[libc] wcscspn implementation (#146158)

Implemented wcscspn and tests.

---------

Co-authored-by: Sriya Pratipati <sriyap@google.com>

7 files changed, 152 insertions, 0 deletions

diff --git a/libc/config/linux/x86_64/entrypoints.txt b/libc/config/linux/x86_64/entrypoints.txt
index 263669803c29..59c248871f83 100644
--- a/libc/config/linux/x86_64/entrypoints.txt
+++ b/libc/config/linux/x86_64/entrypoints.txt
@@ -376,6 +376,7 @@ set(TARGET_LIBC_ENTRYPOINTS
     libc.src.wchar.wcspbrk
     libc.src.wchar.wcsrchr
     libc.src.wchar.wcsspn
+    libc.src.wchar.wcscspn
     libc.src.wchar.wmemcmp
     libc.src.wchar.wmempcpy
     libc.src.wchar.wmemcpy
diff --git a/libc/include/wchar.yaml b/libc/include/wchar.yaml
index 0430a33343ff..02e1ba331b7c 100644
--- a/libc/include/wchar.yaml
+++ b/libc/include/wchar.yaml
@@ -104,6 +104,13 @@ functions:
     arguments:
       - type: const wchar_t *
       - type: const wchar_t *
+  - name: wcscspn
+    standards:
+      - stdc
+    return_type: size_t
+    arguments:
+      - type: const wchar_t *
+      - type: const wchar_t *
   - name: wmemcmp
     standards:
       - stdc
diff --git a/libc/src/wchar/CMakeLists.txt b/libc/src/wchar/CMakeLists.txt
index 833735d71639..f2f4b1d38f0f 100644
--- a/libc/src/wchar/CMakeLists.txt
+++ b/libc/src/wchar/CMakeLists.txt
@@ -214,6 +214,17 @@ add_entrypoint_object(
 )
 
 add_entrypoint_object(
+  wcscspn
+  SRCS
+    wcscspn.cpp
+  HDRS
+    wcscspn.h
+  DEPENDS
+    libc.hdr.wchar_macros
+    libc.hdr.types.size_t
+)
+
+add_entrypoint_object(
   wmemcmp
   SRCS
     wmemcmp.cpp
diff --git a/libc/src/wchar/wcscspn.cpp b/libc/src/wchar/wcscspn.cpp
new file mode 100644
index 000000000000..8869d84cdfde
--- /dev/null
+++ b/libc/src/wchar/wcscspn.cpp
@@ -0,0 +1,34 @@
+//===-- Implementation of wcscspn -----------------------------------------===//
+//
+// 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/wchar/wcscspn.h"
+
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/common.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+bool check(wchar_t c, const wchar_t *s2) {
+  for (int n = 0; s2[n]; ++n) {
+    if (s2[n] == c)
+      return false;
+  }
+  return true;
+}
+LLVM_LIBC_FUNCTION(size_t, wcscspn, (const wchar_t *s1, const wchar_t *s2)) {
+  size_t i = 0;
+  for (; s1[i]; ++i) {
+    if (!check(s1[i], s2))
+      return i;
+  }
+  return i;
+}
+
+} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/wchar/wcscspn.h b/libc/src/wchar/wcscspn.h
new file mode 100644
index 000000000000..cffc2b98c346
--- /dev/null
+++ b/libc/src/wchar/wcscspn.h
@@ -0,0 +1,22 @@
+//===-- Implementation header for wcscspn ---------------------------------===//
+//
+// 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_WCHAR_WCSCSPN_H
+#define LLVM_LIBC_SRC_WCHAR_WCSCSPN_H
+
+#include "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+size_t wcscspn(const wchar_t *s1, const wchar_t *s2);
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_WCHAR_WCSCSPN_H
diff --git a/libc/test/src/wchar/CMakeLists.txt b/libc/test/src/wchar/CMakeLists.txt
index 60f2c2e0f166..657343d07215 100644
--- a/libc/test/src/wchar/CMakeLists.txt
+++ b/libc/test/src/wchar/CMakeLists.txt
@@ -184,6 +184,16 @@ add_libc_test(
 )
 
 add_libc_test(
+  wcscspn_test
+  SUITE
+    libc_wchar_unittests
+  SRCS
+    wcscspn_test.cpp
+  DEPENDS
+    libc.src.wchar.wcscspn
+)
+
+add_libc_test(
   wmemchr_test
   SUITE
     libc_wchar_unittests
diff --git a/libc/test/src/wchar/wcscspn_test.cpp b/libc/test/src/wchar/wcscspn_test.cpp
new file mode 100644
index 000000000000..6318302d1005
--- /dev/null
+++ b/libc/test/src/wchar/wcscspn_test.cpp
@@ -0,0 +1,67 @@
+//===-- Unittests for wcscspn ---------------------------------------------===//
+//
+// 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 "hdr/types/size_t.h"
+#include "hdr/types/wchar_t.h"
+#include "src/wchar/wcscspn.h"
+#include "test/UnitTest/Test.h"
+
+TEST(LlvmLibcWCSCSpnTest, EmptyStringShouldReturnZeroLengthSpan) {
+  // The search should not include the null terminator.
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"", L""), size_t(0));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"_", L""), size_t(1));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"", L"_"), size_t(0));
+}
+
+TEST(LlvmLibcWCSCSpnTest, ShouldNotSpanAnythingAfterNullTerminator) {
+  const wchar_t src[4] = {L'a', L'b', L'\0', L'c'};
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"de"), size_t(2));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"c"), size_t(2));
+
+  // Same goes for the segment to be searched for.
+  const wchar_t segment[4] = {L'1', L'2', L'\0', L'3'};
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"3", segment), size_t(1));
+}
+
+TEST(LlvmLibcWCSCSpnTest, SpanEachIndividualCharacter) {
+  const wchar_t *src = L"12345";
+  // These are all in the segment.
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"1"), size_t(0));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"2"), size_t(1));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"3"), size_t(2));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"4"), size_t(3));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"5"), size_t(4));
+}
+
+TEST(LlvmLibcWCSCSpnTest, UnmatchedCharacterShouldReturnLength) {
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"a", L"b"), size_t(1));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"abcdef", L"1"), size_t(6));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"123", L"4"), size_t(3));
+}
+
+TEST(LlvmLibcWCSCSpnTest, NonSequentialCharactersShouldNotSpan) {
+  const wchar_t *src = L"abc456789";
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"_1_abc_2_def_3_"), size_t(0));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(src, L"67__34xyz12"), size_t(3));
+}
+
+TEST(LlvmLibcWCSCSpnTest, ReverseCharacters) {
+  // These are all in the string.
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"12345", L"54321"), size_t(0));
+  // 1 is not in the span.
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"12345", L"432"), size_t(1));
+  // 1 is in the span.
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"12345", L"51"), size_t(0));
+}
+
+TEST(LlvmLibcWCSCSpnTest, DuplicatedCharactersToBeSearchedForShouldStillMatch) {
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"a", L"aa"), size_t(0));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"aa", L"aa"), size_t(0));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"aaa", L"bb"), size_t(3));
+  EXPECT_EQ(LIBC_NAMESPACE::wcscspn(L"aaaa", L"bb"), size_t(4));
+}