Make the tuple length in com_finder configurable

joomla · Jun 3, 2018 · c26f898 · c26f898
1 parent 3a13f8f
commit c26f898
Show file tree

Hide file tree

Showing 4 changed files with 61 additions and 64 deletions.
diff --git a/administrator/components/com_finder/config.xml b/administrator/components/com_finder/config.xml
@@ -201,6 +201,17 @@
 		description="COM_FINDER_FIELDSET_INDEX_OPTIONS_DESCRIPTION"
 		>
 
+		<field
+			name="tuplecount"
+			type="integer"
+			label="COM_FINDER_CONFIG_TUPLECOUNT_LABEL"
+			default="1"
+			validate="options"
+			first="1"
+			last="10"
+			step="1"
+		/>
+
 		<field
 			name="batch_size"
 			type="list"

diff --git a/administrator/components/com_finder/helpers/indexer/helper.php b/administrator/components/com_finder/helpers/indexer/helper.php
@@ -72,7 +72,7 @@ public static function parse($input, $format = 'html')
 	 */
 	public static function tokenize($input, $lang, $phrase = false)
 	{
-		static $cache;
+		static $cache, $tuplecount;
 		$store = StringHelper::strlen($input) < 128 ? md5($input . '::' . $lang . '::' . $phrase) : null;
 
 		// Check if the string has been tokenized already.
@@ -81,6 +81,12 @@ public static function tokenize($input, $lang, $phrase = false)
 			return $cache[$store];
 		}
 
+		if (!$tuplecount)
+		{
+			$params = ComponentHelper::getParams('com_finder');
+			$tuplecount = $params->get('tuplecount', 1);
+		}
+
 		$tokens = array();
 		$quotes = html_entity_decode('&#8216;&#8217;&#39;', ENT_QUOTES, 'UTF-8');
 
@@ -168,33 +174,28 @@ public static function tokenize($input, $lang, $phrase = false)
 				$tokens[] = new FinderIndexerToken($terms[$i], $lang);
 			}
 
-			// Create two and three word phrase tokens from the individual words.
-			for ($i = 0, $n = count($tokens); $i < $n; $i++)
+			// Create multi-word phrases tokens from the individual words.
+			if ($tuplecount > 1)
 			{
-				// Setup the phrase positions.
-				$i2 = $i + 1;
-				$i3 = $i + 2;
-
-				// Create the two word phrase.
-				if ($i2 < $n && isset($tokens[$i2]))
+				for ($i = 0, $n = count($tokens); $i < $n; $i++)
 				{
-					// Tokenize the two word phrase.
-					$token = new FinderIndexerToken(array($tokens[$i]->term, $tokens[$i2]->term), $lang, $lang === 'zh' ? '' : ' ');
-					$token->derived = true;
+					$temp = array($tokens[$i]->term);
 
-					// Add the token to the stack.
-					$tokens[] = $token;
-				}
+					// Create tokens for 2 to $tuplecount length phrases
+					for ($j = 1; $j < $tuplecount; $j++)
+					{
+						if ($i + $j >= $n || !isset($tokens[$i + $j]))
+						{
+							break;
+						}
 
-				// Create the three word phrase.
-				if ($i3 < $n && isset($tokens[$i3]))
-				{
-					// Tokenize the three word phrase.
-					$token = new FinderIndexerToken(array($tokens[$i]->term, $tokens[$i2]->term, $tokens[$i3]->term), $lang, $lang === 'zh' ? '' : ' ');
-					$token->derived = true;
+						$temp[] = $tokens[$i + $j]->term;
+						$token = new FinderIndexerToken($temp, $lang, $lang === 'zh' ? '' : ' ');
+						$token->derived = true;
 
-					// Add the token to the stack.
-					$tokens[] = $token;
+						// Add the token to the stack.
+						$tokens[] = $token;
+					}
 				}
 			}
 		}

diff --git a/administrator/components/com_finder/helpers/indexer/query.php b/administrator/components/com_finder/helpers/indexer/query.php
@@ -9,6 +9,7 @@
 
 defined('_JEXEC') or die;
 
+use Joomla\CMS\Component\ComponentHelper;
 use Joomla\Registry\Registry;
 use Joomla\String\StringHelper;
 use Joomla\Utilities\ArrayHelper;
@@ -731,11 +732,12 @@ protected function processDates($date1, $date2, $when1, $when2)
 	protected function processString($input, $lang, $mode)
 	{
 		// Clean up the input string.
-		$input = html_entity_decode($input, ENT_QUOTES, 'UTF-8');
-		$input = StringHelper::strtolower($input);
-		$input = preg_replace('#\s+#mi', ' ', $input);
-		$input = trim($input);
-		$debug = JFactory::getConfig()->get('debug_lang');
+		$input	= html_entity_decode($input, ENT_QUOTES, 'UTF-8');
+		$input	= StringHelper::strtolower($input);
+		$input	= preg_replace('#\s+#mi', ' ', $input);
+		$input	= trim($input);
+		$debug	= JFactory::getConfig()->get('debug_lang');
+		$params	= ComponentHelper::getParams('com_finder');
 
 		/*
 		 * First, we need to handle string based modifiers. String based
@@ -893,51 +895,33 @@ protected function processString($input, $lang, $mode)
 
 					// Get the number of words in the phrase.
 					$parts = explode(' ', $match);
+					$tuplecount = $params->get('tuplecount', 1);
 
 					// Check if the phrase is longer than three words.
-					if (count($parts) > 3)
+					if (count($parts) > $tuplecount && $tuplecount > 1)
 					{
+						$chunk = array_slice($parts, 0, $tuplecount);
+						$parts = array_slice($parts, $tuplecount);
+
+						// If the chunk is not empty, add it as a phrase.
+						if (count($chunk))
+						{
+							$phrases[] = implode(' ', $chunk);
+							$terms[] = implode(' ', $chunk);
+						}
+
 						/*
-						 * If the phrase is longer than three words, we need to
+						 * If the phrase is longer than $tuplecount words, we need to
 						 * break it down into smaller chunks of phrases that
-						 * are less than or equal to three words. We overlap
+						 * are less than or equal to $tuplecount words. We overlap
 						 * the chunks so that we can ensure that a match is
 						 * found for the complete phrase and not just portions
 						 * of it.
 						 */
-						for ($i = 0, $c = count($parts); $i < $c; $i += 2)
+						for ($i = 0, $c = count($parts); $i < $c; $i += 1)
 						{
-							// Set up the chunk.
-							$chunk = array();
-
-							// The chunk has to be assembled based on how many
-							// pieces are available to use.
-							switch ($c - $i)
-							{
-								/*
-								 * If only one word is left, we can break from
-								 * the switch and loop because the last word
-								 * was already used at the end of the last
-								 * chunk.
-								 */
-								case 1:
-									break 2;
-
-								// If there words are left, we use them both as
-								// the last chunk of the phrase and we're done.
-								case 2:
-									$chunk[] = $parts[$i];
-									$chunk[] = $parts[$i + 1];
-									break;
-
-								// If there are three or more words left, we
-								// build a three word chunk and continue on.
-								default:
-									$chunk[] = $parts[$i];
-									$chunk[] = $parts[$i + 1];
-									$chunk[] = $parts[$i + 2];
-									break;
-							}
+							array_shift($chunk);
+							$chunk[] = array_shift($parts);
 
 							// If the chunk is not empty, add it as a phrase.
 							if (count($chunk))
@@ -949,7 +933,7 @@ protected function processString($input, $lang, $mode)
 					}
 					else
 					{
-						// The phrase is <= 3 words so we can use it as is.
+						// The phrase is <= $tuplecount words so we can use it as is.
 						$phrases[] = $match;
 						$terms[] = $match;
 					}

diff --git a/administrator/language/en-GB/en-GB.com_finder.ini b/administrator/language/en-GB/en-GB.com_finder.ini
@@ -40,6 +40,7 @@ COM_FINDER_CONFIG_STEMMER_FR="French Only"
 COM_FINDER_CONFIG_STEMMER_LABEL="Select Language Stemmer"
 COM_FINDER_CONFIG_STEMMER_PORTER_EN="English Only"
 COM_FINDER_CONFIG_STEMMER_SNOWBALL="Snowball"
+COM_FINDER_CONFIG_TUPLECOUNT_LABEL="Length indexed tuples"
 COM_FINDER_CONFIG_TEXT_MULTIPLIER_DESCRIPTION="The multiplier is used to control how much influence matching text has on the overall relevance score of a search result. A multiplier is considered in relationship to the other multipliers. The body text comes from the summary and/or body of the content."
 COM_FINDER_CONFIG_TEXT_MULTIPLIER_LABEL="Body Text Weight Multiplier"
 COM_FINDER_CONFIG_TITLE_MULTIPLIER_DESCRIPTION="The multiplier is used to control how much influence matching text has on the overall relevance score of a search result. A multiplier is considered in relationship to the other multipliers. The title text comes from the title of the content."