From 112a83c346a537f1a5eac6fc17ee2ce3143d625b Mon Sep 17 00:00:00 2001
From: Karen Arutyunov <karen@codesynthesis.com>
Date: Thu, 18 Jun 2020 16:40:00 +0300
Subject: Fix lexer to fail on invalid UTF-8 sequences

---
 libbuild2/lexer+utf8.test.testscript | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
 create mode 100644 libbuild2/lexer+utf8.test.testscript

(limited to 'libbuild2/lexer+utf8.test.testscript')

diff --git a/libbuild2/lexer+utf8.test.testscript b/libbuild2/lexer+utf8.test.testscript
new file mode 100644
index 0000000..42c62ea
--- /dev/null
+++ b/libbuild2/lexer+utf8.test.testscript
@@ -0,0 +1,28 @@
+# file      : libbuild2/lexer+utf8.test.testscript
+# license   : MIT; see accompanying LICENSE file
+
+: valid
+:
+$* <<EOI >>EOO
+  Sommerzeit
+  Mitteleuropäische
+  EOI
+  'Sommerzeit'
+  <newline>
+  'Mitteleuropäische'
+  <newline>
+  EOO
+
+: invalid
+:
+: Here we spoil the UTF-8 sequence 'ä' by dropping its second byte.
+:
+cat <<EOI | sed -e 's/(rop.).(isc)/\1\2/' | $* >>EOO 2>>EOE != 0
+  Sommerzeit
+  Mitteleuropäische
+  EOI
+  'Sommerzeit'
+  <newline>
+  EOO
+  <stdin>:2:12: error: invalid UTF-8 sequence second byte (0x69 'i')
+  EOE
-- 
cgit v1.1