Compare commits

...

7 Commits

Author SHA1 Message Date
d71c47f8ce CMakeLists.txt: omit end{if,foreach} expressions
Their usefulness was almost negative.
2020-10-29 15:32:26 +01:00
425ea57b17 CMakeLists.txt: clean up OpenBSD support
A few things might have changed.
2020-10-29 15:31:05 +01:00
8822d06091 Don't suppress -Wimplicit-fallthrough
Might have already been resolved by: 9494e8e da75b6f
2020-10-26 18:25:32 +01:00
9639777814 Fix validation of overlong UTF-8
It was too strict and Egyptian dicks didn't want to pass,
so we'll do it half-arsedly for a subset.
2020-10-24 19:09:09 +02:00
929229a1d7 Fix config PEG grammar to match strtoll() 2020-10-24 08:05:17 +02:00
53bcebc2f0 Split out utf8_validate_cp(), adhere to RFC 3629 2020-10-21 05:20:20 +02:00
b08cf6c29f Reject overlong UTF-8 sequences 2020-10-21 05:08:59 +02:00
3 changed files with 25 additions and 15 deletions

View File

@@ -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})
endforeach (extra)
endif ()
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 ()

View File

@@ -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'

View File

@@ -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 invalid[] = "\xf0\x90\x28\xbc";
soft_assert ( utf8_validate (valid, sizeof valid));
soft_assert (!utf8_validate (invalid, sizeof invalid));
const char valid_1[] = "2H₂ + O₂ ⇌ 2H₂O, R = 4.7 kΩ, ⌀ 200 mm";
const char valid_2[] = "\xf0\x93\x82\xb9";
const char invalid_1[] = "\xf0\x90\x28\xbc";
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;