From 23ee354023272efdca261ee7f52025fe69e15b79 Mon Sep 17 00:00:00 2001
From: Jim Mason <jmason@ibinx.com>
Date: Sat, 7 Jan 2023 11:08:52 +0000
Subject: [PATCH] fixed full text search to split on non-word character, per
 default utf8mb4 collation semantics

partly satisfies #379
---
 INSTALLATION.md         |  2 +-
 engine/impl/Library.php | 20 ++++++++++++--------
 js/playlists.track.js   |  6 +++---
 3 files changed, 16 insertions(+), 12 deletions(-)

diff --git a/INSTALLATION.md b/INSTALLATION.md
index 5938546f..ba7b6f22 100644
--- a/INSTALLATION.md
+++ b/INSTALLATION.md
@@ -91,7 +91,7 @@ particulars, then read on.
    may locate and/or name the file differently.  In the `[mysqld]`
    section, add the lines:
 
-        ft_min_word_len = 3
+        ft_min_word_len = 1
         ft_stopword_file = ''
 
    If `ft_min_word_len` or `ft_stopword_file` is already in the file,
diff --git a/engine/impl/Library.php b/engine/impl/Library.php
index 66ca90d5..6d5dc014 100644
--- a/engine/impl/Library.php
+++ b/engine/impl/Library.php
@@ -3,7 +3,7 @@
  * Zookeeper Online
  *
  * @author Jim Mason <jmason@ibinx.com>
- * @copyright Copyright (C) 1997-2022 Jim Mason <jmason@ibinx.com>
+ * @copyright Copyright (C) 1997-2023 Jim Mason <jmason@ibinx.com>
  * @link https://zookeeper.ibinx.com/
  * @license GPL-3.0
  *
@@ -112,6 +112,13 @@ class LibraryImpl extends DBO implements ILibrary {
         "\"\u{201c}\u{201d}",                // double quotation mark
     ];
 
+    /*
+     * words to exclude from a full-text search
+     */
+    private static $ftExclude = [
+        "a", "an", "and", "or", "the"
+    ];
+
     private static function orderBy($sortBy) {
         if(substr($sortBy, -1) == "-") {
             $sortBy = substr($sortBy, 0, -1);
@@ -839,11 +846,6 @@ public function listLabels($op, $key, $limit) {
         return $result;
     }
 
-    public static function ftfilter($elt) {
-       return strlen($elt) > 1 &&
-          $elt != "an" && $elt != "and" && $elt != "or" && $elt != "the";
-    }
-
     public function searchFullText($type, $key, $size, $offset) {
         $retVal = array();
         $loggedIn = Engine::session()->isAuth("u");
@@ -856,8 +858,10 @@ public function searchFullText($type, $key, $size, $offset) {
         if(substr($key, 0, 2) == "\\\"") {
              $search = $key;
         } else {
-             $words = array_filter(explode(" ", $key), array(__CLASS__, "ftfilter"));
-             $search = "+".implode(" +",$words);
+             $words = array_filter(preg_split('/\W+/u', $key, 0, PREG_SPLIT_NO_EMPTY), function($word) {
+                 return !in_array($word, self::$ftExclude);
+             });
+             $search = "+" . implode(" +", $words);
         }
     
         // JM 2010-09-26 remove semicolons to thwart injection attacks
diff --git a/js/playlists.track.js b/js/playlists.track.js
index 4dcf5c3f..dce3bbc9 100644
--- a/js/playlists.track.js
+++ b/js/playlists.track.js
@@ -2,7 +2,7 @@
 // Zookeeper Online
 //
 // @author Jim Mason <jmason@ibinx.com>
-// @copyright Copyright (C) 1997-2022 Jim Mason <jmason@ibinx.com>
+// @copyright Copyright (C) 1997-2023 Jim Mason <jmason@ibinx.com>
 // @link https://zookeeper.ibinx.com/
 // @license GPL-3.0
 //
@@ -20,7 +20,7 @@
 // http://www.gnu.org/licenses/
 //
 
-/*! Zookeeper Online (C) 1997-2022 Jim Mason <jmason@ibinx.com> | @source: https://zookeeper.ibinx.com/ | @license: magnet:?xt=urn:btih:1f739d935676111cfff4b4693e3816e664797050&dn=gpl-3.0.txt GPL-v3.0 */
+/*! Zookeeper Online (C) 1997-2023 Jim Mason <jmason@ibinx.com> | @source: https://zookeeper.ibinx.com/ | @license: magnet:?xt=urn:btih:1f739d935676111cfff4b4693e3816e664797050&dn=gpl-3.0.txt GPL-v3.0 */
 
 $().ready(function(){
     const NME_ENTRY='nme-entry';
@@ -718,7 +718,7 @@ $().ready(function(){
     }).on('click', function() {
         $(this).autocomplete('search', this.value);
     }).autocomplete({
-        minLength: 3,
+        minLength: 1,
         source: function(rq, rs) {
             var artist = rq.term;