Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
41 changes: 22 additions & 19 deletions libc/src/string/string_utils.h
Original file line number Diff line number Diff line change
Expand Up @@ -172,7 +172,7 @@ LIBC_INLINE size_t complementary_span(const char *src, const char *segment) {
return static_cast<size_t>(src - initial);
}

// Given the similarities between strtok and strtok_r, we can implement both
// Given the similarities between strsep/strtok/strtok_r, we can implement them
// using a utility function. On the first call, 'src' is scanned for the
// first character not found in 'delimiter_string'. Once found, it scans until
// the first character in the 'delimiter_string' or the null terminator is
Expand All @@ -184,33 +184,36 @@ LIBC_INLINE size_t complementary_span(const char *src, const char *segment) {
template <bool SkipDelim = true>
LIBC_INLINE char *string_token(char *__restrict src,
const char *__restrict delimiter_string,
char **__restrict saveptr) {
// Return nullptr immediately if both src AND saveptr are nullptr
if (LIBC_UNLIKELY(src == nullptr && ((src = *saveptr) == nullptr)))
char **__restrict context) {
// Return nullptr immediately if both src AND context are nullptr
if (LIBC_UNLIKELY(src == nullptr && ((src = *context) == nullptr)))
return nullptr;

static_assert(CHAR_BIT == 8, "bitset of 256 assumes char is 8 bits");
cpp::bitset<256> delimiter_set;
cpp::bitset<256> delims;
for (; *delimiter_string != '\0'; ++delimiter_string)
delimiter_set.set(static_cast<size_t>(*delimiter_string));
delims.set(static_cast<size_t>(*delimiter_string));

char *tok_start = src;
if constexpr (SkipDelim)
for (; *src != '\0' && delimiter_set.test(static_cast<size_t>(*src)); ++src)
;
if (*src == '\0') {
*saveptr = src;
while (*tok_start != '\0' && delims.test(static_cast<size_t>(*tok_start)))
++tok_start;
if (*tok_start == '\0' && SkipDelim) {
*context = nullptr;
return nullptr;
}
char *token = src;
for (; *src != '\0'; ++src) {
if (delimiter_set.test(static_cast<size_t>(*src))) {
*src = '\0';
++src;
break;
}

char *tok_end = tok_start;
while (*tok_end != '\0' && !delims.test(static_cast<size_t>(*tok_end)))
++tok_end;

if (*tok_end == '\0') {
*context = nullptr;
} else {
*tok_end = '\0';
*context = tok_end + 1;
}
*saveptr = src;
return token;
return tok_start;
}

LIBC_INLINE size_t strlcpy(char *__restrict dst, const char *__restrict src,
Expand Down
8 changes: 8 additions & 0 deletions libc/test/src/string/strsep_test.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,14 @@ TEST(LlvmLibcStrsepTest, DelimitersShouldNotBeIncludedInToken) {
}
}

TEST(LlvmLibcStrsepTest, SubsequentSearchesReturnNull) {
char s[] = "a";
char *string = s;
ASSERT_STREQ(LIBC_NAMESPACE::strsep(&string, ":"), "a");
ASSERT_EQ(LIBC_NAMESPACE::strsep(&string, ":"), nullptr);
ASSERT_EQ(LIBC_NAMESPACE::strsep(&string, ":"), nullptr);
}

#if defined(LIBC_ADD_NULL_CHECKS)

TEST(LlvmLibcStrsepTest, CrashOnNullPtr) {
Expand Down
9 changes: 9 additions & 0 deletions libc/test/src/string/strtok_r_test.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -122,3 +122,12 @@ TEST(LlvmLibcStrTokReentrantTest, DelimitersShouldNotBeIncludedInToken) {
token = LIBC_NAMESPACE::strtok_r(nullptr, "_:,_", &reserve);
ASSERT_STREQ(token, nullptr);
}

TEST(LlvmLibcStrTokReentrantTest, SubsequentSearchesReturnNull) {
char src[] = "a";
char *reserve = nullptr;
char *token = LIBC_NAMESPACE::strtok_r(src, ":", &reserve);
ASSERT_STREQ(token, "a");
ASSERT_EQ(LIBC_NAMESPACE::strtok_r(nullptr, ":", &reserve), nullptr);
ASSERT_EQ(LIBC_NAMESPACE::strtok_r(nullptr, ":", &reserve), nullptr);
}
7 changes: 7 additions & 0 deletions libc/test/src/string/strtok_test.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -76,3 +76,10 @@ TEST(LlvmLibcStrTokTest, DelimitersShouldNotBeIncludedInToken) {
token = LIBC_NAMESPACE::strtok(nullptr, "_:,_");
ASSERT_STREQ(token, nullptr);
}

TEST(LlvmLibcStrTokTest, SubsequentSearchesReturnNull) {
char src[] = "a";
ASSERT_STREQ("a", LIBC_NAMESPACE::strtok(src, ":"));
ASSERT_EQ(LIBC_NAMESPACE::strtok(nullptr, ":"), nullptr);
ASSERT_EQ(LIBC_NAMESPACE::strtok(nullptr, ":"), nullptr);
}
Loading