From b812485d8ae4a2920a73514df5ce1de8d96fe999 Mon Sep 17 00:00:00 2001
From: Alex Cabal <alex@alexcabal.com>
Date: Fri, 18 Jan 2019 20:51:22 -0600
Subject: [PATCH] Remove flaky iconv and replace with PHP-intl

---
 README.md         |  2 +-
 lib/Ebook.php     |  5 ++---
 lib/Formatter.php | 17 ++++++++++++-----
 3 files changed, 15 insertions(+), 9 deletions(-)

diff --git a/README.md b/README.md
index 573d9e39..fd1bd0f4 100644
--- a/README.md
+++ b/README.md
@@ -12,7 +12,7 @@ You’ll also need to ensure the following:
 
 - Your PHP installation must be configured to have `/standardebooks.org/lib/` in its include path.
 
-- PHP-APCu must be installed. On Ubuntu this can be done with `sudo apt install php-apcu`.
+- [PHP-APCu](http://php.net/manual/en/book.apcu.php) and [PHP-intl](http://php.net/manual/en/book.intl.php) must be installed. On Ubuntu this can be done with `sudo apt install php-apcu php-intl`.
 
 - The URL `^/ebooks/([^\.]+?)/?$` must redirect to `/standardebooks.org/ebooks/ebook.php?url-path=$1`
 
diff --git a/lib/Ebook.php b/lib/Ebook.php
index 621d523d..ed15ca4e 100644
--- a/lib/Ebook.php
+++ b/lib/Ebook.php
@@ -342,9 +342,8 @@ class Ebook{
 		}
 
 		// Remove diacritics and non-alphanumeric characters
-		$searchString = str_replace(['‘,', '’'], '', $searchString);
-		$searchString = trim(preg_replace('|[^a-zA-Z0-9 ]|ius', ' ', iconv('UTF-8', 'ASCII//TRANSLIT', $searchString) ?: '') ?? '');
-		$query = trim(preg_replace('|[^a-zA-Z0-9 ]|ius', ' ', iconv('UTF-8', 'ASCII//TRANSLIT', $query) ?: '') ?? '');
+		$searchString = trim(preg_replace('|[^a-zA-Z0-9 ]|ius', ' ', Formatter::RemoveDiacritics($searchString)) ?? '');
+		$query = trim(preg_replace('|[^a-zA-Z0-9 ]|ius', ' ', Formatter::RemoveDiacritics($query)) ?? '');
 
 		if($query == ''){
 			return false;
diff --git a/lib/Formatter.php b/lib/Formatter.php
index 870a37a5..6c51c417 100644
--- a/lib/Formatter.php
+++ b/lib/Formatter.php
@@ -1,12 +1,19 @@
 <?
 class Formatter{
-	public static function MakeUrlSafe(string $text): string{
-		// Remove apostrophes
-		// We have to do this first so iconv doesn't choke
-		$text = str_replace(['\'', '‘,', '’'], '', $text);
+	public static function RemoveDiacritics(string $text): string{
+		if(!isset($GLOBALS['transliterator'])){
+			$GLOBALS['transliterator'] = Transliterator::createFromRules(':: Any-Latin; :: Latin-ASCII; :: NFD; :: [:Nonspacing Mark:] Remove; :: Lower(); :: NFC;', Transliterator::FORWARD);
+		}
 
+		return $GLOBALS['transliterator']->transliterate($text);
+	}
+
+	public static function MakeUrlSafe(string $text): string{
 		// Remove accent characters
-		$text = iconv('UTF-8', 'ASCII//TRANSLIT', $text) ?: '';
+		$text = self::RemoveDiacritics($text);
+
+		// Remove apostrophes
+		$text = str_replace('\'', '', $text);
 
 		// Trim and convert to lowercase
 		$text = mb_strtolower(trim($text));