AK+Everywhere: Turn bool keep_empty to an enum in split* functions

AK/IPv6Address.h

@@ -125,7 +125,7 @@ class [[gnu::packed]] IPv6Address {
  if (string.is_null())
  return {};
 
- auto const parts = string.split_view(':', true);
+ auto const parts = string.split_view(':', SplitBehavior::KeepEmpty);
  if (parts.is_empty())
  return {};
  if (parts.size() > 9) {

AK/String.cpp

@@ -97,18 +97,19 @@ StringView String::substring_view(size_t start) const
  return { characters() + start, length() - start };
 }
 
-Vector<String> String::split(char separator, bool keep_empty) const
+Vector<String> String::split(char separator, SplitBehavior split_behavior) const
 {
- return split_limit(separator, 0, keep_empty);
+ return split_limit(separator, 0, split_behavior);
 }
 
-Vector<String> String::split_limit(char separator, size_t limit, bool keep_empty) const
+Vector<String> String::split_limit(char separator, size_t limit, SplitBehavior split_behavior) const
 {
  if (is_empty())
  return {};
 
  Vector<String> v;
  size_t substart = 0;
+ bool keep_empty = has_flag(split_behavior, SplitBehavior::KeepEmpty);
  for (size_t i = 0; i < length() && (v.size() + 1) != limit; ++i) {
  char ch = characters()[i];
  if (ch == separator) {
@@ -124,13 +125,14 @@ Vector<String> String::split_limit(char separator, size_t limit, bool keep_empty
  return v;
 }
 
-Vector<StringView> String::split_view(Function<bool(char)> separator, bool keep_empty) const
+Vector<StringView> String::split_view(Function<bool(char)> separator, SplitBehavior split_behavior) const
 {
  if (is_empty())
  return {};
 
  Vector<StringView> v;
  size_t substart = 0;
+ bool keep_empty = has_flag(split_behavior, SplitBehavior::KeepEmpty);
  for (size_t i = 0; i < length(); ++i) {
  char ch = characters()[i];
  if (separator(ch)) {
@@ -146,9 +148,9 @@ Vector<StringView> String::split_view(Function<bool(char)> separator, bool keep_
  return v;
 }
 
-Vector<StringView> String::split_view(char const separator, bool keep_empty) const
+Vector<StringView> String::split_view(char const separator, SplitBehavior split_behavior) const
 {
- return split_view([separator](char ch) { return ch == separator; }, keep_empty);
+ return split_view([separator](char ch) { return ch == separator; }, split_behavior);
 }
 
 ByteBuffer String::to_byte_buffer() const

AK/String.h

@@ -150,10 +150,10 @@ class String {
  [[nodiscard]] bool contains(StringView, CaseSensitivity = CaseSensitivity::CaseSensitive) const;
  [[nodiscard]] bool contains(char, CaseSensitivity = CaseSensitivity::CaseSensitive) const;
 
- [[nodiscard]] Vector<String> split_limit(char separator, size_t limit, bool keep_empty = false) const;
- [[nodiscard]] Vector<String> split(char separator, bool keep_empty = false) const;
- [[nodiscard]] Vector<StringView> split_view(char separator, bool keep_empty = false) const;
- [[nodiscard]] Vector<StringView> split_view(Function<bool(char)> separator, bool keep_empty = false) const;
+ [[nodiscard]] Vector<String> split_limit(char separator, size_t limit, SplitBehavior = SplitBehavior::Nothing) const;
+ [[nodiscard]] Vector<String> split(char separator, SplitBehavior = SplitBehavior::Nothing) const;
+ [[nodiscard]] Vector<StringView> split_view(char separator, SplitBehavior = SplitBehavior::Nothing) const;
+ [[nodiscard]] Vector<StringView> split_view(Function<bool(char)> separator, SplitBehavior = SplitBehavior::Nothing) const;
 
  [[nodiscard]] Optional<size_t> find(char needle, size_t start = 0) const { return StringUtils::find(*this, needle, start); }
  [[nodiscard]] Optional<size_t> find(StringView needle, size_t start = 0) const { return StringUtils::find(*this, needle, start); }

AK/StringUtils.h

@@ -8,6 +8,7 @@
 #pragma once
 
 #include <AK/Concepts.h>
+#include <AK/EnumBits.h>
 #include <AK/Forward.h>
 
 namespace AK {
@@ -38,6 +39,14 @@ enum class TrimWhitespace {
  No,
 };
 
+enum class SplitBehavior : unsigned {
+ Nothing = 0,
+ // If two separators follow each other without any characters
+ // in between, keep a "" in the resulting vector.
+ KeepEmpty = 1,
+};
+AK_ENUM_BITWISE_OPERATORS(SplitBehavior);
+
 struct MaskSpan {
  size_t start;
  size_t length;
@@ -99,5 +108,6 @@ size_t count(StringView, StringView needle);
 
 using AK::CaseSensitivity;
 using AK::ReplaceMode;
+using AK::SplitBehavior;
 using AK::TrimMode;
 using AK::TrimWhitespace;

AK/StringView.cpp

@@ -40,16 +40,16 @@ StringView::StringView(ByteBuffer const& buffer)
 {
 }
 
-Vector<StringView> StringView::split_view(char const separator, bool keep_empty) const
+Vector<StringView> StringView::split_view(char const separator, SplitBehavior split_behavior) const
 {
  StringView seperator_view { &separator, 1 };
- return split_view(seperator_view, keep_empty);
+ return split_view(seperator_view, split_behavior);
 }
 
-Vector<StringView> StringView::split_view(StringView separator, bool keep_empty) const
+Vector<StringView> StringView::split_view(StringView separator, SplitBehavior split_behavior) const
 {
  Vector<StringView> parts;
- for_each_split_view(separator, keep_empty, [&](StringView view) {
+ for_each_split_view(separator, split_behavior, [&](StringView view) {
  parts.append(view);
  });
  return parts;
@@ -61,7 +61,7 @@ Vector<StringView> StringView::lines(bool consider_cr) const
  return {};
 
  if (!consider_cr)
- return split_view('\n', true);
+ return split_view('\n', SplitBehavior::KeepEmpty);
 
  Vector<StringView> v;
  size_t substart = 0;
@@ -264,13 +264,14 @@ Vector<size_t> StringView::find_all(StringView needle) const
  return StringUtils::find_all(*this, needle);
 }
 
-Vector<StringView> StringView::split_view_if(Function<bool(char)> const& predicate, bool keep_empty) const
+Vector<StringView> StringView::split_view_if(Function<bool(char)> const& predicate, SplitBehavior split_behavior) const
 {
  if (is_empty())
  return {};
 
  Vector<StringView> v;
  size_t substart = 0;
+ bool keep_empty = has_flag(split_behavior, SplitBehavior::KeepEmpty);
  for (size_t i = 0; i < length(); ++i) {
  char ch = characters_without_null_termination()[i];
  if (predicate(ch)) {

AK/StringView.h

@@ -8,6 +8,7 @@
 
 #include <AK/Assertions.h>
 #include <AK/Checked.h>
+#include <AK/EnumBits.h>
 #include <AK/Forward.h>
 #include <AK/Optional.h>
 #include <AK/Span.h>
@@ -129,10 +130,10 @@ class StringView {
  return substring_view(start, length() - start);
  }
 
- [[nodiscard]] Vector<StringView> split_view(char, bool keep_empty = false) const;
- [[nodiscard]] Vector<StringView> split_view(StringView, bool keep_empty = false) const;
+ [[nodiscard]] Vector<StringView> split_view(char, SplitBehavior = SplitBehavior::Nothing) const;
+ [[nodiscard]] Vector<StringView> split_view(StringView, SplitBehavior = SplitBehavior::Nothing) const;
 
- [[nodiscard]] Vector<StringView> split_view_if(Function<bool(char)> const& predicate, bool keep_empty = false) const;
+ [[nodiscard]] Vector<StringView> split_view_if(Function<bool(char)> const& predicate, SplitBehavior = SplitBehavior::Nothing) const;
 
  [[nodiscard]] StringView find_last_split_view(char separator) const
  {
@@ -151,14 +152,14 @@ class StringView {
  }
 
  template<VoidFunction<StringView> Callback>
- void for_each_split_view(char separator, bool keep_empty, Callback callback) const
+ void for_each_split_view(char separator, SplitBehavior split_behavior, Callback callback) const
  {
  StringView seperator_view { &separator, 1 };
- for_each_split_view(seperator_view, keep_empty, callback);
+ for_each_split_view(seperator_view, split_behavior, callback);
  }
 
  template<VoidFunction<StringView> Callback>
- void for_each_split_view(StringView separator, bool keep_empty, Callback callback) const
+ void for_each_split_view(StringView separator, SplitBehavior split_behavior, Callback callback) const
  {
  VERIFY(!separator.is_empty());
 
@@ -168,6 +169,7 @@ class StringView {
  StringView view { *this };
 
  auto maybe_separator_index = find(separator);
+ bool keep_empty = has_flag(split_behavior, SplitBehavior::KeepEmpty);
  while (maybe_separator_index.has_value()) {
  auto separator_index = maybe_separator_index.value();
  auto part_with_separator = view.substring_view(0, separator_index + separator.length());

Kernel/CommandLine.cpp

@@ -65,7 +65,7 @@ UNMAP_AFTER_INIT void CommandLine::add_arguments(Vector<StringView> const& args)
  continue;
  }
 
- auto pair = str.split_view('=', false);
+ auto pair = str.split_view('=');
  VERIFY(pair.size() == 2 || pair.size() == 1);
 
  if (pair.size() == 1) {

Kernel/GlobalProcessExposed.cpp

@@ -642,7 +642,7 @@ class ProcFSCPUInformation final : public ProcFSGlobalInformation {
  TRY(obj.add("hypervisor_vendor_id"sv, info.hypervisor_vendor_id_string()));
 
  auto features_array = TRY(obj.add_array("features"sv));
- auto keep_empty = false;
+ auto keep_empty = SplitBehavior::KeepEmpty;
 
  ErrorOr<void> result; // FIXME: Make this nicer
  info.features_string().for_each_split_view(' ', keep_empty, [&](StringView feature) {

Kernel/Storage/StorageManagement.cpp

@@ -213,7 +213,7 @@ UNMAP_AFTER_INIT Array<unsigned, 3> StorageManagement::extract_boot_device_addre
  auto parameters_view = m_boot_argument.substring_view(device_prefix.length()).find_first_split_view(';');
  size_t parts_count = 0;
  bool parse_failure = false;
- parameters_view.for_each_split_view(':', false, [&](StringView parameter_view) {
+ parameters_view.for_each_split_view(':', SplitBehavior::Nothing, [&](StringView parameter_view) {
  if (parse_failure)
  return;
  if (parts_count > 2)

Kernel/Syscalls/pledge.cpp

@@ -29,7 +29,7 @@ ErrorOr<FlatPtr> Process::sys$pledge(Userspace<Syscall::SC_pledge_params const*>
 
  auto parse_pledge = [&](auto pledge_spec, u32& mask) {
  auto found_invalid_pledge = true;
- pledge_spec.for_each_split_view(' ', false, [&mask, &found_invalid_pledge](auto const& part) {
+ pledge_spec.for_each_split_view(' ', SplitBehavior::Nothing, [&mask, &found_invalid_pledge](auto const& part) {
 #define __ENUMERATE_PLEDGE_PROMISE(x) \
  if (part == #x##sv) { \
  mask |= (1u << (u32)Pledge::x); \

Meta/Lagom/Tools/CodeGenerators/LibEDID/GeneratePnpIDs.cpp

@@ -91,7 +91,7 @@ static ErrorOr<String> decode_html_entities(StringView const& str)
 
 static ErrorOr<ApprovalDate> parse_approval_date(StringView const& str)
 {
- auto parts = str.trim_whitespace().split_view('/', true);
+ auto parts = str.trim_whitespace().split_view('/', SplitBehavior::KeepEmpty);
  if (parts.size() != 3)
  return Error::from_string_literal("Failed to parse approval date parts (mm/dd/yyyy)");
 

Meta/Lagom/Tools/CodeGenerators/LibLocale/GeneratePluralRulesData.cpp

@@ -258,7 +258,7 @@ static Relation parse_relation(StringView relation)
  parsed.symbol = lhs[0];
  }
 
- rhs.for_each_split_view(set_operator, false, [&](auto set) {
+ rhs.for_each_split_view(set_operator, SplitBehavior::Nothing, [&](auto set) {
  if (auto index = set.find(range_operator); index.has_value()) {
  auto range_begin = set.substring_view(0, *index).to_uint();
  VERIFY(range_begin.has_value());
@@ -313,10 +313,10 @@ static void parse_condition(StringView category, StringView rule, Conditions& ru
  // and_condition = relation ('and' relation)*
  //
  // This affords some simplicity in that disjunctions are never embedded within a conjunction.
- condition.for_each_split_view(disjunction_keyword, false, [&](auto disjunction) {
+ condition.for_each_split_view(disjunction_keyword, SplitBehavior::Nothing, [&](auto disjunction) {
  Vector<Relation> conjunctions;
 
- disjunction.for_each_split_view(conjunction_keyword, false, [&](auto relation) {
+ disjunction.for_each_split_view(conjunction_keyword, SplitBehavior::Nothing, [&](auto relation) {
  conjunctions.append(parse_relation(relation));
  });
 

Meta/Lagom/Tools/CodeGenerators/LibTimeZone/GenerateTimeZoneData.cpp

@@ -400,7 +400,7 @@ static ErrorOr<void> parse_time_zone_coordinates(Core::Stream::BufferedFile& fil
 
  time_zone_data.time_zone_coordinates.set(zone, { latitude, longitude });
 
- regions.for_each_split_view(',', false, [&](auto region) {
+ regions.for_each_split_view(',', SplitBehavior::Nothing, [&](auto region) {
  auto index = time_zone_data.unique_strings.ensure(zone);
  time_zone_data.time_zone_regions.ensure(region).append(index);
 

Meta/Lagom/Tools/CodeGenerators/LibUnicode/GenerateEmojiData.cpp

@@ -112,7 +112,7 @@ static ErrorOr<void> parse_emoji_serenity_data(Core::Stream::BufferedFile& file,
  emoji.group = Unicode::EmojiGroup::SerenityOS;
  emoji.display_order = display_order++;
 
- line.for_each_split_view(' ', false, [&](auto segment) {
+ line.for_each_split_view(' ', SplitBehavior::Nothing, [&](auto segment) {
  if (segment.starts_with(code_point_header)) {
  segment = segment.substring_view(code_point_header.length());
 

Meta/Lagom/Tools/CodeGenerators/LibUnicode/GenerateUnicodeData.cpp

@@ -223,7 +223,7 @@ static ErrorOr<void> parse_special_casing(Core::Stream::BufferedFile& file, Unic
  if (auto index = line.find('#'); index.has_value())
  line = line.substring_view(0, *index);
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY(segments.size() == 5 || segments.size() == 6);
 
  SpecialCasing casing {};
@@ -233,7 +233,7 @@ static ErrorOr<void> parse_special_casing(Core::Stream::BufferedFile& file, Unic
  casing.uppercase_mapping = parse_code_point_list(segments[3]);
 
  if (auto condition = segments[4].trim_whitespace(); !condition.is_empty()) {
- auto conditions = condition.split_view(' ', true);
+ auto conditions = condition.split_view(' ', SplitBehavior::KeepEmpty);
  VERIFY(conditions.size() == 1 || conditions.size() == 2);
 
  if (conditions.size() == 2) {
@@ -294,7 +294,7 @@ static ErrorOr<void> parse_prop_list(Core::Stream::BufferedFile& file, PropList&
  if (auto index = line.find('#'); index.has_value())
  line = line.substring_view(0, *index);
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY(segments.size() == 2);
 
  auto code_point_range = parse_code_point_range(segments[0].trim_whitespace());
@@ -344,7 +344,7 @@ static ErrorOr<void> parse_alias_list(Core::Stream::BufferedFile& file, PropList
  if (current_property != "Binary Properties"sv)
  continue;
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY((segments.size() == 2) || (segments.size() == 3));
 
  auto alias = segments[0].trim_whitespace();
@@ -370,7 +370,7 @@ static ErrorOr<void> parse_name_aliases(Core::Stream::BufferedFile& file, Unicod
  if (line.is_empty() || line.starts_with('#'))
  continue;
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY(segments.size() == 3);
 
  auto code_point = AK::StringUtils::convert_to_uint_from_hex<u32>(segments[0].trim_whitespace());
@@ -417,7 +417,7 @@ static ErrorOr<void> parse_value_alias_list(Core::Stream::BufferedFile& file, St
  if (auto index = line.find('#'); index.has_value())
  line = line.substring_view(0, *index);
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  auto category = segments[0].trim_whitespace();
 
  if (category != desired_category)
@@ -450,7 +450,7 @@ static ErrorOr<void> parse_normalization_props(Core::Stream::BufferedFile& file,
  if (auto index = line.find('#'); index.has_value())
  line = line.substring_view(0, *index);
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY((segments.size() == 2) || (segments.size() == 3));
 
  auto code_point_range = parse_code_point_range(segments[0].trim_whitespace());
@@ -579,7 +579,7 @@ static ErrorOr<void> parse_block_display_names(Core::Stream::BufferedFile& file,
  if (line.is_empty() || line.starts_with('#'))
  continue;
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY(segments.size() == 2);
 
  auto code_point_range = parse_code_point_range(segments[0].trim_whitespace());
@@ -610,7 +610,7 @@ static ErrorOr<void> parse_unicode_data(Core::Stream::BufferedFile& file, Unicod
  if (line.is_empty())
  continue;
 
- auto segments = line.split_view(';', true);
+ auto segments = line.split_view(';', SplitBehavior::KeepEmpty);
  VERIFY(segments.size() == 15);
 
  CodePointData data {};

Meta/Lagom/Tools/ConfigureComponents/main.cpp

@@ -195,7 +195,7 @@ static Result<Vector<String>, int> run_whiptail(WhiptailMode mode, Vector<Whipta
  auto file = Core::File::construct();
  file->open(read_fd, Core::OpenMode::ReadOnly, Core::File::ShouldCloseFileDescriptor::Yes);
  auto data = String::copy(file->read_all());
- return data.split('\n', false);
+ return data.split('\n');
 }
 
 static bool run_system_command(String const& command, StringView command_name)

Tests/AK/TestString.cpp

@@ -233,7 +233,7 @@ TEST_CASE(split)
  EXPECT_EQ(parts[0], "a");
  EXPECT_EQ(parts[1], "b");
 
- parts = test.split(' ', true);
+ parts = test.split(' ', SplitBehavior::KeepEmpty);
  EXPECT_EQ(parts.size(), 5u);
  EXPECT_EQ(parts[0], "a");
  EXPECT_EQ(parts[1], "");
@@ -243,9 +243,9 @@ TEST_CASE(split)
 
  test = "axxbx";
  EXPECT_EQ(test.split('x').size(), 2u);
- EXPECT_EQ(test.split('x', true).size(), 4u);
+ EXPECT_EQ(test.split('x', SplitBehavior::KeepEmpty).size(), 4u);
  EXPECT_EQ(test.split_view('x').size(), 2u);
- EXPECT_EQ(test.split_view('x', true).size(), 4u);
+ EXPECT_EQ(test.split_view('x', SplitBehavior::KeepEmpty).size(), 4u);
 }
 
 TEST_CASE(builder_zero_initial_capacity)