[libc] add locale free strcoll

[michaelrj]

Summary

The strcoll function is intended to compare strings based on their
ordering in the current locale. Since the locale facilities have not yet
been added, a simple implementation that is the same as strcmp has been
added as a placeholder.

[sivachandra]

libc/src/string/strcoll.cpp

20

I think this is the same implementation as strcmp, but to avoid all type conversion/promotion confusions, we should cast the de-referenced value to int:

int(*reinterpret_cast<const unsigned char *>(left))

libc/test/src/string/strcoll_test.cpp

23

For completeness, you should add ASSERT_GT and ASSERT_LT tests also.

[michaelrj]

address comments

libc/test/src/string/strcoll_test.cpp

23

I'm avoiding adding GT and LT tests because they'll likely be wrong in some locales. Until we have a system for setting the locale per-test, similar to what we have for floating point rounding modes, any inequality comparison is unsafe.

[michaelrj]

rebase

[michaelrj]

add difference tests before land

libc/config/linux/x86_64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     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.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.src.string.strlcpy
     libc.src.string.strlen

libc/src/string/CMakeLists.txt

add_entrypoint_object(
   strcmp
   SRCS
     strcmp.cpp
   HDRS
     strcmp.h
 )
 
 add_entrypoint_object(
+  strcoll
+  SRCS
+    strcoll.cpp
+  HDRS
+    strcoll.h
+)
+
+add_entrypoint_object(
   strcpy
   SRCS
     strcpy.cpp
   HDRS
     strcpy.h
   DEPENDS
     .memory_utils.memcpy_implementation
     .string_utils

libc/src/string/strcoll.h

+//===-- Implementation header for strcoll -----------------------*- 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_STRCOLL_H
+#define LLVM_LIBC_SRC_STRING_STRCOLL_H
+
+namespace __llvm_libc {
+
+int strcoll(const char *left, const char *right);
+
+} // namespace __llvm_libc
+
+#endif // LLVM_LIBC_SRC_STRING_STRCOLL_H

libc/src/string/strcoll.cpp

+//===-- Implementation of strcoll -----------------------------------------===//
+//
+// 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/strcoll.h"
+
+#include "src/__support/common.h"
+
+namespace __llvm_libc {
+
+// TODO: Add support for locales.
+LLVM_LIBC_FUNCTION(int, strcoll, (const char *left, const char *right)) {
+  for (; *left && *left == *right; ++left, ++right)
+    ;
+  return static_cast<int>(*left) - static_cast<int>(*right);
+}
+
+} // namespace __llvm_libc

libc/test/src/string/CMakeLists.txt

SUITE
     libc_string_unittests
   SRCS
     strcmp_test.cpp
   DEPENDS
     libc.src.string.strcmp
 )
 
 add_libc_unittest(
+  strcoll_test
+  SUITE
+    libc_string_unittests
+  SRCS
+    strcoll_test.cpp
+  DEPENDS
+    libc.src.string.strcoll
+)
+
+add_libc_unittest(
   strcpy_test
   SUITE
     libc_string_unittests
   SRCS
     strcpy_test.cpp
   DEPENDS
     libc.src.string.strcpy
 )

libc/test/src/string/strcoll_test.cpp

+//===-- Unittests for strcoll ---------------------------------------------===//
+//
+// 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/strcoll.h"
+#include "utils/UnitTest/Test.h"
+
+// TODO: Add more comprehensive tests once locale support is added.
+
+TEST(LlvmLibcStrcollTest, SimpleTest) {
+  const char *s1 = "abc";
+  const char *s2 = "abc";
+  const char *s3 = "def";
+  int result = __llvm_libc::strcoll(s1, s2);
+  ASSERT_EQ(result, 0);
+
+  // Verify operands reversed.
+  result = __llvm_libc::strcoll(s2, s1);
+  ASSERT_EQ(result, 0);
+
+  result = __llvm_libc::strcoll(s1, s3);
+  ASSERT_LT(result, 0);
+
+  result = __llvm_libc::strcoll(s3, s1);
+  ASSERT_GT(result, 0);
+}