From 95f4db0a8bb5bbf742a0195fbbe17d304fb99507 Mon Sep 17 00:00:00 2001
From: Carl Hetherington <cth@carlh.net>
Date: Thu, 17 Jan 2019 22:33:55 +0000
Subject: [PATCH] Read UTF8 string lengths correctly when checking closed
 captions (part of #1446).

---
 src/lib/hints.cc  |  2 +-
 src/lib/util.cc   | 19 +++++++++++++++++++
 src/lib/util.h    |  1 +
 test/util_test.cc |  8 +++++++-
 4 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/src/lib/hints.cc b/src/lib/hints.cc
index d961b1a30..6238aa991 100644
--- a/src/lib/hints.cc
+++ b/src/lib/hints.cc
@@ -311,7 +311,7 @@ Hints::text (PlayerText text, TextType type, DCPTimePeriod period)
 
 	int lines = text.string.size();
 	BOOST_FOREACH (StringText i, text.string) {
-		if (i.text().length() > CLOSED_CAPTION_LENGTH) {
+		if (utf8_strlen(i.text()) > CLOSED_CAPTION_LENGTH) {
 			++lines;
 			if (!_long_ccap) {
 				_long_ccap = true;
diff --git a/src/lib/util.cc b/src/lib/util.cc
index 7472047e8..cd2d2e753 100644
--- a/src/lib/util.cc
+++ b/src/lib/util.cc
@@ -815,6 +815,25 @@ checked_fread (void* ptr, size_t size, FILE* stream, boost::filesystem::path pat
 	}
 }
 
+size_t
+utf8_strlen (string s)
+{
+	size_t const len = s.length ();
+	int N = 0;
+	for (size_t i = 0; i < len; ++i) {
+		unsigned char c = s[i];
+		if ((c & 0xe0) == 0xc0) {
+			++i;
+		} else if ((c & 0xf0) == 0xe0) {
+			i += 2;
+		} else if ((c & 0xf8) == 0xf0) {
+			i += 3;
+		}
+		++N;
+	}
+	return N;
+}
+
 #ifdef DCPOMATIC_VARIANT_SWAROOP
 
 /* Make up a key from the machine UUID */
diff --git a/src/lib/util.h b/src/lib/util.h
index bdf2480fc..7c0eb4b6d 100644
--- a/src/lib/util.h
+++ b/src/lib/util.h
@@ -100,6 +100,7 @@ extern boost::shared_ptr<AudioBuffers> remap (boost::shared_ptr<const AudioBuffe
 extern Eyes increment_eyes (Eyes e);
 extern void checked_fread (void* ptr, size_t size, FILE* stream, boost::filesystem::path path);
 extern void checked_fwrite (void const * ptr, size_t size, FILE* stream, boost::filesystem::path path);
+extern size_t utf8_strlen (std::string s);
 #ifdef DCPOMATIC_VARIANT_SWAROOP
 extern boost::shared_ptr<dcp::CertificateChain> read_swaroop_chain (boost::filesystem::path path);
 extern void write_swaroop_chain (boost::shared_ptr<const dcp::CertificateChain> chain, boost::filesystem::path output);
diff --git a/test/util_test.cc b/test/util_test.cc
index ba33b42c5..d8cb61fdc 100644
--- a/test/util_test.cc
+++ b/test/util_test.cc
@@ -92,6 +92,13 @@ BOOST_AUTO_TEST_CASE (tidy_for_filename_test)
 	BOOST_CHECK_EQUAL (tidy_for_filename ("abcdefghï"), "abcdefghï");
 }
 
+BOOST_AUTO_TEST_CASE (utf8_strlen_test)
+{
+	BOOST_CHECK_EQUAL (utf8_strlen("hello world"), 11);
+	BOOST_CHECK_EQUAL (utf8_strlen("hëllo world"), 11);
+	BOOST_CHECK_EQUAL (utf8_strlen("hëłlo wørld"), 11);
+}
+
 #ifdef DCPOMATIC_VARIANT_SWAROOP
 BOOST_AUTO_TEST_CASE (swaroop_chain_test)
 {
@@ -111,5 +118,4 @@ BOOST_AUTO_TEST_CASE (swaroop_chain_test)
 
 	BOOST_CHECK (cc->root_to_leaf() == back->root_to_leaf());
 }
-
 #endif
-- 
2.30.2