ladybird/Userland/Libraries/LibGfx/TextLayout.cpp
Timothy Flynn d9349f1510 LibGfx: Treat "\r\n" as a single line break during text layout
We currently insert two line breaks for this sequence. This Windows-
style of line breaking can be seen in documents on the web.
2024-03-08 14:43:33 -05:00

262 lines
8.4 KiB
C++

/*
* Copyright (c) 2018-2020, Andreas Kling <kling@serenityos.org>
* Copyright (c) 2021, sin-ack <sin-ack@protonmail.com>
*
* SPDX-License-Identifier: BSD-2-Clause
*/
#include "TextLayout.h"
#include "Font/Emoji.h"
#include <AK/Debug.h>
#include <LibUnicode/CharacterTypes.h>
#include <LibUnicode/Emoji.h>
namespace Gfx {
enum class BlockType {
Newline,
Whitespace,
Word
};
struct Block {
BlockType type;
Utf8View characters;
};
FloatRect TextLayout::bounding_rect(TextWrapping wrapping) const
{
auto lines = wrap_lines(TextElision::None, wrapping);
if (lines.is_empty()) {
return {};
}
FloatRect bounding_rect = {
0, 0, 0, (static_cast<float>(lines.size()) * (m_font_metrics.ascent + m_font_metrics.descent + m_font_metrics.line_gap)) - m_font_metrics.line_gap
};
for (auto& line : lines) {
auto line_width = m_font.width(line);
if (line_width > bounding_rect.width())
bounding_rect.set_width(line_width);
}
return bounding_rect;
}
Vector<ByteString, 32> TextLayout::wrap_lines(TextElision elision, TextWrapping wrapping) const
{
Vector<Block> blocks;
Optional<BlockType> current_block_type;
size_t block_start_offset = 0;
size_t offset = 0;
for (auto it = m_text.begin(); !it.done(); ++it) {
offset = m_text.iterator_offset(it);
switch (*it) {
case '\t':
case ' ': {
if (current_block_type.has_value() && current_block_type.value() != BlockType::Whitespace) {
blocks.append({
current_block_type.value(),
m_text.substring_view(block_start_offset, offset - block_start_offset),
});
current_block_type.clear();
}
if (!current_block_type.has_value()) {
current_block_type = BlockType::Whitespace;
block_start_offset = offset;
}
continue;
}
case '\r':
if (it.peek(1) == static_cast<u32>('\n'))
++it;
[[fallthrough]];
case '\n': {
if (current_block_type.has_value()) {
blocks.append({
current_block_type.value(),
m_text.substring_view(block_start_offset, offset - block_start_offset),
});
current_block_type.clear();
}
blocks.append({ BlockType::Newline, Utf8View {} });
continue;
}
default: {
if (current_block_type.has_value() && current_block_type.value() != BlockType::Word) {
blocks.append({
current_block_type.value(),
m_text.substring_view(block_start_offset, offset - block_start_offset),
});
current_block_type.clear();
}
if (!current_block_type.has_value()) {
current_block_type = BlockType::Word;
block_start_offset = offset;
}
}
}
}
if (current_block_type.has_value()) {
blocks.append({
current_block_type.value(),
m_text.substring_view(block_start_offset, m_text.byte_length() - block_start_offset),
});
}
Vector<ByteString> lines;
StringBuilder builder;
float line_width = 0;
size_t current_block = 0;
for (Block& block : blocks) {
switch (block.type) {
case BlockType::Newline: {
lines.append(builder.to_byte_string());
builder.clear();
line_width = 0;
current_block++;
continue;
}
case BlockType::Whitespace:
case BlockType::Word: {
float block_width = m_font.width(block.characters);
// FIXME: This should look at the specific advance amount of the
// last character, but we don't support that yet.
if (current_block != blocks.size() - 1) {
block_width += m_font.glyph_spacing();
}
if (wrapping == TextWrapping::Wrap && line_width + block_width > m_rect.width()) {
lines.append(builder.to_byte_string());
builder.clear();
line_width = 0;
}
builder.append(block.characters.as_string());
line_width += block_width;
current_block++;
}
}
}
auto last_line = builder.to_byte_string();
if (!last_line.is_empty())
lines.append(last_line);
switch (elision) {
case TextElision::None:
break;
case TextElision::Right: {
lines.at(lines.size() - 1) = elide_text_from_right(Utf8View { lines.at(lines.size() - 1) });
break;
}
}
return lines;
}
ByteString TextLayout::elide_text_from_right(Utf8View text) const
{
float text_width = m_font.width(text);
if (text_width > static_cast<float>(m_rect.width())) {
float ellipsis_width = m_font.width("..."sv);
float current_width = ellipsis_width;
size_t glyph_spacing = m_font.glyph_spacing();
// FIXME: This code will break when the font has glyphs with advance
// amounts different from the actual width of the glyph
// (which is the case with many TrueType fonts).
if (ellipsis_width < text_width) {
size_t offset = 0;
for (auto it = text.begin(); !it.done(); ++it) {
auto glyph_width = m_font.glyph_or_emoji_width(it);
// NOTE: Glyph spacing should not be added after the last glyph on the line,
// but since we are here because the last glyph does not actually fit on the line,
// we don't have to worry about spacing.
auto width_with_this_glyph_included = current_width + glyph_width + glyph_spacing;
if (width_with_this_glyph_included > m_rect.width())
break;
current_width += glyph_width + glyph_spacing;
offset = text.iterator_offset(it);
}
StringBuilder builder;
builder.append(text.substring_view(0, offset).as_string());
builder.append("..."sv);
return builder.to_byte_string();
}
}
return text.as_string();
}
DrawGlyphOrEmoji prepare_draw_glyph_or_emoji(FloatPoint point, Utf8CodePointIterator& it, Font const& font)
{
u32 code_point = *it;
auto next_code_point = it.peek(1);
ScopeGuard consume_variation_selector = [&, initial_it = it] {
static auto const variation_selector = Unicode::property_from_string("Variation_Selector"sv);
if (!variation_selector.has_value())
return;
// If we advanced the iterator to consume an emoji sequence, don't look for another variation selector.
if (initial_it != it)
return;
// Otherwise, discard one code point if it's a variation selector.
if (next_code_point.has_value() && Unicode::code_point_has_property(*next_code_point, *variation_selector))
++it;
};
// NOTE: We don't check for emoji
auto font_contains_glyph = font.contains_glyph(code_point);
auto check_for_emoji = !font.has_color_bitmaps() && Unicode::could_be_start_of_emoji_sequence(it, font_contains_glyph ? Unicode::SequenceType::EmojiPresentation : Unicode::SequenceType::Any);
// If the font contains the glyph, and we know it's not the start of an emoji, draw a text glyph.
if (font_contains_glyph && !check_for_emoji) {
return DrawGlyph {
.position = point,
.code_point = code_point,
.font = font,
};
}
// If we didn't find a text glyph, or have an emoji variation selector or regional indicator, try to draw an emoji glyph.
if (auto const* emoji = Emoji::emoji_for_code_point_iterator(it)) {
return DrawEmoji {
.position = point,
.emoji = emoji,
.font = font,
};
}
// If that failed, but we have a text glyph fallback, draw that.
if (font_contains_glyph) {
return DrawGlyph {
.position = point,
.code_point = code_point,
.font = font,
};
}
// No suitable glyph found, draw a replacement character.
dbgln_if(EMOJI_DEBUG, "Failed to find a glyph or emoji for code_point {}", code_point);
return DrawGlyph {
.position = point,
.code_point = 0xFFFD,
.font = font,
};
}
}