Resolves #122 by falling back to a full-spec YAML parser by default

src/Addon.php

@@ -18,7 +18,7 @@ class Addon
     public const CODE_ADDON_NAME = 'meerkat';
     public const ROUTE_PREFIX = 'meerkat';
     public const ADDON_ID = 'stillat/meerkat';
-    public const VERSION = '2.1.5-beta2';
+    public const VERSION = '2.1.6-beta2';
 
     /**
      * Gets the addon API prefix.

src/Core/Comments/Comment.php

@@ -655,6 +655,17 @@ public function getStorableAttributes()
         return $this->getAttributesToSave();
     }
 
+    /**
+     * Indicates that runtime attributes should already be considered resolved.
+     *
+     * @internal
+     * @since 2.1.6
+     */
+    public function flagRuntimeAttributesResolved()
+    {
+        $this->runTimeAttributesResolved = true;
+    }
+
     /**
      * Parses the raw attributes and reloads the current attributes.
      *

src/Core/Configuration.php

@@ -131,6 +131,20 @@ class Configuration extends ConfigurationContainer
      */
     public $addressToSendEmailTo = [];
 
+    /**
+     * Indicates if a full, or off-spec, parser should be used for comment prototype parsing.
+     *
+     * If all comments of a site adhere to Meerkat's recommended structure, this can be
+     * set to `true` to receive comment parsing and loading performance improvements.
+     *
+     * Leave this set to `false` if you are doing weird things with your data,
+     * or are having issues with migrating comments from old Meerkat versions.
+     *
+     * @since 2.1.6
+     * @var bool
+     */
+    public $useSlimCommentPrototypeParser = false;
+
     /**
      * The comment length limit. If this limit is reached when
      * reading the comment's content data will be truncated.

src/Core/Contracts/Parsing/PrototypeParserContract.php

@@ -0,0 +1,50 @@
+<?php
+
+namespace Stillat\Meerkat\Core\Contracts\Parsing;
+
+use Stillat\Meerkat\Core\Configuration;
+
+/**
+ * Interface PrototypeParserContract
+ *
+ * Represents a parser capable of converting a YAML
+ * text file into a Meerkat Comment prototype.
+ *
+ * @package Stillat\Meerkat\Core\Contracts\Parsing
+ * @since 2.1.6
+ */
+interface PrototypeParserContract
+{
+
+    /**
+     * Sets the comment's truthy prototype elements.
+     *
+     * @param array $elements The truthy prototype elements.
+     */
+    public function setTruthyElements($elements);
+
+    /**
+     * Sets the Meerkat Core configuration instance.
+     *
+     * @param Configuration $configuration The configuration.
+     */
+    public function setConfig(Configuration $configuration);
+
+    /**
+     * Sets the prototype elements.
+     *
+     * @param array $elements The prototype elements.
+     */
+    public function setPrototypeElements($elements);
+
+    /**
+     * Retrieves only the core meta-data for the comment.
+     *
+     * Supplemental data and content are ignored during this phase.
+     *
+     * @param string $path The full path to the comment data.
+     * @return array
+     */
+    public function getCommentPrototype($path);
+
+}

src/Core/Parsing/CommentPrototypeParser.php

@@ -5,6 +5,7 @@
 use Exception;
 use Stillat\Meerkat\Core\Configuration;
 use Stillat\Meerkat\Core\Contracts\Comments\CommentContract;
+use Stillat\Meerkat\Core\Contracts\Parsing\PrototypeParserContract;
 use Stillat\Meerkat\Core\Errors;
 use Stillat\Meerkat\Core\Logging\ExceptionLoggerFactory;
 use Stillat\Meerkat\Core\Logging\LocalErrorCodeRepository;
@@ -20,7 +21,7 @@
  * @package Stillat\Meerkat\Core\Parsing
  * @since 2.0.11
  */
-class CommentPrototypeParser
+class CommentPrototypeParser implements PrototypeParserContract
 {
     const UNIX_EPOCH_ID = '0000000000';
 

src/Core/Parsing/FullCommentPrototypeParser.php

@@ -0,0 +1,129 @@
+<?php
+
+namespace Stillat\Meerkat\Core\Parsing;
+
+use Stillat\Meerkat\Core\Configuration;
+use Stillat\Meerkat\Core\Contracts\Comments\CommentContract;
+use Stillat\Meerkat\Core\Contracts\Parsing\PrototypeParserContract;
+use Stillat\Meerkat\Core\Contracts\Parsing\YAMLParserContract;
+use Stillat\Meerkat\Core\Storage\Drivers\Local\LocalCommentStorageManager;
+
+/**
+ * Class FullCommentPrototypeParser
+ *
+ * Utilizes a full YAML-spec parser to produce a comment prototype.
+ *
+ * This prototype parser produces the most consistent results across a wide
+ * range of Comment data structures, but with a higher performance cost.
+ *
+ * @package Stillat\Meerkat\Core\Parsing
+ * @since 2.1.6
+ */
+class FullCommentPrototypeParser implements PrototypeParserContract
+{
+
+    /**
+     * The prototype comment elements to parse.
+     *
+     * @var array
+     */
+    protected $prototypeElements = [];
+
+    /**
+     * The Meerkat Core configuration.
+     *
+     * @var Configuration
+     */
+    protected $config = null;
+
+    /**
+     * The prototype elements that should be converted to boolean values.
+     *
+     * @var array
+     */
+    protected $truthyPrototypeElements = [];
+
+    /**
+     * The YAMLParserContract implementation instance.
+     *
+     * @var YAMLParserContract
+     */
+    protected $yamlParser = null;
+
+    public function __construct(YAMLParserContract $yamlParser)
+    {
+        $this->yamlParser = $yamlParser;
+    }
+
+    /**
+     * Sets the comment's truthy prototype elements.
+     *
+     * @param array $elements The truthy prototype elements.
+     */
+    public function setTruthyElements($elements)
+    {
+        $this->truthyPrototypeElements = $elements;
+    }
+
+    /**
+     * Sets the Meerkat Core configuration instance.
+     *
+     * @param Configuration $configuration The configuration.
+     */
+    public function setConfig(Configuration $configuration)
+    {
+        $this->config = $configuration;
+    }
+
+    /**
+     * Sets the prototype elements.
+     *
+     * @param array $elements The prototype elements.
+     */
+    public function setPrototypeElements($elements)
+    {
+        $this->prototypeElements = $elements;
+    }
+
+    /**
+     * Retrieves only the core meta-data for the comment.
+     *
+     * Supplemental data and content are ignored during this phase.
+     *
+     * @param string $path The full path to the comment data.
+     * @return array
+     */
+    public function getCommentPrototype($path)
+    {
+        $contents = file_get_contents($path);
+        $parsedComment = $this->yamlParser->parseDocument($contents);
+        $commentContent = '';
+        $alreadyFoundContent = false;
+
+        if (array_key_exists(CommentContract::KEY_LEGACY_COMMENT, $parsedComment)) {
+            $commentContent = $parsedComment[CommentContract::KEY_LEGACY_COMMENT];
+            $alreadyFoundContent = true;
+
+            unset($parsedComment[CommentContract::KEY_LEGACY_COMMENT]);
+        } else if (array_key_exists(CommentContract::KEY_CONTENT, $parsedComment)) {
+            $commentContent = $parsedComment[CommentContract::KEY_CONTENT];
+
+            unset($parsedComment[CommentContract::KEY_CONTENT]);
+        }
+
+        // Reset some types.
+        if (array_key_exists(CommentContract::KEY_ID, $parsedComment)) {
+            if (is_int($parsedComment[CommentContract::KEY_ID])) {
+                $parsedComment[CommentContract::KEY_ID] = strval($parsedComment[CommentContract::KEY_ID]);
+            }
+        }
+
+        return [
+            LocalCommentStorageManager::KEY_HEADERS => $parsedComment,
+            LocalCommentStorageManager::KEY_RAW_HEADERS => $parsedComment,
+            LocalCommentStorageManager::KEY_CONTENT => $commentContent,
+            LocalCommentStorageManager::KEY_NEEDS_MIGRATION => $alreadyFoundContent
+        ];
+    }
+
+}

src/Core/Storage/Drivers/Local/LocalCommentStorageManager.php

@@ -19,6 +19,7 @@
 use Stillat\Meerkat\Core\Contracts\Comments\CommentMutationPipelineContract;
 use Stillat\Meerkat\Core\Contracts\Identity\IdentityManagerContract;
 use Stillat\Meerkat\Core\Contracts\Parsing\MarkdownParserContract;
+use Stillat\Meerkat\Core\Contracts\Parsing\PrototypeParserContract;
 use Stillat\Meerkat\Core\Contracts\Parsing\YAMLParserContract;
 use Stillat\Meerkat\Core\Contracts\Storage\CommentChangeSetStorageManagerContract;
 use Stillat\Meerkat\Core\Contracts\Storage\CommentStorageManagerContract;
@@ -31,6 +32,7 @@
 use Stillat\Meerkat\Core\Exceptions\MutationException;
 use Stillat\Meerkat\Core\Logging\ExceptionLoggerFactory;
 use Stillat\Meerkat\Core\Parsing\CommentPrototypeParser;
+use Stillat\Meerkat\Core\Parsing\FullCommentPrototypeParser;
 use Stillat\Meerkat\Core\Paths\PathUtilities;
 use Stillat\Meerkat\Core\RuntimeStateGuard;
 use Stillat\Meerkat\Core\Storage\Data\CommentAuthorRetriever;
@@ -219,9 +221,9 @@ class LocalCommentStorageManager implements CommentStorageManagerContract
     private $changeSetManager = null;
 
     /**
-     * The CommentPrototypeParser instance.
+     * The PrototypeParserContract implementation instance.
      *
-     * @var CommentPrototypeParser
+     * @var PrototypeParserContract
      */
     private $commentParser = null;
 
@@ -235,7 +237,13 @@ public function __construct(
         IdentityManagerContract $identityManager,
         CommentChangeSetStorageManagerContract $changeSetManager)
     {
-        $this->commentParser = new CommentPrototypeParser();
+
+        if ($config->useSlimCommentPrototypeParser === false) {
+            $this->commentParser = new FullCommentPrototypeParser($yamlParser);
+        } else {
+            $this->commentParser = new CommentPrototypeParser();
+        }
+
         $this->commentShadowIndex = new ShadowIndex($config);
         $this->commentStructureResolver = new LocalCommentStructureResolver();
         $this->authorRetriever = $authorRetriever;
@@ -395,6 +403,10 @@ private function getThreadHierarchy($threadPath, $threadId, $commentPaths)
                 $comment->setDataAttribute(CommentContract::INTERNAL_STRUCTURE_NEEDS_MIGRATION, true);
             }
 
+            if ($this->config->useSlimCommentPrototypeParser === false) {
+                $comment->flagRuntimeAttributesResolved();
+            }
+
             $commentPrototypes[$commentId] = $comment;
         }
 

src/ServiceProvider.php

@@ -228,6 +228,9 @@ private function registerMeerkatGlobalConfiguration()
             $globalConfiguration->searchableAttributes = $this->getConfig('search.attributes', []);
             $globalConfiguration->onlyAcceptCommentsFromAuthenticatedUser = $this->getConfig('publishing.only_accept_comments_from_authenticated_users', false);
 
+            // Internal parser configuration.
+            $globalConfiguration->useSlimCommentPrototypeParser = $this->getConfig('internals.useFasterSlimCommentParser', false);
+
             // Storage directories.
             $globalConfiguration->storageDirectory = PathProvider::contentPath();
             $globalConfiguration->indexDirectory = storage_path('meerkat/index');