text, array $options = array()) {
+
+        if (!is_string($text) or empty($text)) {
+            // Non string data can not be filtered anyway.
+            return $text;
+        }
+
+        if (stripos($text, 'http') === false) {
+            return $text;
+        }
+
+        $allowedsources = get_config('filter_h5p', 'allowedsources');
+        $allowedsources = array_filter(array_map('trim', explode("\n", $allowedsources)));
+        if (empty($allowedsources)) {
+            return $text;
+        }
+
+        $params = array(
+            'tagbegin' => "<iframe src=",
+            'tagend' => "</iframe>"
+        );
+
+        foreach ($allowedsources as $source) {
+            // It is needed to add "/embed" at the end of URLs like https:://*.h5p.com/content/12345 (H5P.com).
+            $params['urlmodifier'] = '';
+            if (!(stripos($source, 'embed'))) {
+                $params['urlmodifier'] = '/embed';
+            }
+
+            // Convert special chars.
+            $specialchars = ['*', '?', '&'];
+            $escapedspecialchars = ['[^.]+', '\?', '&amp;'];
+            $sourceid = str_replace('[id]', '[0-9]+', $source);
+            $escapechars = str_replace($specialchars, $escapedspecialchars, $sourceid);
+            $ultimatepattern = '#(' . $escapechars . ')#';
+
+            $h5pcontenturl = new filterobject($source, null, null, false,
+                   false, null, [$this, 'filterobject_prepare_replacement_callback'], $params);
+
+            $h5pcontenturl->workregexp = $ultimatepattern;
+            $h5pcontents[] = $h5pcontenturl;
+        }
+
+        return filter_phrases($text, $h5pcontents, null, null, false, true);
+    }
+
+    /**
+     * Callback used by filterobject / filter_phrases.
+     *
+     * @param string $tagbegin HTML to insert before any match
+     * @param string $tagend HTML to insert after any match
+     * @param string $urlmodifier string to add to the match URL
+     * @return array [$hreftagbegin, $hreftagend, $replacementphrase] for filterobject.
+     */
+    public function filterobject_prepare_replacement_callback($tagbegin, $tagend, $urlmodifier) {
+
+        $sourceurl = "$1";
+        if ($urlmodifier !== "") {
+            $sourceurl .= $urlmodifier;
+        }
+
+        $h5piframesrc = "\"".$sourceurl."\" width=\"100%\" height=\"637\" allowfullscreen=\"allowfullscreen\" style=\"border: 0;\">";
+
+        // We want to request the resizing script only once.
+        if (self::$loadresizerjs) {
+            $tagend .= '<script src="https://h5p.org/sites/all/modules/h5p/library/js/h5p-resizer.js"></script>';
+            self::$loadresizerjs = false;
+        }
+
+        return [$tagbegin, $tagend, $h5piframesrc];
+    }
+}
diff --git a/filter/h5p/lang/en/filter_h5p.php b/filter/h5p/lang/en/filter_h5p.php
new file mode 100644
index 00000000000..38f446044dc
--- /dev/null
+++ b/filter/h5p/lang/en/filter_h5p.php
@@ -0,0 +1,34 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// Moodle is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * Strings for filter_h5p
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+defined('MOODLE_INTERNAL') || die;
+
+$string['allowedsourceslist'] = 'Allowed sources';
+$string['allowedsourceslistdesc'] = 'List of sources from which users can embed H5P content. If empty, the filter won\'t convert any external URL.
+
+<b>[id]</b> is a placeholder for the H5P content id in the external source.
+
+<b>*</b> wildcard is supported. For example, *.example.com will embed H5P content from any subdomain of example.com, but not from the example.com domain.';
+$string['filtername'] = 'H5P';
+$string['privacy:metadata'] = 'This H5P filter does not store any personal data.';
diff --git a/filter/h5p/settings.php b/filter/h5p/settings.php
new file mode 100644
index 00000000000..6102eebbe38
--- /dev/null
+++ b/filter/h5p/settings.php
@@ -0,0 +1,31 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// Moodle is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * H5P filter settings
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+defined('MOODLE_INTERNAL') || die;
+
+if ($ADMIN->fulltree) {
+    $settings->add(new admin_setting_configtextarea('filter_h5p/allowedsources', get_string('allowedsourceslist', 'filter_h5p'),
+            get_string('allowedsourceslistdesc', 'filter_h5p'),
+            "https://h5p.org/h5p/embed/[id]\nhttps://*.h5p.com/content/[id]/embed\nhttps://*.h5p.com/content/[id]"));
+}
diff --git a/filter/h5p/tests/filter_test.php b/filter/h5p/tests/filter_test.php
new file mode 100644
index 00000000000..2e17c105ab0
--- /dev/null
+++ b/filter/h5p/tests/filter_test.php
@@ -0,0 +1,87 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle 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.
+//
+// Moodle 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.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * Unit tests for the filter_h5p
+ *
+ * @package    filter_h5p
+ * @category   test
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+defined('MOODLE_INTERNAL') || die();
+
+global $CFG;
+require_once($CFG->dirroot.'/filter/h5p/filter.php');
+
+/**
+ * Unit tests for the H5P filter.
+ *
+ * @copyright 2019 Victor Deniz <victor@moodle.com>
+ * @license   http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+class filter_h5p_testcase extends advanced_testcase {
+
+    public function setUp() {
+        parent::setUp();
+
+        $this->resetAfterTest(true);
+
+        set_config('allowedsources', "https://h5p.org/h5p/embed/[id]\nhttps://*.h5p.com/content/[id]/embed\nhttps://*.h5p.com/content/[id]
+                \nhttps://generic.wordpress.soton.ac.uk/altc/wp-admin/admin-ajax.php?action=h5p_embed&id=[id]", 'filter_h5p');
+        // Enable h5p filter at top level.
+        filter_set_global_state('h5p', TEXTFILTER_ON);
+    }
+
+    /**
+     * Check that h5p tags with urls from allowed domains are filtered.
+     *
+     * @param string $text Original text
+     * @param string $filteredtextpattern Text pattern after H5P filter
+     *
+     * @dataProvider texts_provider
+     */
+    public function test_filter_urls($text, $filteredtextpattern) {
+
+        $filterplugin = new filter_h5p(null, array());
+
+        $filteredtext = $filterplugin->filter($text);
+        $this->assertRegExp($filteredtextpattern, $filteredtext);
+    }
+
+    /**
+     * Provides texts to filter for the {@link self::test_filter_urls} method.
+     *
+     * @return array
+     */
+    public function texts_provider() {
+        return [
+            ["http:://example.com", "#http:://example.com#"],
+            ["http://google.es/h5p/embed/3425234", "#http://google.es/h5p/embed/3425234#"],
+            ["https://h5p.org/h5p/embed/547225", "#<iframe src=\"https://h5p.org/h5p/embed/547225\"[^>]+?>#"],
+            ["https://moodle.h5p.com/content/1290729733828858779/embed", "#<iframe src=\"https://moodle.h5p.com/content/1290729733828858779/embed\"[^>]+?>#"],
+            ["https://moodle.h5p.com/content/1290729733828858779", "#<iframe src=\"https://moodle.h5p.com/content/1290729733828858779/embed\"[^>]+?>#"],
+            ["<a href=\"https://h5p.org/h5p/embed/547225\">link</a>",  "#^((?!iframe).)*$#"],
+            ["this is a text with an h5p url https://h5p.org/h5p/embed/547225 inside",
+                    "#this is a text with an h5p url <iframe src=\"https://h5p.org/h5p/embed/547225\"(.|\n)*> inside#"],
+            ["https://generic.wordpress.soton.ac.uk/altc/wp-admin/admin-ajax.php?action=h5p_embed&amp;id=13",
+                    "#<iframe src=\"https://generic.wordpress.soton.ac.uk/altc/wp-admin/admin-ajax.php\?action=h5p_embed\&amp\;id=13\"[^>]+?>#"],
+            ["https://h5p.org/h5p/embed/547225 another content in the same page https://moodle.h5p.com/content/1290729733828858779/embed",
+                    "#<iframe src=\"https://h5p.org/h5p/embed/547225\"[^>]+?>((?!<iframe).)*<iframe src=\"https://moodle.h5p.com/content/1290729733828858779/embed\"[^>]+?>#"]
+        ];
+    }
+}
\ No newline at end of file
diff --git a/filter/h5p/version.php b/filter/h5p/version.php
new file mode 100644
index 00000000000..343b5d2ffac
--- /dev/null
+++ b/filter/h5p/version.php
@@ -0,0 +1,29 @@
+<?php
+// This file is part of Moodle - http://moodle.org/
+//
+// Moodle 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.
+//
+// Moodle 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.
+//
+// You should have received a copy of the GNU General Public License
+// along with Moodle.  If not, see <http://www.gnu.org/licenses/>.
+
+/**
+ * Version of filter_h5p.
+ *
+ * @package    filter_h5p
+ * @copyright  2019 Victor Deniz <victor@moodle.com>
+ * @license    http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
+ */
+
+defined('MOODLE_INTERNAL') || die;
+
+$plugin->version  = 2019092000;
+$plugin->requires = 2019092000;
+$plugin->component = 'filter_h5p';
diff --git a/lib/classes/plugin_manager.php b/lib/classes/plugin_manager.php
index 1e65d4f1fe1..45ed1633242 100644
--- a/lib/classes/plugin_manager.php
+++ b/lib/classes/plugin_manager.php
@@ -1778,7 +1778,7 @@ class core_plugin_manager {
 
             'filter' => array(
                 'activitynames', 'algebra', 'censor', 'emailprotect',
-                'emoticon', 'mathjaxloader', 'mediaplugin', 'multilang', 'tex', 'tidy',
+                'emoticon', 'h5p', 'mathjaxloader', 'mediaplugin', 'multilang', 'tex', 'tidy',
                 'urltolink', 'data', 'glossary'
             ),
 
-- 
2.11.4.GIT