Add a setting to retrieve bookmark metadata asynchrounously

  - There is a new standalone script (metadata.js) which requests
    a new controller to get bookmark metadata and fill the form async
  - This feature is enabled with the new setting: general.enable_async_metadata
    (enabled by default)
  - general.retrieve_description is now enabled by default
  - A small rotating loader animation has a been added to bookmark inputs
    when metadata is being retrieved (default template)
  - Custom JS htmlentities has been removed and  mathiasbynens/he
    library is used instead

Fixes #1563

6 files changed, 118 insertions, 25 deletions

diff --git a/application/config/ConfigManager.php b/application/config/ConfigManager.php
index 4c98be30..fb085023 100644
--- a/application/config/ConfigManager.php
+++ b/application/config/ConfigManager.php
@@ -366,7 +366,8 @@ class ConfigManager
         $this->setEmpty('general.links_per_page', 20);
         $this->setEmpty('general.enabled_plugins', self::$DEFAULT_PLUGINS);
         $this->setEmpty('general.default_note_title', 'Note: ');
-        $this->setEmpty('general.retrieve_description', false);
+        $this->setEmpty('general.retrieve_description', true);
+        $this->setEmpty('general.enable_async_metadata', true);
 
         $this->setEmpty('updates.check_updates', false);
         $this->setEmpty('updates.check_updates_branch', 'stable');
diff --git a/application/container/ContainerBuilder.php b/application/container/ContainerBuilder.php
index c21d58dd..fd94a1c3 100644
--- a/application/container/ContainerBuilder.php
+++ b/application/container/ContainerBuilder.php
@@ -14,6 +14,7 @@ use Shaarli\Front\Controller\Visitor\ErrorController;
 use Shaarli\Front\Controller\Visitor\ErrorNotFoundController;
 use Shaarli\History;
 use Shaarli\Http\HttpAccess;
+use Shaarli\Http\MetadataRetriever;
 use Shaarli\Netscape\NetscapeBookmarkUtils;
 use Shaarli\Plugin\PluginManager;
 use Shaarli\Render\PageBuilder;
@@ -90,6 +91,10 @@ class ContainerBuilder
             );
         };
 
+        $container['metadataRetriever'] = function (ShaarliContainer $container): MetadataRetriever {
+            return new MetadataRetriever($container->conf, $container->httpAccess);
+        };
+
         $container['pageBuilder'] = function (ShaarliContainer $container): PageBuilder {
             return new PageBuilder(
                 $container->conf,
diff --git a/application/container/ShaarliContainer.php b/application/container/ShaarliContainer.php
index 66e669aa..3a7c238f 100644
--- a/application/container/ShaarliContainer.php
+++ b/application/container/ShaarliContainer.php
@@ -10,6 +10,7 @@ use Shaarli\Feed\FeedBuilder;
 use Shaarli\Formatter\FormatterFactory;
 use Shaarli\History;
 use Shaarli\Http\HttpAccess;
+use Shaarli\Http\MetadataRetriever;
 use Shaarli\Netscape\NetscapeBookmarkUtils;
 use Shaarli\Plugin\PluginManager;
 use Shaarli\Render\PageBuilder;
@@ -35,6 +36,7 @@ use Slim\Container;
  * @property History                  $history
  * @property HttpAccess               $httpAccess
  * @property LoginManager             $loginManager
+ * @property MetadataRetriever        $metadataRetriever
  * @property NetscapeBookmarkUtils    $netscapeBookmarkUtils
  * @property callable                 $notFoundHandler       Overrides default Slim exception display
  * @property PageBuilder              $pageBuilder
diff --git a/application/front/controller/admin/ManageShaareController.php b/application/front/controller/admin/ManageShaareController.php
index bb083486..df2f1631 100644
--- a/application/front/controller/admin/ManageShaareController.php
+++ b/application/front/controller/admin/ManageShaareController.php
@@ -53,36 +53,22 @@ class ManageShaareController extends ShaarliAdminController
 
             // If this is an HTTP(S) link, we try go get the page to extract
             // the title (otherwise we will to straight to the edit form.)
-            if (empty($title) && strpos(get_url_scheme($url) ?: '', 'http') !== false) {
-                $retrieveDescription = $this->container->conf->get('general.retrieve_description');
-                // Short timeout to keep the application responsive
-                // The callback will fill $charset and $title with data from the downloaded page.
-                $this->container->httpAccess->getHttpResponse(
-                    $url,
-                    $this->container->conf->get('general.download_timeout', 30),
-                    $this->container->conf->get('general.download_max_size', 4194304),
-                    $this->container->httpAccess->getCurlDownloadCallback(
-                        $charset,
-                        $title,
-                        $description,
-                        $tags,
-                        $retrieveDescription
-                    )
-                );
-                if (! empty($title) && strtolower($charset) !== 'utf-8' && mb_check_encoding($charset)) {
-                    $title = mb_convert_encoding($title, 'utf-8', $charset);
-                }
+            if (true !== $this->container->conf->get('general.enable_async_metadata', true)
+                && empty($title)
+                && strpos(get_url_scheme($url) ?: '', 'http') !== false
+            ) {
+                $metadata = $this->container->metadataRetriever->retrieve($url);
             }
 
-            if (empty($url) && empty($title)) {
-                $title = $this->container->conf->get('general.default_note_title', t('Note: '));
+            if (empty($url)) {
+                $metadata['title'] = $this->container->conf->get('general.default_note_title', t('Note: '));
             }
 
             $link = [
-                'title' => $title,
+                'title' => $title ?? $metadata['title'] ?? '',
                 'url' => $url ?? '',
-                'description' => $description ?? '',
-                'tags' => $tags ?? '',
+                'description' => $description ?? $metadata['description'] ?? '',
+                'tags' => $tags ?? $metadata['tags'] ?? '',
                 'private' => $private,
             ];
         } else {
@@ -352,6 +338,8 @@ class ManageShaareController extends ShaarliAdminController
             'source' => $request->getParam('source') ?? '',
             'tags' => $tags,
             'default_private_links' => $this->container->conf->get('privacy.default_private_links', false),
+            'async_metadata' => $this->container->conf->get('general.enable_async_metadata', true),
+            'retrieve_description' => $this->container->conf->get('general.retrieve_description', false),
         ]);
 
         $this->executePageHooks('render_editlink', $data, TemplatePage::EDIT_LINK);
diff --git a/application/front/controller/admin/MetadataController.php b/application/front/controller/admin/MetadataController.php
new file mode 100644
index 00000000..ff845944
--- /dev/null
+++ b/application/front/controller/admin/MetadataController.php
@@ -0,0 +1,29 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Shaarli\Front\Controller\Admin;
+
+use Slim\Http\Request;
+use Slim\Http\Response;
+
+/**
+ * Controller used to retrieve/update bookmark's metadata.
+ */
+class MetadataController extends ShaarliAdminController
+{
+    /**
+     * GET /admin/metadata/{url} - Attempt to retrieve the bookmark title from provided URL.
+     */
+    public function ajaxRetrieveTitle(Request $request, Response $response): Response
+    {
+        $url = $request->getParam('url');
+
+        // Only try to extract metadata from URL with HTTP(s) scheme
+        if (!empty($url) && strpos(get_url_scheme($url) ?: '', 'http') !== false) {
+            return $response->withJson($this->container->metadataRetriever->retrieve($url));
+        }
+
+        return $response->withJson([]);
+    }
+}
diff --git a/application/http/MetadataRetriever.php b/application/http/MetadataRetriever.php
new file mode 100644
index 00000000..2ca982e2
--- /dev/null
+++ b/application/http/MetadataRetriever.php
@@ -0,0 +1,68 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Shaarli\Http;
+
+use Shaarli\Config\ConfigManager;
+
+/**
+ * HTTP Tool used to extract metadata from external URL (title, description, etc.).
+ */
+class MetadataRetriever
+{
+    /** @var ConfigManager */
+    protected $conf;
+
+    /** @var HttpAccess */
+    protected $httpAccess;
+
+    public function __construct(ConfigManager $conf, HttpAccess $httpAccess)
+    {
+        $this->conf = $conf;
+        $this->httpAccess = $httpAccess;
+    }
+
+    /**
+     * Retrieve metadata for given URL.
+     *
+     * @return array [
+     *                  'title' => <remote title>,
+     *                  'description' => <remote description>,
+     *                  'tags' => <remote keywords>,
+     *               ]
+     */
+    public function retrieve(string $url): array
+    {
+        $charset = null;
+        $title = null;
+        $description = null;
+        $tags = null;
+        $retrieveDescription = $this->conf->get('general.retrieve_description');
+
+        // Short timeout to keep the application responsive
+        // The callback will fill $charset and $title with data from the downloaded page.
+        $this->httpAccess->getHttpResponse(
+            $url,
+            $this->conf->get('general.download_timeout', 30),
+            $this->conf->get('general.download_max_size', 4194304),
+            $this->httpAccess->getCurlDownloadCallback(
+                $charset,
+                $title,
+                $description,
+                $tags,
+                $retrieveDescription
+            )
+        );
+
+        if (!empty($title) && strtolower($charset) !== 'utf-8') {
+            $title = mb_convert_encoding($title, 'utf-8', $charset);
+        }
+
+        return [
+            'title' => $title,
+            'description' => $description,
+            'tags' => $tags,
+        ];
+    }
+}