init

Author: shubh@DOE

libc/src/stdio/printf_core/CMakeLists.txt

@@ -67,6 +67,7 @@ add_header_library(
     parser.h
   DEPENDS
     .core_structs
+    libc.hdr.types.wint_t.h
     libc.src.__support.arg_list
     libc.src.__support.ctype_utils
     libc.src.__support.str_to_integer
@@ -111,6 +112,8 @@ add_header_library(
     .printf_config
     .writer
     libc.include.inttypes
+    libc.hdr.types.wchar_t
+    libc.hdr.types.wint_t
     libc.src.__support.big_int
     libc.src.__support.common
     libc.src.__support.CPP.limits
@@ -123,6 +126,8 @@ add_header_library(
     libc.src.__support.integer_to_string
     libc.src.__support.libc_assert
     libc.src.__support.uint128
+    libc.src.__support.wchar.mbstate
+    libc.src.__support.wchar.wcrtomb
     libc.src.__support.StringUtil.error_to_string
     libc.src.string.memory_utils.inline_memcpy
 )

libc/src/stdio/printf_core/char_converter.h

@@ -1,4 +1,4 @@
-//===-- String Converter for printf -----------------------------*- C++ -*-===//
+//===-- Character Converter for printf --------------------------*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
@@ -9,7 +9,11 @@
 #ifndef LLVM_LIBC_SRC_STDIO_PRINTF_CORE_CHAR_CONVERTER_H
 #define LLVM_LIBC_SRC_STDIO_PRINTF_CORE_CHAR_CONVERTER_H
 
+#include "hdr/types/wchar_t.h"
+#include "hdr/types/wint_t.h"
 #include "src/__support/macros/config.h"
+#include "src/__support/wchar/mbstate.h"
+#include "src/__support/wchar/wcrtomb.h"
 #include "src/stdio/printf_core/converter_utils.h"
 #include "src/stdio/printf_core/core_structs.h"
 #include "src/stdio/printf_core/writer.h"
@@ -21,7 +25,6 @@ template <WriteMode write_mode>
 LIBC_INLINE int convert_char(Writer<write_mode> *writer,
                              const FormatSection &to_conv) {
   char c = static_cast<char>(to_conv.conv_val_raw);
-
   constexpr int STRING_LEN = 1;
 
   size_t padding_spaces =
@@ -33,7 +36,27 @@ LIBC_INLINE int convert_char(Writer<write_mode> *writer,
     RET_IF_RESULT_NEGATIVE(writer->write(' ', padding_spaces));
   }
 
-  RET_IF_RESULT_NEGATIVE(writer->write(c));
+  if (to_conv.length_modifier == LengthModifier::l) {
+    wint_t wi = static_cast<wint_t>(to_conv.conv_val_raw);
+
+    if (wi == WEOF) {
+      return -1;
+    }
+
+    char mb_str[MB_LEN_MAX];
+    internal::mbstate mbstate;
+    wchar_t wc = static_cast<wchar_t>(wi);
+
+    auto ret = internal::wcrtomb(mb_str, wc, &internal_mbstate);
+    if (!ret.has_value()) {
+      return -1;
+    }
+
+    RET_IF_RESULT_NEGATIVE(writer->write({mb_str, ret.value()}));
+
+  } else {
+    RET_IF_RESULT_NEGATIVE(writer->write(c));
+  }
 
   // If the padding is on the right side, write the spaces last.
   if (padding_spaces > 0 &&

libc/src/stdio/printf_core/parser.h

@@ -9,6 +9,7 @@
 #ifndef LLVM_LIBC_SRC_STDIO_PRINTF_CORE_PARSER_H
 #define LLVM_LIBC_SRC_STDIO_PRINTF_CORE_PARSER_H
 
+#include "hdr/types/wint_t.h"
 #include "include/llvm-libc-macros/stdfix-macros.h"
 #include "src/__support/CPP/algorithm.h" // max
 #include "src/__support/CPP/limits.h"
@@ -73,9 +74,9 @@ template <typename ArgProvider> class Parser {
   ArgProvider args_cur;
 
 #ifndef LIBC_COPT_PRINTF_DISABLE_INDEX_MODE
-  // args_start stores the start of the va_args, which is allows getting the
-  // value of arguments that have already been passed. args_index is tracked so
-  // that we know which argument args_cur is on.
+  // args_start stores the start of the va_args, which helps in getting the
+  // number of arguments that have already been passed. args_index is tracked
+  // so that we know which argument args_cur is on.
   ArgProvider args_start;
   size_t args_index = 1;
 
@@ -173,7 +174,11 @@ template <typename ArgProvider> class Parser {
         section.has_conv = true;
         break;
       case ('c'):
-        WRITE_ARG_VAL_SIMPLEST(section.conv_val_raw, int, conv_index);
+        if (section.length_modifier == LengthModifier::l) {
+          WRITE_ARG_VAL_SIMPLEST(section.conv_val_raw, wint_t, conv_index);
+        } else {
+          WRITE_ARG_VAL_SIMPLEST(section.conv_val_raw, int, conv_index);
+        }
         break;
       case ('d'):
       case ('i'):

libc/test/src/stdio/printf_core/CMakeLists.txt

@@ -7,6 +7,7 @@ add_libc_unittest(
   LINK_LIBRARIES
     LibcPrintfHelpers
   DEPENDS
+    libc.hdr.types.wchar_t
     libc.src.stdio.printf_core.parser
     libc.src.stdio.printf_core.core_structs
     libc.src.__support.CPP.string_view
@@ -32,6 +33,7 @@ add_libc_unittest(
   SRCS
     converter_test.cpp
   DEPENDS
+    libc.hdr.types.wchar_t
     libc.src.stdio.printf_core.converter
     libc.src.stdio.printf_core.writer
     libc.src.stdio.printf_core.core_structs

libc/test/src/stdio/printf_core/converter_test.cpp

@@ -6,10 +6,10 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "hdr/types/wchar_t.h"
 #include "src/stdio/printf_core/converter.h"
 #include "src/stdio/printf_core/core_structs.h"
 #include "src/stdio/printf_core/writer.h"
-
 #include "test/UnitTest/Test.h"
 
 class LlvmLibcPrintfConverterTest : public LIBC_NAMESPACE::testing::Test {
@@ -255,3 +255,83 @@ TEST_F(LlvmLibcPrintfConverterTest, OctConversion) {
   ASSERT_STREQ(str, "1234");
   ASSERT_EQ(writer.get_chars_written(), size_t{4});
 }
+
+TEST_F(LlvmLibcPrintfConverterTest, WideCharConversion) {
+  LIBC_NAMESPACE::printf_core::FormatSection section;
+  section.has_conv = true;
+  section.raw_string = "%lc";
+  section.conv_name = 'c';
+  section.length_modifier = LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  section.conv_val_raw = static_cast<wchar_t>(L'€');
+
+  LIBC_NAMESPACE::printf_core::convert(&writer, section);
+
+  wb.buff[wb.buff_cur] = '\0';
+
+  ASSERT_STREQ(str, "€");
+  ASSERT_EQ(writer.get_chars_written(), size_t{1});
+}
+
+TEST_F(LlvmLibcPrintfConverterTest, WideCharConversionLeftJustified) {
+  LIBC_NAMESPACE::printf_core::FormatSection left_justified_conv;
+  left_justified_conv.has_conv = true;
+  left_justified_conv.raw_string = "%-4lc";
+  left_justified_conv.conv_name = 'c';
+  left_justified_conv.length_modifier =
+      LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  left_justified_conv.flags =
+      LIBC_NAMESPACE::printf_core::FormatFlags::LEFT_JUSTIFIED;
+  left_justified_conv.min_width = 4;
+  left_justified_conv.conv_val_raw = static_cast<wchar_t>(L'€');
+
+  LIBC_NAMESPACE::printf_core::convert(&writer, left_justified_conv);
+  wb.buff[wb.buff_cur] = '\0';
+
+  ASSERT_STREQ(str, "€   ");
+  ASSERT_EQ(writer.get_chars_written(), size_t{4});
+}
+
+TEST_F(LlvmLibcPrintfConverterTest, WideCharConversionRightJustified) {
+  LIBC_NAMESPACE::printf_core::FormatSection right_justified_conv;
+  right_justified_conv.has_conv = true;
+  right_justified_conv.raw_string = "%4lc";
+  right_justified_conv.conv_name = 'c';
+  right_justified_conv.length_modifier =
+      LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  right_justified_conv.min_width = 4;
+  right_justified_conv.conv_val_raw = static_cast<wchar_t>(L'€');
+
+  LIBC_NAMESPACE::printf_core::convert(&writer, right_justified_conv);
+  wb.buff[wb.buff_cur] = '\0';
+
+  ASSERT_STREQ(str, "   €");
+  ASSERT_EQ(writer.get_chars_written(), size_t{4});
+}
+
+TEST_F(LlvmLibcPrintfConverterTest, WideCharConversionInvalid) {
+  LIBC_NAMESPACE::printf_core::FormatSection section;
+  section.has_conv = true;
+  section.raw_string = "%lc";
+  section.conv_name = 'c';
+  section.length_modifier = LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  // An invalid wide character.
+  section.conv_val_raw = static_cast<wchar_t>(0xFFFFFFFF);
+
+  int ret = LIBC_NAMESPACE::printf_core::convert(&writer, section);
+
+  ASSERT_EQ(ret, -1);
+}
+
+TEST_F(LlvmLibcPrintfConverterTest, WideCharWEOFConversion) {
+  LIBC_NAMESPACE::printf_core::FormatSection section;
+  section.has_conv = true;
+  section.raw_string = "%lc";
+  section.conv_name = 'c';
+  section.length_modifier = LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  // WEOF value.
+  section.conv_val_raw = static_cast<wchar_t>(WEOF);
+
+  int ret = LIBC_NAMESPACE::printf_core::convert(&writer, section);
+
+  ASSERT_EQ(ret, -1);
+}

libc/test/src/stdio/printf_core/parser_test.cpp

@@ -6,6 +6,7 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "hdr/types/wchar_t.h"
 #include "src/__support/CPP/bit.h"
 #include "src/__support/CPP/string_view.h"
 #include "src/__support/arg_list.h"
@@ -370,6 +371,24 @@ TEST(LlvmLibcPrintfParserTest,
   ASSERT_PFORMAT_EQ(expected, format_arr[0]);
 }
 
+TEST(LlvmLibcPrintfParserTest, EvalOneArgWithWideCharacter) {
+  LIBC_NAMESPACE::printf_core::FormatSection format_arr[2];
+  const char *str = "%lc";
+  wchar_t arg1 = L'€';
+  evaluate(format_arr, str, arg1);
+
+  LIBC_NAMESPACE::printf_core::FormatSection expected;
+  expected.has_conv = true;
+
+  expected.raw_string = {str, 3};
+  expected.length_modifier = LIBC_NAMESPACE::printf_core::LengthModifier::l;
+  expected.conv_val_raw =
+      static_cast<LIBC_NAMESPACE::fputil::FPBits<double>::StorageType>(arg1);
+  expected.conv_name = 'c';
+
+  ASSERT_PFORMAT_EQ(expected, format_arr[0]);
+}
+
 #ifndef LIBC_COPT_PRINTF_DISABLE_INDEX_MODE
 
 TEST(LlvmLibcPrintfParserTest, IndexModeOneArg) {