CMakeLists.txt: omit end{if,foreach} expressions

Their usefulness was almost negative.

CMakeLists.txt

@@ -4,9 +4,9 @@ cmake_minimum_required (VERSION 2.8.5)
 # Moar warnings
 if ("${CMAKE_C_COMPILER_ID}" MATCHES "GNU" OR CMAKE_COMPILER_IS_GNUCC)
 	# -Wunused-function is pretty annoying here, as everything is static
-	set (wdisabled "-Wno-unused-function -Wno-implicit-fallthrough")
+	set (wdisabled "-Wno-unused-function")
 	set (CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -std=c99 -Wall -Wextra ${wdisabled}")
-endif ("${CMAKE_C_COMPILER_ID}" MATCHES "GNU" OR CMAKE_COMPILER_IS_GNUCC)
+endif ()
 
 # Dependencies
 set (CMAKE_MODULE_PATH ${PROJECT_SOURCE_DIR}/cmake)
@@ -16,11 +16,9 @@ find_package (PkgConfig REQUIRED)
 pkg_check_modules (libssl REQUIRED libssl libcrypto)
 
 if ("${CMAKE_SYSTEM_NAME}" MATCHES "BSD")
-	include_directories (/usr/local/include)
-	link_directories (/usr/local/lib)
 	# Our POSIX version macros make these undefined
 	add_definitions (-D__BSD_VISIBLE=1 -D_BSD_SOURCE=1)
-endif ("${CMAKE_SYSTEM_NAME}" MATCHES "BSD")
+endif ()
 
 set (common_libraries ${libssl_LIBRARIES})
 include_directories (${libssl_INCLUDE_DIRS})
@@ -32,8 +30,8 @@ foreach (extra iconv rt)
 	find_library (extra_lib_${extra} ${extra})
 	if (extra_lib_${extra})
 		list (APPEND common_libraries ${extra})
-	endif (extra_lib_${extra})
+	endif ()
-endforeach (extra)
+endforeach ()
 
 # Build some unit tests
 include_directories (${PROJECT_SOURCE_DIR})
@@ -43,4 +41,4 @@ foreach (name liberty proto)
 	add_threads (test-${name})
 	target_link_libraries (test-${name} ${common_libraries})
 	add_test (NAME test-${name} COMMAND test-${name})
-endforeach (name)
+endforeach ()

liberty.c

@@ -2748,7 +2748,8 @@ utf8_decode (const char **s, size_t len)
 	}
 
 	// In the middle of a character
-	if (sequence_len == 1)
+	// or an overlong sequence (subset, possibly MUTF-8, not supported)
+	if (sequence_len == 1 || *p == 0xC0 || *p == 0xC1)
 		return -1;
 
 	// Check the rest of the sequence
@@ -2765,6 +2766,13 @@ utf8_decode (const char **s, size_t len)
 	return cp;
 }
 
+static inline bool
+utf8_validate_cp (int32_t cp)
+{
+	// RFC 3629, CESU-8 not allowed
+	return cp >= 0 && cp <= 0x10FFFF && (cp < 0xD800 || cp > 0xDFFF);
+}
+
 /// Very rough UTF-8 validation, just makes sure codepoints can be iterated
 static bool
 utf8_validate (const char *s, size_t len)
@@ -2772,7 +2780,7 @@ utf8_validate (const char *s, size_t len)
 	const char *end = s + len;
 	int32_t codepoint;
 	while ((codepoint = utf8_decode (&s, end - s)) >= 0
-		&& codepoint <= 0x10FFFF /* TODO: better validations */)
+		&& utf8_validate_cp (codepoint))
 		;
 	return s == end;
 }
@@ -4409,7 +4417,7 @@ socket_io_try_write (int socket_fd, struct str *wb)
 //   char    = [\0-\177]  # or any Unicode codepoint in the UTF-8 encoding
 //   escape  = [\\"abfnrtv] / [xX][0-9A-Fa-f][0-9A-Fa-f]? / [0-7][0-7]?[0-7]?
 //
-//   integer = lws '-'? [0-9]+  # whatever strtoll() accepts on your system
+//   integer = lws [-+]? [0-9]+  # whatever strtoll() accepts on your system
 //   null    = lws 'null'
 //   boolean = lws 'yes'  / lws 'YES'  / lws 'no'    / lws 'NO'
 //           / lws 'on'   / lws 'ON'   / lws 'off'   / lws 'OFF'

tests/liberty.c

@@ -331,10 +331,14 @@ test_utf8 (void)
 	soft_assert (utf8_decode (&partial, 1) == -2);
 	soft_assert (utf8_decode (&empty,   0) == -1);
 
-	const char valid  [] = "2H₂ + O₂ ⇌ 2H₂O, R = 4.7 kΩ, ⌀ 200 mm";
+	const char valid_1[] = "2H₂ + O₂ ⇌ 2H₂O, R = 4.7 kΩ, ⌀ 200 mm";
-	const char invalid[] = "\xf0\x90\x28\xbc";
+	const char valid_2[] = "\xf0\x93\x82\xb9";
-	soft_assert ( utf8_validate (valid,   sizeof valid));
+	const char invalid_1[] = "\xf0\x90\x28\xbc";
-	soft_assert (!utf8_validate (invalid, sizeof invalid));
+	const char invalid_2[] = "\xc0\x80";
+	soft_assert ( utf8_validate (valid_1,   sizeof valid_1));
+	soft_assert ( utf8_validate (valid_2,   sizeof valid_2));
+	soft_assert (!utf8_validate (invalid_1, sizeof invalid_1));
+	soft_assert (!utf8_validate (invalid_2, sizeof invalid_2));
 
 	struct utf8_iter iter = utf8_iter_make ("fóọ");
 	size_t ch_len;