From e7e451b7d9339883db29bad79df5256f9b410652 Mon Sep 17 00:00:00 2001
From: Aki Tuomi <aki.tuomi@dovecot.fi>
Date: Thu, 19 Oct 2017 11:55:01 +0300
Subject: [PATCH] test-unichar: Update test to conform RFC3629

---
 src/lib/test-unichar.c | 75 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 75 insertions(+)

diff --git a/src/lib/test-unichar.c b/src/lib/test-unichar.c
index 21826ebf97..d16040ebbb 100644
--- a/src/lib/test-unichar.c
+++ b/src/lib/test-unichar.c
@@ -36,6 +36,77 @@ static void test_unichar_uni_utf8_partial_strlen_n(void)
 	test_end();
 }
 
+static void test_unichar_valid_unicode(void)
+{
+	struct {
+		const char *input;
+		bool valid;
+		unichar_t expected;
+	} test_cases[] = {
+		{ "a", TRUE, 'a' },
+		{ "\xc3\xb1", TRUE, 0x00F1 }, /* U+00F1 */
+		{ "\xc3\x28", FALSE, 0x0 }, /* has invalid 2nd octet */
+		{ "\xa0\xa1", FALSE, 0x0 }, /* invalid sequence identifier */
+		{ "\xed\xb2\x80", FALSE, 0x0 }, /* UTF-8B */
+		{ "\xed\xa0\x80", FALSE, 0x0 }, /* surrogate halves, U+D800 .. */
+		{ "\xed\xa0\x80", FALSE, 0x0 },
+		{ "\xed\xa1\x80", FALSE, 0x0 },
+		{ "\xed\xa2\x80", FALSE, 0x0 },
+		{ "\xed\xa3\x80", FALSE, 0x0 },
+		{ "\xed\xa4\x80", FALSE, 0x0 },
+		{ "\xed\xa5\x80", FALSE, 0x0 },
+		{ "\xed\xa6\x80", FALSE, 0x0 },
+		{ "\xed\xa7\x80", FALSE, 0x0 },
+		{ "\xed\xa8\x80", FALSE, 0x0 },
+		{ "\xed\xa9\x80", FALSE, 0x0 },
+		{ "\xed\xaa\x80", FALSE, 0x0 },
+		{ "\xed\xab\x80", FALSE, 0x0 },
+		{ "\xed\xac\x80", FALSE, 0x0 },
+		{ "\xed\xad\x80", FALSE, 0x0 },
+		{ "\xed\xaf\x80", FALSE, 0x0 },
+		{ "\xed\xb0\x80", FALSE, 0x0 },
+		{ "\xed\xb1\x80", FALSE, 0x0 },
+		{ "\xed\xb2\x80", FALSE, 0x0 },
+		{ "\xed\xb3\x80", FALSE, 0x0 },
+		{ "\xed\xb4\x80", FALSE, 0x0 },
+		{ "\xed\xb5\x80", FALSE, 0x0 },
+		{ "\xed\xb6\x80", FALSE, 0x0 },
+		{ "\xed\xb7\x80", FALSE, 0x0 },
+		{ "\xed\xb8\x80", FALSE, 0x0 },
+		{ "\xed\xb9\x80", FALSE, 0x0 },
+		{ "\xed\xba\x80", FALSE, 0x0 },
+		{ "\xed\xbb\x80", FALSE, 0x0 },
+		{ "\xed\xbc\x80", FALSE, 0x0 },
+		{ "\xed\xbd\x80", FALSE, 0x0 },
+		{ "\xed\xbf\x80", FALSE, 0x0 }, /* .. U+DFFF */
+		{ "\xe2\x82\xa1", TRUE, 0x20A1 },  /* U+20A1 */
+		{ "\xe2\x28\xa1", FALSE, 0x0 }, /* invalid 2nd octet */
+		{ "\xe2\x82\x28", FALSE, 0x0 }, /* invalid 3rd octet */
+		{ "\xf0\x90\x8c\xbc", TRUE, 0x1033C },  /* U+1033C */
+		{ "\xf0\x28\x8c\xbc", FALSE, 0x0 }, /*invalid 2nd octet*/
+		{ "\xf0\x90\x28\xbc", FALSE, 0x0 }, /* invalid 3rd octet */
+		{ "\xf0\x28\x8c\x28", FALSE, 0x0 }, /* invalid 4th octet */
+		{ "\xf4\x80\x80\x80", TRUE, 0x100000 }, /* U+100000, supplementary plane start */
+		{ "\xf4\x8f\xbf\xbf", TRUE, 0x10FFFF }, /* U+10FFFF, maximum value */
+		{ "\xf8\xa1\xa1\xa1\xa1", FALSE, 0x0 }, /* invalid unicode */
+		{ "\xfc\xa1\xa1\xa1\xa1\xa1", FALSE, 0x0 }, /* invalid unicode */
+	};
+
+	test_begin("unichar valid unicode");
+
+	for(size_t i = 0; i < N_ELEMENTS(test_cases); i++) {
+		unichar_t chr;
+		if (test_cases[i].valid) {
+			test_assert_idx(uni_utf8_get_char(test_cases[i].input, &chr) > 0, i);
+			test_assert_idx(test_cases[i].expected == chr, i);
+		} else {
+			test_assert_idx(uni_utf8_get_char(test_cases[i].input, &chr) < 1, i);
+		}
+	}
+
+	test_end();
+}
+
 void test_unichar(void)
 {
 	static const char overlong_utf8[] = "\xf8\x80\x95\x81\xa1";
@@ -47,6 +118,9 @@ void test_unichar(void)
 
 	test_begin("unichars encode/decode");
 	for (chr = 0; chr <= 0x10ffff; chr++) {
+		/* skip surrogates */
+		if ((chr & 0xfff800) == 0xd800)
+			continue;
 		/* The bottom 6 bits should be irrelevant to code coverage,
 		   only test 000000, 111111, and something in between. */
 		if ((chr & 63) == 1)
@@ -92,4 +166,5 @@ void test_unichar(void)
 
 	test_unichar_uni_utf8_strlen();
 	test_unichar_uni_utf8_partial_strlen_n();
+	test_unichar_valid_unicode();
 }
-- 
2.47.3