From ccad5c83600646cfcbb7833a31e923a7d033015e Mon Sep 17 00:00:00 2001
From: Matthew de Marillac <mattmarillac@hotmail.com>
Date: Sat, 25 Nov 2017 13:07:24 +0100
Subject: [PATCH] Remove BOM from UTF files.

Implementation similar to suggestion from, and thus closes #83
---
 parsecsv.lib.php | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/parsecsv.lib.php b/parsecsv.lib.php
index 13725ee..8001442 100644
--- a/parsecsv.lib.php
+++ b/parsecsv.lib.php
@@ -838,6 +838,20 @@ class parseCSV {
                 $data = ltrim($strip[1]);
             }
 
+            if (strpos($data, "\xef\xbb\xbf") === 0) {
+                // strip off BOM (UTF-8)
+                $data = substr($data, 3);
+            }
+            else if (strpos($data, "\xff\xfe") === 0) {
+                // strip off BOM (UTF-16 little endian)
+                $data = substr($data, 2);
+            }
+
+            else if (strpos($data, "\xfe\xff") === 0) {
+                // strip off BOM (UTF-16 big endian)
+                $data = substr($data, 2);
+            }
+
             if ($this->convert_encoding) {
                 $data = iconv($this->input_encoding, $this->output_encoding, $data);
             }