[libc] Add strchrnul implementation

[Caslyn]

Summary

Introduce strchrnul implementation and unit tests.

Ticket: https://fxbug.dev/124219

[Caslyn]

Remove extraneous line

[michaelrj]

Mostly LGTM with minor changes.

libc/spec/gnu_ext.td

79

this second ArgSpec should be IntType

libc/src/string/strchrnul.cpp

18

In general, we avoid calling other entrypoints from with our functions to keep the library more modular. Having strchrnul include strchr would mean that to include strchrnul in a final build it would also need to include strchr. It's unlikely to be a problem here, but it is a problem for functions like atoi and strtol.

In this case the strchr function is short enough that it can probably just be copied in here.

[Caslyn]

• Corrected IntType parameter
• Removed strchr dep, copied in logic

libc/spec/gnu_ext.td

79

thanks for catching!

libc/src/string/strchrnul.cpp

18

Gotcha - will remember that. Copied in the strchr logic in the updated patch.

[michaelrj]

LGTM after addressing the last comment.

libc/src/string/strchrnul.cpp

21–22

I think this can be simplified to just return ch.

[Caslyn]

• Simplified return statement

libc/src/string/strchrnul.cpp

21–22

Done (assumed you mean src).

libc/config/baremetal/arm/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncasecmp
     libc.src.string.strncat

libc/config/darwin/arm/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncasecmp
     libc.src.string.strncat

libc/config/darwin/x86_64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memmove
     libc.src.string.mempcpy
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncat
     libc.src.string.strncmp

libc/config/gpu/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncasecmp
     libc.src.string.strncat

libc/config/linux/aarch64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strdup
     libc.src.string.strerror
     libc.src.string.strerror_r
     libc.src.string.strlcat
     libc.src.string.strlcpy

libc/config/linux/arm/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncasecmp
     libc.src.string.strncat

libc/config/linux/riscv64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcoll
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strdup
     libc.src.string.strerror
     libc.src.string.strerror_r
     libc.src.string.strlcat

libc/config/linux/x86_64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcoll
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strdup
     libc.src.string.strerror
     libc.src.string.strerror_r
     libc.src.string.strlcat

libc/config/windows/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memrchr
     libc.src.string.memset
     libc.src.string.stpcpy
     libc.src.string.stpncpy
     libc.src.string.strcasecmp
     libc.src.string.strcasestr
     libc.src.string.strcat
     libc.src.string.strchr
+    libc.src.string.strchrnul
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlcat
     libc.src.string.strlcpy
     libc.src.string.strlen
     libc.src.string.strncasecmp
     libc.src.string.strncat

libc/spec/gnu_ext.td

HeaderSpec String = HeaderSpec<
             RetValSpec<CharPtr>,
             [ArgSpec<IntType>, ArgSpec<CharPtr>, ArgSpec<SizeTType>]
         >,
         FunctionSpec<
             "strcasestr",
             RetValSpec<CharPtr>,
             [ArgSpec<ConstCharPtr>, ArgSpec<ConstCharPtr>]
         >,
+        FunctionSpec<
+            "strchrnul",
+            RetValSpec<CharPtr>,
+            [ArgSpec<ConstCharPtr>, ArgSpec<ConstCharPtr>]
+        >,
       ]
   >;
 
   HeaderSpec FEnv = HeaderSpec<
       "fenv.h",
       [], // Macros
       [], // Types
       [], // Enumerations

libc/src/string/CMakeLists.txt

add_entrypoint_object(
   strchr
   SRCS
     strchr.cpp
   HDRS
     strchr.h
 )
 
 add_entrypoint_object(
+  strchrnul
+  SRCS
+    strchrnul.cpp
+  HDRS
+    strchrnul.h
+  DEPENDS
+    .strchr
+    .string_utils
+)
+
+add_entrypoint_object(
   strcmp
   SRCS
     strcmp.cpp
   HDRS
     strcmp.h
   DEPENDS
     .memory_utils.strcmp_implementation
 )

libc/src/string/strchrnul.h

+//===-- Implementation header for strchrnul --------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_SRC_STRING_STRCHRNUL_H
+#define LLVM_LIBC_SRC_STRING_STRCHRNUL_H
+
+namespace __llvm_libc {
+
+char *strchrnul(const char *src, int c);
+
+} // namespace __llvm_libc
+
+#endif // LLVM_LIBC_SRC_STRING_STRCHRNUL_H

libc/src/string/strchrnul.cpp

+//===-- Implementation of strchrnul --------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/strchrnul.h"
+#include "src/string/strchr.h"
+#include "src/string/string_utils.h"
+
+#include "src/__support/common.h"
+
+namespace __llvm_libc {
+
+LLVM_LIBC_FUNCTION(char *, strchrnul, (const char *src, int c)) {
+  char *ch = __llvm_libc::strchr(src, c);
+  return ch ? ch : const_cast<char *>(src) + internal::string_length(src);
+}
+
+} // namespace __llvm_libc

libc/test/src/string/CMakeLists.txt

SUITE
     libc_string_unittests
   SRCS
     strchr_test.cpp
   DEPENDS
     libc.src.string.strchr
 )
 
 add_libc_unittest(
+  strchrnul_test
+  SUITE
+    libc_string_unittests
+  SRCS
+    strchrnul_test.cpp
+  DEPENDS
+    libc.src.string.strchrnul
+)
+
+add_libc_unittest(
   strcmp_test
   SUITE
     libc_string_unittests
   SRCS
     strcmp_test.cpp
   DEPENDS
     libc.src.string.strcmp
 )

libc/test/src/string/strchr_test.cpp

 TEST(LlvmLibcStrChrTest, TheSourceShouldNotChange) {
   const char *src = "abcde";
   // When the character is found, the source string should not change.
   __llvm_libc::strchr(src, 'd');
   ASSERT_STREQ(src, "abcde");
   // Same case for when the character is not found.
   __llvm_libc::strchr(src, 'z');
   ASSERT_STREQ(src, "abcde");
-  // Same case for when looking for nullptr.
+  // Same case for when looking for null terminator.
   __llvm_libc::strchr(src, '\0');
   ASSERT_STREQ(src, "abcde");
 }
 
 TEST(LlvmLibcStrChrTest, ShouldFindFirstOfDuplicates) {
   // '1' is duplicated in the string, but it should find the first copy.
   ASSERT_STREQ(__llvm_libc::strchr("abc1def1ghi", '1'), "1def1ghi");
 

libc/test/src/string/strchrnul_test.cpp

+//===-- Unittests for strchrnul -------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/strchrnul.h"
+#include "test/UnitTest/Test.h"
+
+TEST(LlvmLibcStrChrNulTest, FindsFirstCharacter) {
+  const char *src = "abcde";
+
+  // Should return original string since 'a' is the first character.
+  ASSERT_STREQ(__llvm_libc::strchrnul(src, 'a'), "abcde");
+  // Source string should not change.
+  ASSERT_STREQ(src, "abcde");
+}
+
+TEST(LlvmLibcStrChrNulTest, FindsMiddleCharacter) {
+  const char *src = "abcde";
+
+  // Should return characters after (and including) 'c'.
+  ASSERT_STREQ(__llvm_libc::strchrnul(src, 'c'), "cde");
+  // Source string should not change.
+  ASSERT_STREQ(src, "abcde");
+}
+
+TEST(LlvmLibcStrChrNulTest, FindsLastCharacterThatIsNotNullTerminator) {
+  const char *src = "abcde";
+
+  // Should return 'e' and null-terminator.
+  ASSERT_STREQ(__llvm_libc::strchrnul(src, 'e'), "e");
+  // Source string should not change.
+  ASSERT_STREQ(src, "abcde");
+}
+
+TEST(LlvmLibcStrChrNulTest, FindsNullTerminator) {
+  const char *src = "abcde";
+
+  // Should return null terminator.
+  ASSERT_STREQ(__llvm_libc::strchrnul(src, '\0'), "");
+  // Source string should not change.
+  ASSERT_STREQ(src, "abcde");
+}
+
+TEST(LlvmLibcStrChrNulTest,
+     CharacterNotWithinStringShouldReturnNullTerminator) {
+  const char *src = "123?";
+
+  // Since 'z' is not within the string, should return a pointer to the source
+  // string's null terminator.
+  char *result = __llvm_libc::strchrnul(src, 'z');
+  ASSERT_EQ(*result, '\0');
+
+  char *term = const_cast<char *>(src) + 4;
+  ASSERT_EQ(result, term);
+}
+
+TEST(LlvmLibcStrChrNulTest, TheSourceShouldNotChange) {
+  const char *src = "abcde";
+  // When the character is found, the source string should not change.
+  __llvm_libc::strchrnul(src, 'd');
+  ASSERT_STREQ(src, "abcde");
+  // Same case for when the character is not found.
+  __llvm_libc::strchrnul(src, 'z');
+  ASSERT_STREQ(src, "abcde");
+  // Same case for when looking for null terminator.
+  __llvm_libc::strchrnul(src, '\0');
+  ASSERT_STREQ(src, "abcde");
+}
+
+TEST(LlvmLibcStrChrNulTest, ShouldFindFirstOfDuplicates) {
+  // '1' is duplicated in the string, but it should find the first copy.
+  ASSERT_STREQ(__llvm_libc::strchrnul("abc1def1ghi", '1'), "1def1ghi");
+
+  const char *dups = "XXXXX";
+  // Should return original string since 'X' is the first character.
+  ASSERT_STREQ(__llvm_libc::strchrnul(dups, 'X'), dups);
+}
+
+TEST(LlvmLibcStrChrNulTest, EmptyStringShouldOnlyMatchNullTerminator) {
+  // Null terminator should match.
+  ASSERT_STREQ(__llvm_libc::strchrnul("", '\0'), "");
+
+  // All other characters should not match.
+  char *result = __llvm_libc::strchrnul("", 'Z');
+  ASSERT_EQ(*result, '\0');
+
+  result = __llvm_libc::strchrnul("", '3');
+  ASSERT_EQ(*result, '\0');
+
+  result = __llvm_libc::strchrnul("", '*');
+  ASSERT_EQ(*result, '\0');
+}