Add a test to ensure source files do not have Byte Order Mark

.gitignore

@@ -110,6 +110,7 @@ tests/*.check
 !tests/ssml.test
 !tests/translate.test
 !tests/non-executable-files-with-executable-bit.test
+!tests/bom.test
 
 espeak-ng.pc
 

Makefile.am

@@ -256,7 +256,8 @@ check:	tests/encoding.check \
 	tests/language-pronunciation.check \
 	tests/language-numbers-cardinal.check \
 	tests/language-numbers-ordinal.check \
-	tests/non-executable-files-with-executable-bit.check
+	tests/non-executable-files-with-executable-bit.check \
+	tests/bom.check
 
 ##### fuzzer:
 

tests/bom.test

@@ -0,0 +1,22 @@
+#!/bin/bash
+# This script checks for Byte Order Marks in source files
+# which in some places may break compilation/runtime of eSpeak NG
+# Look for more at: https://en.wikipedia.org/wiki/Byte_order_mark
+
+echo -n "testing for Byte Order Marks in source files ... "
+
+# TODO checking for other encodings (e.g. UTF-16, UTF-32, UTF-7) needs more elaborate search,
+# otherwise it shows many false positives
+
+egrep -rlI $'\xEF\xBB\xBF' . | tee tests/bom.check >/dev/null # UTF-8
+
+if [ -s tests/bom.check ] ; then
+	echo "found:"
+	cat tests/bom.check
+	false
+else
+	rm tests/bom.check
+	echo "none found"
+	true
+fi
+