beenull/ladybird
1
0
Fork 0
mirror of https://github.com/LadybirdBrowser/ladybird.git synced 2024-11-25 17:10:23 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 7
ladybird/AK/Format.cpp

698 lines
22 KiB
C++
Raw Blame History

/*
* Copyright (c) 2020, the SerenityOS developers.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* 1. Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* 2. Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
* DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <AK/Format.h>
#include <AK/GenericLexer.h>
#include <AK/String.h>
#include <AK/StringBuilder.h>
#include <ctype.h>
#ifdef KERNEL
# include <Kernel/Process.h>
# include <Kernel/Thread.h>
#else
# include <stdio.h>
# include <unistd.h>
#endif
namespace AK {
namespace {
constexpr size_t use_next_index = NumericLimits<size_t>::max();
// The worst case is that we have the largest 64-bit value formatted as binary number, this would take
// 65 bytes. Choosing a larger power of two won't hurt and is a bit of mitigation against out-of-bounds accesses.
inline size_t convert_unsigned_to_string(u64 value, Array<u8, 128>& buffer, u8 base, bool upper_case)
{
ASSERT(base >= 2 && base <= 16);
static constexpr const char* lowercase_lookup = "0123456789abcdef";
static constexpr const char* uppercase_lookup = "0123456789ABCDEF";
if (value == 0) {
buffer[0] = '0';
return 1;
}
size_t used = 0;
while (value > 0) {
if (upper_case)
buffer[used++] = uppercase_lookup[value % base];
else
buffer[used++] = lowercase_lookup[value % base];
value /= base;
}
for (size_t i = 0; i < used / 2; ++i)
swap(buffer[i], buffer[used - i - 1]);
return used;
}
void vformat_impl(TypeErasedFormatParams& params, FormatBuilder& builder, FormatParser& parser)
{
const auto literal = parser.consume_literal();
builder.put_literal(literal);
FormatParser::FormatSpecifier specifier;
if (!parser.consume_specifier(specifier)) {
ASSERT(parser.is_eof());
return;
}
if (specifier.index == use_next_index)
specifier.index = params.take_next_index();
auto& parameter = params.parameters().at(specifier.index);
FormatParser argparser { specifier.flags };
parameter.formatter(params, builder, argparser, parameter.value);
vformat_impl(params, builder, parser);
}
} // namespace AK::{anonymous}
size_t TypeErasedParameter::to_size() const
{
i64 svalue;
if (type == TypeErasedParameter::Type::UInt8)
svalue = *reinterpret_cast<const u8*>(value);
else if (type == TypeErasedParameter::Type::UInt16)
svalue = *reinterpret_cast<const u16*>(value);
else if (type == TypeErasedParameter::Type::UInt32)
svalue = *reinterpret_cast<const u32*>(value);
else if (type == TypeErasedParameter::Type::UInt64)
svalue = *reinterpret_cast<const u64*>(value);
else if (type == TypeErasedParameter::Type::Int8)
svalue = *reinterpret_cast<const i8*>(value);
else if (type == TypeErasedParameter::Type::Int16)
svalue = *reinterpret_cast<const i16*>(value);
else if (type == TypeErasedParameter::Type::Int32)
svalue = *reinterpret_cast<const i32*>(value);
else if (type == TypeErasedParameter::Type::Int64)
svalue = *reinterpret_cast<const i64*>(value);
else
ASSERT_NOT_REACHED();
ASSERT(svalue >= 0);
return static_cast<size_t>(svalue);
}
FormatParser::FormatParser(StringView input)
: GenericLexer(input)
{
}
StringView FormatParser::consume_literal()
{
const auto begin = tell();
while (!is_eof()) {
if (consume_specific("{{"))
continue;
if (consume_specific("}}"))
continue;
if (next_is(is_any_of("{}")))
return m_input.substring_view(begin, tell() - begin);
consume();
}
return m_input.substring_view(begin);
}
bool FormatParser::consume_number(size_t& value)
{
value = 0;
bool consumed_at_least_one = false;
while (next_is(isdigit)) {
value *= 10;
value += consume() - '0';
consumed_at_least_one = true;
}
return consumed_at_least_one;
}
bool FormatParser::consume_specifier(FormatSpecifier& specifier)
{
ASSERT(!next_is('}'));
if (!consume_specific('{'))
return false;
if (!consume_number(specifier.index))
specifier.index = use_next_index;
if (consume_specific(':')) {
const auto begin = tell();
size_t level = 1;
while (level > 0) {
ASSERT(!is_eof());
if (consume_specific('{')) {
++level;
continue;
}
if (consume_specific('}')) {
--level;
continue;
}
consume();
}
specifier.flags = m_input.substring_view(begin, tell() - begin - 1);
} else {
if (!consume_specific('}'))
ASSERT_NOT_REACHED();
specifier.flags = "";
}
return true;
}
bool FormatParser::consume_replacement_field(size_t& index)
{
if (!consume_specific('{'))
return false;
if (!consume_number(index))
index = use_next_index;
if (!consume_specific('}'))
ASSERT_NOT_REACHED();
return true;
}
void FormatBuilder::put_padding(char fill, size_t amount)
{
for (size_t i = 0; i < amount; ++i)
m_builder.append(fill);
}
void FormatBuilder::put_literal(StringView value)
{
for (size_t i = 0; i < value.length(); ++i) {
m_builder.append(value[i]);
if (value[i] == '{' || value[i] == '}')
++i;
}
}
void FormatBuilder::put_string(
StringView value,
Align align,
size_t min_width,
size_t max_width,
char fill)
{
const auto used_by_string = min(max_width, value.length());
const auto used_by_padding = max(min_width, used_by_string) - used_by_string;
if (used_by_string < value.length())
value = value.substring_view(0, used_by_string);
if (align == Align::Left || align == Align::Default) {
m_builder.append(value);
put_padding(fill, used_by_padding);
} else if (align == Align::Center) {
const auto used_by_left_padding = used_by_padding / 2;
const auto used_by_right_padding = ceil_div<size_t, size_t>(used_by_padding, 2);
put_padding(fill, used_by_left_padding);
m_builder.append(value);
put_padding(fill, used_by_right_padding);
} else if (align == Align::Right) {
put_padding(fill, used_by_padding);
m_builder.append(value);
}
}
void FormatBuilder::put_u64(
u64 value,
u8 base,
bool prefix,
bool upper_case,
bool zero_pad,
Align align,
size_t min_width,
char fill,
SignMode sign_mode,
bool is_negative)
{
if (align == Align::Default)
align = Align::Right;
Array<u8, 128> buffer;
const auto used_by_digits = convert_unsigned_to_string(value, buffer, base, upper_case);
size_t used_by_prefix = 0;
if (align == Align::Right && zero_pad) {
// We want String::formatted("{:#08x}", 32) to produce '0x00000020' instead of '0x000020'. This
// behaviour differs from both fmtlib and printf, but is more intuitive.
used_by_prefix = 0;
} else {
if (is_negative || sign_mode != SignMode::OnlyIfNeeded)
used_by_prefix += 1;
if (prefix) {
if (base == 8)
used_by_prefix += 1;
else if (base == 16)
used_by_prefix += 2;
else if (base == 2)
used_by_prefix += 2;
}
}
const auto used_by_field = used_by_prefix + used_by_digits;
const auto used_by_padding = max(used_by_field, min_width) - used_by_field;
const auto put_prefix = [&]() {
if (is_negative)
m_builder.append('-');
else if (sign_mode == SignMode::Always)
m_builder.append('+');
else if (sign_mode == SignMode::Reserved)
m_builder.append(' ');
if (prefix) {
if (base == 2) {
if (upper_case)
m_builder.append("0B");
else
m_builder.append("0b");
} else if (base == 8) {
m_builder.append("0");
} else if (base == 16) {
if (upper_case)
m_builder.append("0X");
else
m_builder.append("0x");
}
}
};
const auto put_digits = [&]() {
for (size_t i = 0; i < used_by_digits; ++i)
m_builder.append(buffer[i]);
};
if (align == Align::Left) {
const auto used_by_right_padding = used_by_padding;
put_prefix();
put_digits();
put_padding(fill, used_by_right_padding);
} else if (align == Align::Center) {
const auto used_by_left_padding = used_by_padding / 2;
const auto used_by_right_padding = ceil_div<size_t, size_t>(used_by_padding, 2);
put_padding(fill, used_by_left_padding);
put_prefix();
put_digits();
put_padding(fill, used_by_right_padding);
} else if (align == Align::Right) {
const auto used_by_left_padding = used_by_padding;
if (zero_pad) {
put_prefix();
put_padding('0', used_by_left_padding);
put_digits();
} else {
put_padding(fill, used_by_left_padding);
put_prefix();
put_digits();
}
}
}
void FormatBuilder::put_i64(
i64 value,
u8 base,
bool prefix,
bool upper_case,
bool zero_pad,
Align align,
size_t min_width,
char fill,
SignMode sign_mode)
{
const auto is_negative = value < 0;
value = is_negative ? -value : value;
put_u64(static_cast<size_t>(value), base, prefix, upper_case, zero_pad, align, min_width, fill, sign_mode, is_negative);
}
#ifndef KERNEL
void FormatBuilder::put_f64(
double value,
u8 base,
bool upper_case,
Align align,
size_t min_width,
size_t precision,
char fill,
SignMode sign_mode)
{
StringBuilder string_builder;
FormatBuilder format_builder { string_builder };
format_builder.put_i64(static_cast<i64>(value), base, false, upper_case, false, Align::Right, 0, ' ', sign_mode);
string_builder.append('.');
if (precision > 0) {
// FIXME: This is a terrible approximation but doing it properly would be a lot of work. If someone is up for that, a good
// place to start would be the following video from CppCon 2019:
// https://youtu.be/4P_kbF0EbZM (Stephan T. Lavavej “Floating-Point <charconv>: Making Your Code 10x Faster With C++17's Final Boss”)
value -= static_cast<i64>(value);
if (value < 0)
value = -value;
for (u32 i = 0; i < precision; ++i)
value *= 10;
format_builder.put_u64(static_cast<u64>(value), base, false, upper_case, true, Align::Right, precision);
// FIXME: Cut off trailing zeroes by default?
}
put_string(string_builder.string_view(), align, min_width, NumericLimits<size_t>::max(), fill);
}
#endif
void vformat(StringBuilder& builder, StringView fmtstr, TypeErasedFormatParams params)
{
FormatBuilder fmtbuilder { builder };
FormatParser parser { fmtstr };
vformat_impl(params, fmtbuilder, parser);
}
void vformat(const LogStream& stream, StringView fmtstr, TypeErasedFormatParams params)
{
StringBuilder builder;
vformat(builder, fmtstr, params);
stream << builder.to_string();
}
void StandardFormatter::parse(TypeErasedFormatParams& params, FormatParser& parser)
{
if (StringView { "<^>" }.contains(parser.peek(1))) {
ASSERT(!parser.next_is(is_any_of("{}")));
m_fill = parser.consume();
}
if (parser.consume_specific('<'))
m_align = FormatBuilder::Align::Left;
else if (parser.consume_specific('^'))
m_align = FormatBuilder::Align::Center;
else if (parser.consume_specific('>'))
m_align = FormatBuilder::Align::Right;
if (parser.consume_specific('-'))
m_sign_mode = FormatBuilder::SignMode::OnlyIfNeeded;
else if (parser.consume_specific('+'))
m_sign_mode = FormatBuilder::SignMode::Always;
else if (parser.consume_specific(' '))
m_sign_mode = FormatBuilder::SignMode::Reserved;
if (parser.consume_specific('#'))
m_alternative_form = true;
if (parser.consume_specific('0'))
m_zero_pad = true;
if (size_t index = 0; parser.consume_replacement_field(index)) {
if (index == use_next_index)
index = params.take_next_index();
m_width = params.parameters().at(index).to_size();
} else if (size_t width = 0; parser.consume_number(width)) {
m_width = width;
}
if (parser.consume_specific('.')) {
if (size_t index = 0; parser.consume_replacement_field(index)) {
if (index == use_next_index)
index = params.take_next_index();
m_precision = params.parameters().at(index).to_size();
} else if (size_t precision = 0; parser.consume_number(precision)) {
m_precision = precision;
}
}
if (parser.consume_specific('b'))
m_mode = Mode::Binary;
else if (parser.consume_specific('B'))
m_mode = Mode::BinaryUppercase;
else if (parser.consume_specific('d'))
m_mode = Mode::Decimal;
else if (parser.consume_specific('o'))
m_mode = Mode::Octal;
else if (parser.consume_specific('x'))
m_mode = Mode::Hexadecimal;
else if (parser.consume_specific('X'))
m_mode = Mode::HexadecimalUppercase;
else if (parser.consume_specific('c'))
m_mode = Mode::Character;
else if (parser.consume_specific('s'))
m_mode = Mode::String;
else if (parser.consume_specific('p'))
m_mode = Mode::Pointer;
else if (parser.consume_specific('f'))
m_mode = Mode::Float;
else if (parser.consume_specific('a'))
m_mode = Mode::Hexfloat;
else if (parser.consume_specific('A'))
m_mode = Mode::HexfloatUppercase;
if (!parser.is_eof())
dbgln("{} did not consume '{}'", __PRETTY_FUNCTION__, parser.remaining());
ASSERT(parser.is_eof());
}
void Formatter<StringView>::format(FormatBuilder& builder, StringView value)
{
if (m_sign_mode != FormatBuilder::SignMode::Default)
ASSERT_NOT_REACHED();
if (m_alternative_form)
ASSERT_NOT_REACHED();
if (m_zero_pad)
ASSERT_NOT_REACHED();
if (m_mode != Mode::Default && m_mode != Mode::String && m_mode != Mode::Character)
ASSERT_NOT_REACHED();
if (m_width.has_value() && m_precision.has_value())
ASSERT_NOT_REACHED();
m_width = m_width.value_or(0);
m_precision = m_precision.value_or(NumericLimits<size_t>::max());
builder.put_string(value, m_align, m_width.value(), m_precision.value(), m_fill);
}
template<typename T>
void Formatter<T, typename EnableIf<IsIntegral<T>::value>::Type>::format(FormatBuilder& builder, T value)
{
if (m_mode == Mode::Character) {
// FIXME: We just support ASCII for now, in the future maybe unicode?
ASSERT(value >= 0 && value <= 127);
m_mode = Mode::String;
Formatter<StringView> formatter { *this };
return formatter.format(builder, StringView { reinterpret_cast<const char*>(&value), 1 });
}
if (m_precision.has_value())
ASSERT_NOT_REACHED();
if (m_mode == Mode::Pointer) {
if (m_sign_mode != FormatBuilder::SignMode::Default)
ASSERT_NOT_REACHED();
if (m_align != FormatBuilder::Align::Default)
ASSERT_NOT_REACHED();
if (m_alternative_form)
ASSERT_NOT_REACHED();
if (m_width.has_value())
ASSERT_NOT_REACHED();
m_mode = Mode::Hexadecimal;
m_alternative_form = true;
m_width = 2 * sizeof(void*);
m_zero_pad = true;
}
u8 base = 0;
bool upper_case = false;
if (m_mode == Mode::Binary) {
base = 2;
} else if (m_mode == Mode::BinaryUppercase) {
base = 2;
upper_case = true;
} else if (m_mode == Mode::Octal) {
base = 8;
} else if (m_mode == Mode::Decimal || m_mode == Mode::Default) {
base = 10;
} else if (m_mode == Mode::Hexadecimal) {
base = 16;
} else if (m_mode == Mode::HexadecimalUppercase) {
base = 16;
upper_case = true;
} else {
ASSERT_NOT_REACHED();
}
m_width = m_width.value_or(0);
if (IsSame<typename MakeUnsigned<T>::Type, T>::value)
builder.put_u64(value, base, m_alternative_form, upper_case, m_zero_pad, m_align, m_width.value(), m_fill, m_sign_mode);
else
builder.put_i64(value, base, m_alternative_form, upper_case, m_zero_pad, m_align, m_width.value(), m_fill, m_sign_mode);
}
void Formatter<char>::format(FormatBuilder& builder, char value)
{
if (m_mode == Mode::Binary || m_mode == Mode::BinaryUppercase || m_mode == Mode::Decimal || m_mode == Mode::Octal || m_mode == Mode::Hexadecimal || m_mode == Mode::HexadecimalUppercase) {
// Trick: signed char != char. (Sometimes weird features are actually helpful.)
Formatter<signed char> formatter { *this };
return formatter.format(builder, static_cast<signed char>(value));
} else {
Formatter<StringView> formatter { *this };
return formatter.format(builder, { &value, 1 });
}
}
void Formatter<bool>::format(FormatBuilder& builder, bool value)
{
if (m_mode == Mode::Binary || m_mode == Mode::BinaryUppercase || m_mode == Mode::Decimal || m_mode == Mode::Octal || m_mode == Mode::Hexadecimal || m_mode == Mode::HexadecimalUppercase) {
Formatter<u8> formatter { *this };
return formatter.format(builder, static_cast<u8>(value));
} else {
Formatter<StringView> formatter { *this };
return formatter.format(builder, value ? "true" : "false");
}
}
#ifndef KERNEL
void Formatter<double>::format(FormatBuilder& builder, double value)
{
u8 base;
bool upper_case;
if (m_mode == Mode::Default || m_mode == Mode::Float) {
base = 10;
upper_case = false;
} else if (m_mode == Mode::Hexfloat) {
base = 16;
upper_case = false;
} else if (m_mode == Mode::HexfloatUppercase) {
base = 16;
upper_case = true;
} else {
ASSERT_NOT_REACHED();
}
m_width = m_width.value_or(0);
m_precision = m_precision.value_or(6);
builder.put_f64(value, base, upper_case, m_align, m_width.value(), m_precision.value(), m_fill, m_sign_mode);
}
void Formatter<float>::format(FormatBuilder& builder, float value)
{
Formatter<double> formatter { *this };
formatter.format(builder, value);
}
#endif
#ifndef KERNEL
void vout(FILE* file, StringView fmtstr, TypeErasedFormatParams params, bool newline)
{
StringBuilder builder;
vformat(builder, fmtstr, params);
if (newline)
builder.append('\n');
const auto string = builder.string_view();
const auto retval = ::fwrite(string.characters_without_null_termination(), 1, string.length(), file);
ASSERT(static_cast<size_t>(retval) == string.length());
}
#endif
void vdbgln(StringView fmtstr, TypeErasedFormatParams params)
{
StringBuilder builder;
// FIXME: This logic is redundant with the stuff in LogStream.cpp.
#if defined(__serenity__)
# ifdef KERNEL
if (Kernel::Processor::is_initialized() && Kernel::Thread::current()) {
auto& thread = *Kernel::Thread::current();
builder.appendff("\033[34;1m[{}({}:{})]\033[0m: ", thread.process().name(), thread.pid().value(), thread.tid().value());
} else {
builder.appendff("\033[34;1m[Kernel]\033[0m: ");
}
# else
static TriState got_process_name = TriState::Unknown;
static char process_name_buffer[256];
if (got_process_name == TriState::Unknown) {
if (get_process_name(process_name_buffer, sizeof(process_name_buffer)) == 0)
got_process_name = TriState::True;
else
got_process_name = TriState::False;
}
if (got_process_name == TriState::True)
builder.appendff("\033[33;1m{}({})\033[0m: ", process_name_buffer, getpid());
# endif
#endif
vformat(builder, fmtstr, params);
builder.append('\n');
const auto string = builder.string_view();
const auto retval = dbgputstr(string.characters_without_null_termination(), string.length());
ASSERT(retval == 0);
}
template struct Formatter<unsigned char, void>;
template struct Formatter<unsigned short, void>;
template struct Formatter<unsigned int, void>;
template struct Formatter<unsigned long, void>;
template struct Formatter<unsigned long long, void>;
template struct Formatter<short, void>;
template struct Formatter<int, void>;
template struct Formatter<long, void>;
template struct Formatter<long long, void>;
template struct Formatter<signed char, void>;
} // namespace AK
Reference in a new issue View git blame Copy permalink