Merge PR #832 branch 'dhruvgoel92/feature/postgresql-fulltext-search' into develop

* dhruvgoel92/feature/postgresql-fulltext-search: (28 commits) [feature/postgresql-fulltext-search] supports_phrase_search function [feature/postgresql-fulltext-search] use proper variable name [feature/postgresql-fulltext-search] each config is checked if it is set [feature/postgresql-fulltext-search] fix formatting [feature/postgresql-fulltext-search] add docblocks to functions [feature/postgresql-fulltext-search] remove phrase search code [feature/postgresql-fulltext-search] define access control specifiers [feature/postgresql-fulltext-search] use version_compare [feature/postgresql-fulltext-search] fix language key [feature/postgresql-fulltext-search] remove Readme [feature/postgresql-fulltext-search] removes pcre check [feature/postgresql-fulltext-search] fix language tsearch2 to text search [feature/postgresql-fulltext-search] remove mbstring support [feature/postgresql-fulltext-search] change language for pgsql < 8.3 [feature/postgresql-fulltext-search] use phpbb_pcre_utf8_support() [feature/postgresql-fulltext-search] fix variable name [feature/postgresql-fulltext-search] Fix version comparison for 9.0+. [feature/postgresql-fulltext-search] fixing license [feature/postgresql-fulltext-search] assign empty stats [feature/postgresql-fulltext-search] remove backward compatibility ...
author: Oleg Pudeyev <oleg@bsdpower.com> 2012-07-07 16:19:53 -0400
committer: Oleg Pudeyev <oleg@bsdpower.com> 2012-07-07 16:19:53 -0400
commit: 1526886e3ec2b475bb7f578767f6430dbfa58097 (patch)
tree: e21dfc5b9eac9fd26b075bee1851efc47272d5e4 /phpBB
parent: d9fd0cce0a5b323bff785af401e90370a3d73972 (diff)
parent: 70a0caee28559bd1bd52b4ecf9d7d33acd732fe2 (diff)
download: forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar
forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.gz
forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.bz2
forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.xz
forums-1526886e3ec2b475bb7f578767f6430dbfa58097.zip
7 files changed, 898 insertions, 0 deletions
diff --git a/phpBB/includes/search/fulltext_postgres.php b/phpBB/includes/search/fulltext_postgres.php
new file mode 100644
index 0000000000..84ce674564
--- /dev/null
+++ b/phpBB/includes/search/fulltext_postgres.php
@@ -0,0 +1,856 @@
+<?php
+/**
+*
+* @package search
+* @copyright (c) 2005 phpBB Group
+* @license http://opensource.org/licenses/gpl-2.0.php GNU General Public License v2
+*
+*/
+
+/**
+* @ignore
+*/
+if (!defined('IN_PHPBB'))
+{
+	exit;
+}
+
+/**
+* fulltext_postgres
+* Fulltext search for PostgreSQL
+* @package search
+*/
+class phpbb_search_fulltext_postgres extends phpbb_search_base
+{
+	private $stats = array();
+	private $split_words = array();
+	private $tsearch_usable = false;
+	private $version;
+	private $tsearch_query;
+	private $phrase_search = false;
+	public $search_query;
+	public $common_words = array();
+	public $word_length = array();
+
+	/**
+	 * Constructor
+	 * Creates a new phpbb_search_fulltext_postgres, which is used as a search backend.
+	 *
+	 * @param string|bool $error Any error that occurs is passed on through this reference variable otherwise false
+	 */
+	public function __construct(&$error)
+	{
+		global $db, $config;
+
+		$this->word_length = array('min' => $config['fulltext_postgres_min_word_len'], 'max' => $config['fulltext_postgres_max_word_len']);
+
+
+		if ($db->sql_layer == 'postgres')
+		{
+			$pgsql_version = explode(',', substr($db->sql_server_info(), 10));
+			$this->version = trim($pgsql_version[0]);
+			if (version_compare($this->version, '8.3', '>='))
+			{
+				$this->tsearch_usable = true;
+			}
+		}
+
+		$error = false;
+	}
+
+	/**
+	* Returns the name of this search backend to be displayed to administrators
+	*
+	* @return string Name
+	*
+	* @access public
+	*/
+	public function get_name()
+	{
+		return 'PostgreSQL Fulltext';
+	}
+
+	/**
+	 * Returns if phrase search is supported or not
+	 *
+	 * @return bool
+	 *
+	 * @access public
+	 */
+	public function supports_phrase_search()
+	{
+		return $this->phrase_search;
+	}
+
+	/**
+	* Checks for correct PostgreSQL version and stores min/max word length in the config
+	*
+	* @return string|bool Language key of the error/incompatiblity occured
+	*
+	* @access public
+	*/
+	function init()
+	{
+		global $db, $user;
+
+		if ($db->sql_layer != 'postgres')
+		{
+			return $user->lang['FULLTEXT_POSTGRES_INCOMPATIBLE_DATABASE'];
+		}
+
+		if (!$this->tsearch_usable)
+		{
+			return $user->lang['FULLTEXT_POSTGRES_TS_NOT_USABLE'];
+		}
+
+		return false;
+	}
+
+	/**
+	* Splits keywords entered by a user into an array of words stored in $this->split_words
+	* Stores the tidied search query in $this->search_query
+	*
+	* @param	string	&$keywords	Contains the keyword as entered by the user
+	* @param	string	$terms	is either 'all' or 'any'
+	* @return	bool	false	if no valid keywords were found and otherwise true
+	*
+	* @access	public
+	*/
+	function split_keywords(&$keywords, $terms)
+	{
+		global $config;
+
+		if ($terms == 'all')
+		{
+			$match		= array('#\sand\s#iu', '#\sor\s#iu', '#\snot\s#iu', '#\+#', '#-#', '#\|#');
+			$replace	= array(' +', ' |', ' -', ' +', ' -', ' |');
+
+			$keywords = preg_replace($match, $replace, $keywords);
+		}
+
+		// Filter out as above
+		$split_keywords = preg_replace("#[\"\n\r\t]+#", ' ', trim(htmlspecialchars_decode($keywords)));
+
+		// Split words
+		$split_keywords = preg_replace('#([^\p{L}\p{N}\'*"()])#u', '$1$1', str_replace('\'\'', '\' \'', trim($split_keywords)));
+		$matches = array();
+		preg_match_all('#(?:[^\p{L}\p{N}*"()]|^)([+\-|]?(?:[\p{L}\p{N}*"()]+\'?)*[\p{L}\p{N}*"()])(?:[^\p{L}\p{N}*"()]|$)#u', $split_keywords, $matches);
+		$this->split_words = $matches[1];
+
+		foreach ($this->split_words as $i => $word)
+		{
+			$clean_word = preg_replace('#^[+\-|"]#', '', $word);
+
+			// check word length
+			$clean_len = utf8_strlen(str_replace('*', '', $clean_word));
+			if (($clean_len < $config['fulltext_postgres_min_word_len']) || ($clean_len > $config['fulltext_postgres_max_word_len']))
+			{
+				$this->common_words[] = $word;
+				unset($this->split_words[$i]);
+			}
+		}
+
+		if ($terms == 'any')
+		{
+			$this->search_query = '';
+			$this->tsearch_query = '';
+			foreach ($this->split_words as $word)
+			{
+				if ((strpos($word, '+') === 0) || (strpos($word, '-') === 0) || (strpos($word, '|') === 0))
+				{
+					$word = substr($word, 1);
+				}
+				$this->search_query .= $word . ' ';
+				$this->tsearch_query .= '|' . $word . ' ';
+			}
+		}
+		else
+		{
+			$this->search_query = '';
+			$this->tsearch_query = '';
+			foreach ($this->split_words as $word)
+			{
+				if (strpos($word, '+') === 0)
+				{
+					$this->search_query .= $word . ' ';
+					$this->tsearch_query .= '&' . substr($word, 1) . ' ';
+				}
+				elseif (strpos($word, '-') === 0)
+				{
+					$this->search_query .= $word . ' ';
+					$this->tsearch_query .= '&!' . substr($word, 1) . ' ';
+				}
+				elseif (strpos($word, '|') === 0)
+				{
+					$this->search_query .= $word . ' ';
+					$this->tsearch_query .= '|' . substr($word, 1) . ' ';
+				}
+				else
+				{
+					$this->search_query .= '+' . $word . ' ';
+					$this->tsearch_query .= '&' . $word . ' ';
+				}
+			}
+		}
+
+		$this->tsearch_query = substr($this->tsearch_query, 1);
+		$this->search_query = utf8_htmlspecialchars($this->search_query);
+
+		if ($this->search_query)
+		{
+			$this->split_words = array_values($this->split_words);
+			sort($this->split_words);
+			return true;
+		}
+		return false;
+	}
+
+	/**
+	* Turns text into an array of words
+	* @param string $text contains post text/subject
+	*
+	* @access public
+	*/
+	function split_message($text)
+	{
+		global $config;
+
+		// Split words
+		$text = preg_replace('#([^\p{L}\p{N}\'*])#u', '$1$1', str_replace('\'\'', '\' \'', trim($text)));
+		$matches = array();
+		preg_match_all('#(?:[^\p{L}\p{N}*]|^)([+\-|]?(?:[\p{L}\p{N}*]+\'?)*[\p{L}\p{N}*])(?:[^\p{L}\p{N}*]|$)#u', $text, $matches);
+		$text = $matches[1];
+
+		// remove too short or too long words
+		$text = array_values($text);
+		for ($i = 0, $n = sizeof($text); $i < $n; $i++)
+		{
+			$text[$i] = trim($text[$i]);
+			if (utf8_strlen($text[$i]) < $config['fulltext_postgres_min_word_len'] || utf8_strlen($text[$i]) > $config['fulltext_postgres_max_word_len'])
+			{
+				unset($text[$i]);
+			}
+		}
+
+		return array_values($text);
+	}
+
+	/**
+	* Performs a search on keywords depending on display specific params. You have to run split_keywords() first.
+	*
+	* @param	string		$type				contains either posts or topics depending on what should be searched for
+	* @param	string		$fields				contains either titleonly (topic titles should be searched), msgonly (only message bodies should be searched), firstpost (only subject and body of the first post should be searched) or all (all post bodies and subjects should be searched)
+	* @param	string		$terms				is either 'all' (use query as entered, words without prefix should default to "have to be in field") or 'any' (ignore search query parts and just return all posts that contain any of the specified words)
+	* @param	array		$sort_by_sql		contains SQL code for the ORDER BY part of a query
+	* @param	string		$sort_key			is the key of $sort_by_sql for the selected sorting
+	* @param	string		$sort_dir			is either a or d representing ASC and DESC
+	* @param	string		$sort_days			specifies the maximum amount of days a post may be old
+	* @param	array		$ex_fid_ary			specifies an array of forum ids which should not be searched
+	* @param	array		$m_approve_fid_ary	specifies an array of forum ids in which the searcher is allowed to view unapproved posts
+	* @param	int			$topic_id			is set to 0 or a topic id, if it is not 0 then only posts in this topic should be searched
+	* @param	array		$author_ary			an array of author ids if the author should be ignored during the search the array is empty
+	* @param	string		$author_name		specifies the author match, when ANONYMOUS is also a search-match
+	* @param	array		&$id_ary			passed by reference, to be filled with ids for the page specified by $start and $per_page, should be ordered
+	* @param	int			$start				indicates the first index of the page
+	* @param	int			$per_page			number of ids each page is supposed to contain
+	* @return	boolean|int						total number of results
+	*
+	* @access	public
+	*/
+	function keyword_search($type, $fields, $terms, $sort_by_sql, $sort_key, $sort_dir, $sort_days, $ex_fid_ary, $m_approve_fid_ary, $topic_id, $author_ary, $author_name, &$id_ary, $start, $per_page)
+	{
+		global $config, $db;
+
+		// No keywords? No posts.
+		if (!$this->search_query)
+		{
+			return false;
+		}
+
+		// generate a search_key from all the options to identify the results
+		$search_key = md5(implode('#', array(
+			implode(', ', $this->split_words),
+			$type,
+			$fields,
+			$terms,
+			$sort_days,
+			$sort_key,
+			$topic_id,
+			implode(',', $ex_fid_ary),
+			implode(',', $m_approve_fid_ary),
+			implode(',', $author_ary)
+		)));
+
+		// try reading the results from cache
+		$result_count = 0;
+		if ($this->obtain_ids($search_key, $result_count, $id_ary, $start, $per_page, $sort_dir) == SEARCH_RESULT_IN_CACHE)
+		{
+			return $result_count;
+		}
+
+		$id_ary = array();
+
+		$join_topic = ($type == 'posts') ? false : true;
+
+		// Build sql strings for sorting
+		$sql_sort = $sort_by_sql[$sort_key] . (($sort_dir == 'a') ? ' ASC' : ' DESC');
+		$sql_sort_table = $sql_sort_join = '';
+
+		switch ($sql_sort[0])
+		{
+			case 'u':
+				$sql_sort_table	= USERS_TABLE . ' u, ';
+				$sql_sort_join	= ($type == 'posts') ? ' AND u.user_id = p.poster_id ' : ' AND u.user_id = t.topic_poster ';
+			break;
+
+			case 't':
+				$join_topic = true;
+			break;
+
+			case 'f':
+				$sql_sort_table	= FORUMS_TABLE . ' f, ';
+				$sql_sort_join	= ' AND f.forum_id = p.forum_id ';
+			break;
+		}
+
+		// Build some display specific sql strings
+		switch ($fields)
+		{
+			case 'titleonly':
+				$sql_match = 'p.post_subject';
+				$sql_match_where = ' AND p.post_id = t.topic_first_post_id';
+				$join_topic = true;
+			break;
+
+			case 'msgonly':
+				$sql_match = 'p.post_text';
+				$sql_match_where = '';
+			break;
+
+			case 'firstpost':
+				$sql_match = 'p.post_subject, p.post_text';
+				$sql_match_where = ' AND p.post_id = t.topic_first_post_id';
+				$join_topic = true;
+			break;
+
+			default:
+				$sql_match = 'p.post_subject, p.post_text';
+				$sql_match_where = '';
+			break;
+		}
+
+		if (!sizeof($m_approve_fid_ary))
+		{
+			$m_approve_fid_sql = ' AND p.post_approved = 1';
+		}
+		else if ($m_approve_fid_ary === array(-1))
+		{
+			$m_approve_fid_sql = '';
+		}
+		else
+		{
+			$m_approve_fid_sql = ' AND (p.post_approved = 1 OR ' . $db->sql_in_set('p.forum_id', $m_approve_fid_ary, true) . ')';
+		}
+
+		$sql_select			= ($type == 'posts') ? 'p.post_id' : 'DISTINCT t.topic_id';
+		$sql_from			= ($join_topic) ? TOPICS_TABLE . ' t, ' : '';
+		$field				= ($type == 'posts') ? 'post_id' : 'topic_id';
+		$sql_author			= (sizeof($author_ary) == 1) ? ' = ' . $author_ary[0] : 'IN (' . implode(', ', $author_ary) . ')';
+
+		if (sizeof($author_ary) && $author_name)
+		{
+			// first one matches post of registered users, second one guests and deleted users
+			$sql_author = '(' . $db->sql_in_set('p.poster_id', array_diff($author_ary, array(ANONYMOUS)), false, true) . ' OR p.post_username ' . $author_name . ')';
+		}
+		else if (sizeof($author_ary))
+		{
+			$sql_author = ' AND ' . $db->sql_in_set('p.poster_id', $author_ary);
+		}
+		else
+		{
+			$sql_author = '';
+		}
+
+		$sql_where_options = $sql_sort_join;
+		$sql_where_options .= ($topic_id) ? ' AND p.topic_id = ' . $topic_id : '';
+		$sql_where_options .= ($join_topic) ? ' AND t.topic_id = p.topic_id' : '';
+		$sql_where_options .= (sizeof($ex_fid_ary)) ? ' AND ' . $db->sql_in_set('p.forum_id', $ex_fid_ary, true) : '';
+		$sql_where_options .= $m_approve_fid_sql;
+		$sql_where_options .= $sql_author;
+		$sql_where_options .= ($sort_days) ? ' AND p.post_time >= ' . (time() - ($sort_days * 86400)) : '';
+		$sql_where_options .= $sql_match_where;
+
+		$tmp_sql_match = array();
+		foreach (explode(',', $sql_match) as $sql_match_column)
+		{
+			$tmp_sql_match[] = "to_tsvector ('" . $db->sql_escape($config['fulltext_postgres_ts_name']) . "', " . $sql_match_column . ") @@ to_tsquery ('" . $db->sql_escape($config['fulltext_postgres_ts_name']) . "', '" . $db->sql_escape($this->tsearch_query) . "')";
+		}
+
+		$sql = "SELECT $sql_select
+			FROM $sql_from$sql_sort_table" . POSTS_TABLE . " p
+			WHERE (" . implode(' OR ', $tmp_sql_match) . ")
+			$sql_where_options
+			ORDER BY $sql_sort";
+		$result = $db->sql_query_limit($sql, $config['search_block_size'], $start);
+
+		while ($row = $db->sql_fetchrow($result))
+		{
+			$id_ary[] = $row[$field];
+		}
+		$db->sql_freeresult($result);
+
+		$id_ary = array_unique($id_ary);
+
+		if (!sizeof($id_ary))
+		{
+			return false;
+		}
+
+		// if the total result count is not cached yet, retrieve it from the db
+		if (!$result_count)
+		{
+			$result_count = sizeof ($id_ary);
+
+			if (!$result_count)
+			{
+				return false;
+			}
+		}
+
+		// store the ids, from start on then delete anything that isn't on the current page because we only need ids for one page
+		$this->save_ids($search_key, implode(' ', $this->split_words), $author_ary, $result_count, $id_ary, $start, $sort_dir);
+		$id_ary = array_slice($id_ary, 0, (int) $per_page);
+
+		return $result_count;
+	}
+
+	/**
+	* Performs a search on an author's posts without caring about message contents. Depends on display specific params
+	*
+	* @param	string		$type				contains either posts or topics depending on what should be searched for
+	* @param	boolean		$firstpost_only		if true, only topic starting posts will be considered
+	* @param	array		$sort_by_sql		contains SQL code for the ORDER BY part of a query
+	* @param	string		$sort_key			is the key of $sort_by_sql for the selected sorting
+	* @param	string		$sort_dir			is either a or d representing ASC and DESC
+	* @param	string		$sort_days			specifies the maximum amount of days a post may be old
+	* @param	array		$ex_fid_ary			specifies an array of forum ids which should not be searched
+	* @param	array		$m_approve_fid_ary	specifies an array of forum ids in which the searcher is allowed to view unapproved posts
+	* @param	int			$topic_id			is set to 0 or a topic id, if it is not 0 then only posts in this topic should be searched
+	* @param	array		$author_ary			an array of author ids
+	* @param	string		$author_name		specifies the author match, when ANONYMOUS is also a search-match
+	* @param	array		&$id_ary			passed by reference, to be filled with ids for the page specified by $start and $per_page, should be ordered
+	* @param	int			$start				indicates the first index of the page
+	* @param	int			$per_page			number of ids each page is supposed to contain
+	* @return	boolean|int						total number of results
+	*
+	* @access	public
+	*/
+	function author_search($type, $firstpost_only, $sort_by_sql, $sort_key, $sort_dir, $sort_days, $ex_fid_ary, $m_approve_fid_ary, $topic_id, $author_ary, $author_name, &$id_ary, $start, $per_page)
+	{
+		global $config, $db;
+
+		// No author? No posts.
+		if (!sizeof($author_ary))
+		{
+			return 0;
+		}
+
+		// generate a search_key from all the options to identify the results
+		$search_key = md5(implode('#', array(
+			'',
+			$type,
+			($firstpost_only) ? 'firstpost' : '',
+			'',
+			'',
+			$sort_days,
+			$sort_key,
+			$topic_id,
+			implode(',', $ex_fid_ary),
+			implode(',', $m_approve_fid_ary),
+			implode(',', $author_ary),
+			$author_name,
+		)));
+
+		// try reading the results from cache
+		$result_count = 0;
+		if ($this->obtain_ids($search_key, $result_count, $id_ary, $start, $per_page, $sort_dir) == SEARCH_RESULT_IN_CACHE)
+		{
+			return $result_count;
+		}
+
+		$id_ary = array();
+
+		// Create some display specific sql strings
+		if ($author_name)
+		{
+			// first one matches post of registered users, second one guests and deleted users
+			$sql_author = '(' . $db->sql_in_set('p.poster_id', array_diff($author_ary, array(ANONYMOUS)), false, true) . ' OR p.post_username ' . $author_name . ')';
+		}
+		else
+		{
+			$sql_author = $db->sql_in_set('p.poster_id', $author_ary);
+		}
+		$sql_fora		= (sizeof($ex_fid_ary)) ? ' AND ' . $db->sql_in_set('p.forum_id', $ex_fid_ary, true) : '';
+		$sql_topic_id	= ($topic_id) ? ' AND p.topic_id = ' . (int) $topic_id : '';
+		$sql_time		= ($sort_days) ? ' AND p.post_time >= ' . (time() - ($sort_days * 86400)) : '';
+		$sql_firstpost = ($firstpost_only) ? ' AND p.post_id = t.topic_first_post_id' : '';
+
+		// Build sql strings for sorting
+		$sql_sort = $sort_by_sql[$sort_key] . (($sort_dir == 'a') ? ' ASC' : ' DESC');
+		$sql_sort_table = $sql_sort_join = '';
+		switch ($sql_sort[0])
+		{
+			case 'u':
+				$sql_sort_table	= USERS_TABLE . ' u, ';
+				$sql_sort_join	= ($type == 'posts') ? ' AND u.user_id = p.poster_id ' : ' AND u.user_id = t.topic_poster ';
+			break;
+
+			case 't':
+				$sql_sort_table	= ($type == 'posts' && !$firstpost_only) ? TOPICS_TABLE . ' t, ' : '';
+				$sql_sort_join	= ($type == 'posts' && !$firstpost_only) ? ' AND t.topic_id = p.topic_id ' : '';
+			break;
+
+			case 'f':
+				$sql_sort_table	= FORUMS_TABLE . ' f, ';
+				$sql_sort_join	= ' AND f.forum_id = p.forum_id ';
+			break;
+		}
+
+		if (!sizeof($m_approve_fid_ary))
+		{
+			$m_approve_fid_sql = ' AND p.post_approved = 1';
+		}
+		else if ($m_approve_fid_ary == array(-1))
+		{
+			$m_approve_fid_sql = '';
+		}
+		else
+		{
+			$m_approve_fid_sql = ' AND (p.post_approved = 1 OR ' . $db->sql_in_set('p.forum_id', $m_approve_fid_ary, true) . ')';
+		}
+
+		// Build the query for really selecting the post_ids
+		if ($type == 'posts')
+		{
+			$sql = "SELECT p.post_id
+				FROM " . $sql_sort_table . POSTS_TABLE . ' p' . (($firstpost_only) ? ', ' . TOPICS_TABLE . ' t ' : ' ') . "
+				WHERE $sql_author
+					$sql_topic_id
+					$sql_firstpost
+					$m_approve_fid_sql
+					$sql_fora
+					$sql_sort_join
+					$sql_time
+				ORDER BY $sql_sort";
+			$field = 'post_id';
+		}
+		else
+		{
+			$sql = "SELECT t.topic_id
+				FROM " . $sql_sort_table . TOPICS_TABLE . ' t, ' . POSTS_TABLE . " p
+				WHERE $sql_author
+					$sql_topic_id
+					$sql_firstpost
+					$m_approve_fid_sql
+					$sql_fora
+					AND t.topic_id = p.topic_id
+					$sql_sort_join
+					$sql_time
+				GROUP BY t.topic_id, $sort_by_sql[$sort_key]
+				ORDER BY $sql_sort";
+			$field = 'topic_id';
+		}
+
+		// Only read one block of posts from the db and then cache it
+		$result = $db->sql_query_limit($sql, $config['search_block_size'], $start);
+
+		while ($row = $db->sql_fetchrow($result))
+		{
+			$id_ary[] = $row[$field];
+		}
+		$db->sql_freeresult($result);
+
+		// retrieve the total result count if needed
+		if (!$result_count)
+		{
+			$result_count = sizeof ($id_ary);
+
+			if (!$result_count)
+			{
+				return false;
+			}
+		}
+
+		if (sizeof($id_ary))
+		{
+			$this->save_ids($search_key, '', $author_ary, $result_count, $id_ary, $start, $sort_dir);
+			$id_ary = array_slice($id_ary, 0, $per_page);
+
+			return $result_count;
+		}
+		return false;
+	}
+
+	/**
+	* Destroys cached search results, that contained one of the new words in a post so the results won't be outdated.
+	*
+	* @param	string		$mode		contains the post mode: edit, post, reply, quote ...
+	* @param	int			$post_id	contains the post id of the post to index
+	* @param	string		$message	contains the post text of the post
+	* @param	string		$subject	contains the subject of the post to index
+	* @param	int			$poster_id	contains the user id of the poster
+	* @param	int			$forum_id	contains the forum id of parent forum of the post
+	*
+	* @access public
+	*/
+	function index($mode, $post_id, &$message, &$subject, $poster_id, $forum_id)
+	{
+		global $db;
+
+		// Split old and new post/subject to obtain array of words
+		$split_text = $this->split_message($message);
+		$split_title = ($subject) ? $this->split_message($subject) : array();
+
+		$words = array_unique(array_merge($split_text, $split_title));
+
+		unset($split_text);
+		unset($split_title);
+
+		// destroy cached search results containing any of the words removed or added
+		$this->destroy_cache($words, array($poster_id));
+
+		unset($words);
+	}
+
+	/**
+	* Destroy cached results, that might be outdated after deleting a post
+	*
+	* @access public
+	*/
+	function index_remove($post_ids, $author_ids, $forum_ids)
+	{
+		$this->destroy_cache(array(), $author_ids);
+	}
+
+	/**
+	* Destroy old cache entries
+	*
+	* @access public
+	*/
+	function tidy()
+	{
+		global $db, $config;
+
+		// destroy too old cached search results
+		$this->destroy_cache(array());
+
+		set_config('search_last_gc', time(), true);
+	}
+
+	/**
+	* Create fulltext index
+	*
+	* @return string|bool error string is returned incase of errors otherwise false
+	*
+	* @access public
+	*/
+	function create_index($acp_module, $u_action)
+	{
+		global $db, $config;
+
+		// Make sure we can actually use PostgreSQL with fulltext indexes
+		if ($error = $this->init())
+		{
+			return $error;
+		}
+
+		if (empty($this->stats))
+		{
+			$this->get_stats();
+		}
+
+		if (!isset($this->stats['post_subject']))
+		{
+			$db->sql_query("CREATE INDEX " . POSTS_TABLE . "_" . $config['fulltext_postgres_ts_name'] . "_post_subject ON " . POSTS_TABLE . " USING gin (to_tsvector ('" . $db->sql_escape($config['fulltext_postgres_ts_name']) . "', post_subject))");
+		}
+
+		if (!isset($this->stats['post_text']))
+		{
+			$db->sql_query("CREATE INDEX " . POSTS_TABLE . "_" . $config['fulltext_postgres_ts_name'] . "_post_text ON " . POSTS_TABLE . " USING gin (to_tsvector ('" . $db->sql_escape($config['fulltext_postgres_ts_name']) . "', post_text))");
+		}
+
+		$db->sql_query('TRUNCATE TABLE ' . SEARCH_RESULTS_TABLE);
+
+		return false;
+	}
+
+	/**
+	* Drop fulltext index
+	*
+	* @return string|bool error string is returned incase of errors otherwise false
+	*
+	* @access public
+	*/
+	function delete_index($acp_module, $u_action)
+	{
+		global $db;
+
+		// Make sure we can actually use PostgreSQL with fulltext indexes
+		if ($error = $this->init())
+		{
+			return $error;
+		}
+
+		if (empty($this->stats))
+		{
+			$this->get_stats();
+		}
+
+		if (isset($this->stats['post_subject']))
+		{
+			$db->sql_query('DROP INDEX ' . $this->stats['post_subject']['relname']);
+		}
+
+		if (isset($this->stats['post_text']))
+		{
+			$db->sql_query('DROP INDEX ' . $this->stats['post_text']['relname']);
+		}
+
+		$db->sql_query('TRUNCATE TABLE ' . SEARCH_RESULTS_TABLE);
+
+		return false;
+	}
+
+	/**
+	* Returns true if both FULLTEXT indexes exist
+	*
+	* @access public
+	*/
+	function index_created()
+	{
+		if (empty($this->stats))
+		{
+			$this->get_stats();
+		}
+
+		return (isset($this->stats['post_text']) && isset($this->stats['post_subject'])) ? true : false;
+	}
+
+	/**
+	* Returns an associative array containing information about the indexes
+	*
+	* @access public
+	*/
+	function index_stats()
+	{
+		global $user;
+
+		if (empty($this->stats))
+		{
+			$this->get_stats();
+		}
+
+		return array(
+			$user->lang['FULLTEXT_POSTGRES_TOTAL_POSTS']			=> ($this->index_created()) ? $this->stats['total_posts'] : 0,
+		);
+	}
+
+	/**
+	 * Computes the stats and store them in the $this->stats associative array
+	 *
+	 * @access private
+	 */
+	function get_stats()
+	{
+		global $db, $config;
+
+		if ($db->sql_layer != 'postgres')
+		{
+			$this->stats = array();
+			return;
+		}
+
+		$sql = "SELECT c2.relname, pg_catalog.pg_get_indexdef(i.indexrelid, 0, true) AS indexdef
+			  FROM pg_catalog.pg_class c1, pg_catalog.pg_index i, pg_catalog.pg_class c2
+			 WHERE c1.relname = '" . POSTS_TABLE . "'
+			   AND pg_catalog.pg_table_is_visible(c1.oid)
+			   AND c1.oid = i.indrelid
+			   AND i.indexrelid = c2.oid";
+		$result = $db->sql_query($sql);
+
+		while ($row = $db->sql_fetchrow($result))
+		{
+			// deal with older PostgreSQL versions which didn't use Index_type
+			if (strpos($row['indexdef'], 'to_tsvector') !== false)
+			{
+				if ($row['relname'] == POSTS_TABLE . '_' . $config['fulltext_postgres_ts_name'] . '_post_text' || $row['relname'] == POSTS_TABLE . '_post_text')
+				{
+					$this->stats['post_text'] = $row;
+				}
+				else if ($row['relname'] == POSTS_TABLE . '_' . $config['fulltext_postgres_ts_name'] . '_post_subject' || $row['relname'] == POSTS_TABLE . '_post_subject')
+				{
+					$this->stats['post_subject'] = $row;
+				}
+			}
+		}
+		$db->sql_freeresult($result);
+
+		$this->stats['total_posts'] = $config['num_posts'];
+	}
+
+	/**
+	* Display various options that can be configured for the backend from the acp
+	*
+	* @return associative array containing template and config variables
+	*
+	* @access public
+	*/
+	function acp()
+	{
+		global $user, $config, $db;
+
+		$tpl = '
+		<dl>
+			<dt><label>' . $user->lang['FULLTEXT_POSTGRES_VERSION_CHECK'] . '</label><br /><span>' . $user->lang['FULLTEXT_POSTGRES_VERSION_CHECK_EXPLAIN'] . '</span></dt>
+			<dd>' . (($this->tsearch_usable) ? $user->lang['YES'] : $user->lang['NO']) . ' (PostgreSQL ' . $this->version . ')</dd>
+		</dl>
+		<dl>
+			<dt><label>' . $user->lang['FULLTEXT_POSTGRES_TS_NAME'] . '</label><br /><span>' . $user->lang['FULLTEXT_POSTGRES_TS_NAME_EXPLAIN'] . '</span></dt>
+			<dd><select name="config[fulltext_postgres_ts_name]">';
+
+		if ($db->sql_layer == 'postgres' && $this->tsearch_usable)
+		{
+			$sql = 'SELECT cfgname AS ts_name
+				  FROM pg_ts_config';
+			$result = $db->sql_query($sql);
+
+			while ($row = $db->sql_fetchrow($result))
+			{
+				$tpl .= '<option value="' . $row['ts_name'] . '"' . ($row['ts_name'] === $config['fulltext_postgres_ts_name'] ? ' selected="selected"' : '') . '>' . $row['ts_name'] . '</option>';
+			}
+			$db->sql_freeresult($result);
+		}
+		else
+		{
+			$tpl .= '<option value="' . $config['fulltext_postgres_ts_name'] . '" selected="selected">' . $config['fulltext_postgres_ts_name'] . '</option>';
+		}
+
+		$tpl .= '</select></dd>
+		</dl>
+                <dl>
+                        <dt><label for="fulltext_postgres_min_word_len">' . $user->lang['FULLTEXT_POSTGRES_MIN_WORD_LEN'] . ':</label><br /><span>' . $user->lang['FULLTEXT_POSTGRES_MIN_WORD_LEN_EXPLAIN'] . '</span></dt>
+                        <dd><input id="fulltext_postgres_min_word_len" type="text" size="3" maxlength="3" name="config[fulltext_postgres_min_word_len]" value="' . (int) $config['fulltext_postgres_min_word_len'] . '" /></dd>
+                </dl>
+                <dl>
+                        <dt><label for="fulltext_postgres_max_word_len">' . $user->lang['FULLTEXT_POSTGRES_MAX_WORD_LEN'] . ':</label><br /><span>' . $user->lang['FULLTEXT_POSTGRES_MAX_WORD_LEN_EXPLAIN'] . '</span></dt>
+                        <dd><input id="fulltext_postgres_max_word_len" type="text" size="3" maxlength="3" name="config[fulltext_postgres_max_word_len]" value="' . (int) $config['fulltext_postgres_max_word_len'] . '" /></dd>
+                </dl>
+		';
+
+		// These are fields required in the config table
+		return array(
+			'tpl'		=> $tpl,
+			'config'	=> array('fulltext_postgres_ts_name' => 'string', 'fulltext_postgres_min_word_len' => 'integer:0:255', 'fulltext_postgres_max_word_len' => 'integer:0:255')
+		);
+	}
+}
diff --git a/phpBB/install/database_update.php b/phpBB/install/database_update.php
index b7490aca0b..594397c815 100644
--- a/phpBB/install/database_update.php
+++ b/phpBB/install/database_update.php
@@ -2245,6 +2245,21 @@ function change_database_data(&$no_updates, $version)
 				set_config('search_type', 'phpbb_search_' . $config['search_type']);
 			}
 
+			if (!isset($config['fulltext_postgres_ts_name']))
+			{
+				set_config('fulltext_postgres_ts_name', 'simple');
+			}
+
+			if (!isset($config['fulltext_postgres_min_word_len']))
+			{
+				set_config('fulltext_postgres_min_word_len', 4);
+			}
+
+			if (!isset($config['fulltext_postgres_max_word_len']))
+			{
+				set_config('fulltext_postgres_max_word_len', 254);
+			}
+
 			if (!isset($config['load_jquery_cdn']))
 			{
 				set_config('load_jquery_cdn', 0);
diff --git a/phpBB/install/schemas/schema_data.sql b/phpBB/install/schemas/schema_data.sql
index 5489fd4e3d..deefdafbc4 100644
--- a/phpBB/install/schemas/schema_data.sql
+++ b/phpBB/install/schemas/schema_data.sql
@@ -125,6 +125,9 @@ INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_native_co
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_native_load_upd', '1');
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_native_max_chars', '14');
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_native_min_chars', '3');
+INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_postgres_max_word_len', '254');
+INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_postgres_min_word_len', '4');
+INSERT INTO phpbb_config (config_name, config_value) VALUES ('fulltext_postgres_ts_name', 'simple');
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('gzip_compress', '0');
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('hot_threshold', '25');
 INSERT INTO phpbb_config (config_name, config_value) VALUES ('icons_path', 'images/icons');
diff --git a/phpBB/language/en/acp/search.php b/phpBB/language/en/acp/search.php
index 2f46856557..cd319c66a9 100644
--- a/phpBB/language/en/acp/search.php
+++ b/phpBB/language/en/acp/search.php
@@ -57,6 +57,18 @@ $lang = array_merge($lang, array(
 	'FULLTEXT_MYSQL_MIN_SEARCH_CHARS_EXPLAIN'	=> 'Words with at least this many characters will be indexed for searching. You or your host can only change this setting by changing the mysql configuration.',
 	'FULLTEXT_MYSQL_MAX_SEARCH_CHARS_EXPLAIN'	=> 'Words with no more than this many characters will be indexed for searching. You or your host can only change this setting by changing the mysql configuration.',
 
+	'FULLTEXT_POSTGRES_INCOMPATIBLE_DATABASE'	=> 'The PostgreSQL fulltext backend can only be used with PostgreSQL.',
+	'FULLTEXT_POSTGRES_TS_NOT_USABLE'	=> 'The PostgreSQL fulltext backend can only be used with PostgreSQL 8.3 and above.',
+	'FULLTEXT_POSTGRES_TOTAL_POSTS'			=> 'Total number of indexed posts',
+	'FULLTEXT_POSTGRES_VERSION_CHECK'		=> 'PostgreSQL version',
+	'FULLTEXT_POSTGRES_TS_NAME'				=> 'Text search Configuration Profile:',
+	'FULLTEXT_POSTGRES_MIN_WORD_LEN'			=> 'Minimum word length for keywords',
+	'FULLTEXT_POSTGRES_MAX_WORD_LEN'			=> 'Maximum word length for keywords',
+	'FULLTEXT_POSTGRES_VERSION_CHECK_EXPLAIN'		=> 'This search backend requires PostgreSQL version 8.3 and above.',
+	'FULLTEXT_POSTGRES_TS_NAME_EXPLAIN'				=> 'The Text search configuration profile used to determine the parser and dictionary.',
+	'FULLTEXT_POSTGRES_MIN_WORD_LEN_EXPLAIN'			=> 'Words with at least this many characters will be included in the query to the database.',
+	'FULLTEXT_POSTGRES_MAX_WORD_LEN_EXPLAIN'			=> 'Words with no more than this many characters will be included in the query to the database.',
+
 	'GENERAL_SEARCH_SETTINGS'				=> 'General search settings',
 	'GO_TO_SEARCH_INDEX'					=> 'Go to search index page',
 
diff --git a/phpBB/language/en/search.php b/phpBB/language/en/search.php
index 5b6fdce0e7..2fbb5122ad 100644
--- a/phpBB/language/en/search.php
+++ b/phpBB/language/en/search.php
@@ -72,6 +72,7 @@ $lang = array_merge($lang, array(
 	'WORDS_IN_NO_POST'		=> 'No posts were found because the words <strong>%s</strong> are not contained in any post.',
 
 	'POST_CHARACTERS'		=> 'characters of posts',
+	'PHRASE_SEARCH_DISABLED'	=> 'Searching by exact phrase is not supported on this board.',
 
 	'RECENT_SEARCHES'		=> 'Recent searches',
 	'RESULT_DAYS'			=> 'Limit results to previous',
diff --git a/phpBB/search.php b/phpBB/search.php
index 2b463aec9c..1123d7c5b9 100644
--- a/phpBB/search.php
+++ b/phpBB/search.php
@@ -596,6 +596,13 @@ if ($keywords || $author || $author_id || $search_id || $submit)
 	$u_search .= ($search_fields != 'all') ? '&amp;sf=' . $search_fields : '';
 	$u_search .= ($return_chars != 300) ? '&amp;ch=' . $return_chars : '';
 
+	// Check if search backend supports phrase search or not
+	$phrase_search_disabled = '';
+	if (strpos(html_entity_decode($keywords), '"') !== false && method_exists($search, 'supports_phrase_search'))
+	{
+		$phrase_search_disabled = $search->supports_phrase_search() ? false : true;
+	}
+
 	$template->assign_vars(array(
 		'SEARCH_TITLE'		=> $l_search_title,
 		'SEARCH_MATCHES'	=> $l_search_matches,
@@ -603,6 +610,9 @@ if ($keywords || $author || $author_id || $search_id || $submit)
 		'IGNORED_WORDS'		=> (sizeof($search->common_words)) ? implode(' ', $search->common_words) : '',
 		'PAGINATION'		=> generate_pagination($u_search, $total_match_count, $per_page, $start),
 		'PAGE_NUMBER'		=> on_page($total_match_count, $per_page, $start),
+
+		'PHRASE_SEARCH_DISABLED'		=> $phrase_search_disabled,
+
 		'TOTAL_MATCHES'		=> $total_match_count,
 		'SEARCH_IN_RESULTS'	=> ($search_id) ? false : true,
 
diff --git a/phpBB/styles/prosilver/template/search_results.html b/phpBB/styles/prosilver/template/search_results.html
index 5d75bd3d56..56712db75a 100644
--- a/phpBB/styles/prosilver/template/search_results.html
+++ b/phpBB/styles/prosilver/template/search_results.html
@@ -2,6 +2,7 @@
 
 <h2><!-- IF SEARCH_TITLE -->{SEARCH_TITLE}<!-- ELSE -->{SEARCH_MATCHES}<!-- ENDIF --><!-- IF SEARCH_WORDS -->: <a href="{U_SEARCH_WORDS}">{SEARCH_WORDS}</a><!-- ENDIF --></h2>
 <!-- IF IGNORED_WORDS --> <p>{L_IGNORED_TERMS}: <strong>{IGNORED_WORDS}</strong></p><!-- ENDIF -->
+<!-- IF PHRASE_SEARCH_DISABLED --> <p><strong>{L_PHRASE_SEARCH_DISABLED}</strong></p><!-- ENDIF -->
 
 <!-- IF SEARCH_TOPIC -->
 	<p><a class="{S_CONTENT_FLOW_BEGIN}" href="{U_SEARCH_TOPIC}">{L_RETURN_TO}: {SEARCH_TOPIC}</a></p>
author	Oleg Pudeyev <oleg@bsdpower.com>	2012-07-07 16:19:53 -0400
committer	Oleg Pudeyev <oleg@bsdpower.com>	2012-07-07 16:19:53 -0400
commit	1526886e3ec2b475bb7f578767f6430dbfa58097 (patch)
tree	e21dfc5b9eac9fd26b075bee1851efc47272d5e4 /phpBB
parent	d9fd0cce0a5b323bff785af401e90370a3d73972 (diff)
parent	70a0caee28559bd1bd52b4ecf9d7d33acd732fe2 (diff)
download	forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.gz forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.bz2 forums-1526886e3ec2b475bb7f578767f6430dbfa58097.tar.xz forums-1526886e3ec2b475bb7f578767f6430dbfa58097.zip