diff options
Diffstat (limited to 'src/cguittfont')
| -rw-r--r-- | src/cguittfont/CGUITTFont.cpp | 1146 | ||||
| -rw-r--r-- | src/cguittfont/CGUITTFont.h | 377 | ||||
| -rw-r--r-- | src/cguittfont/CMakeLists.txt | 17 | ||||
| -rw-r--r-- | src/cguittfont/irrUString.h | 3877 | ||||
| -rw-r--r-- | src/cguittfont/xCGUITTFont.cpp | 5 | ||||
| -rw-r--r-- | src/cguittfont/xCGUITTFont.h | 7 |
6 files changed, 5429 insertions, 0 deletions
diff --git a/src/cguittfont/CGUITTFont.cpp b/src/cguittfont/CGUITTFont.cpp new file mode 100644 index 000000000..fb8199e21 --- /dev/null +++ b/src/cguittfont/CGUITTFont.cpp @@ -0,0 +1,1146 @@ +/* + CGUITTFont FreeType class for Irrlicht + Copyright (c) 2009-2010 John Norman + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any + damages arising from the use of this software. + + Permission is granted to anyone to use this software for any + purpose, including commercial applications, and to alter it and + redistribute it freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you + must not claim that you wrote the original software. If you use + this software in a product, an acknowledgment in the product + documentation would be appreciated but is not required. + + 2. Altered source versions must be plainly marked as such, and + must not be misrepresented as being the original software. + + 3. This notice may not be removed or altered from any source + distribution. + + The original version of this class can be located at: + http://irrlicht.suckerfreegames.com/ + + John Norman + john@suckerfreegames.com +*/ + +#include <irrlicht.h> +#include "CGUITTFont.h" + +namespace irr +{ +namespace gui +{ + +// Manages the FT_Face cache. +struct SGUITTFace : public virtual irr::IReferenceCounted +{ + SGUITTFace() : face_buffer(0), face_buffer_size(0) + { + memset((void*)&face, 0, sizeof(FT_Face)); + } + + ~SGUITTFace() + { + FT_Done_Face(face); + delete[] face_buffer; + } + + FT_Face face; + FT_Byte* face_buffer; + FT_Long face_buffer_size; +}; + +// Static variables. +FT_Library CGUITTFont::c_library; +core::map<io::path, SGUITTFace*> CGUITTFont::c_faces; +bool CGUITTFont::c_libraryLoaded = false; +scene::IMesh* CGUITTFont::shared_plane_ptr_ = 0; +scene::SMesh CGUITTFont::shared_plane_; + +// + +video::IImage* SGUITTGlyph::createGlyphImage(const FT_Bitmap& bits, video::IVideoDriver* driver) const +{ + // Determine what our texture size should be. + // Add 1 because textures are inclusive-exclusive. + core::dimension2du d(bits.width + 1, bits.rows + 1); + core::dimension2du texture_size; + //core::dimension2du texture_size(bits.width + 1, bits.rows + 1); + + // Create and load our image now. + video::IImage* image = 0; + switch (bits.pixel_mode) + { + case FT_PIXEL_MODE_MONO: + { + // Create a blank image and fill it with transparent pixels. + texture_size = d.getOptimalSize(true, true); + image = driver->createImage(video::ECF_A1R5G5B5, texture_size); + image->fill(video::SColor(0, 255, 255, 255)); + + // Load the monochrome data in. + const u32 image_pitch = image->getPitch() / sizeof(u16); + u16* image_data = (u16*)image->lock(); + u8* glyph_data = bits.buffer; + for (s32 y = 0; y < bits.rows; ++y) + { + u16* row = image_data; + for (s32 x = 0; x < bits.width; ++x) + { + // Monochrome bitmaps store 8 pixels per byte. The left-most pixel is the bit 0x80. + // So, we go through the data each bit at a time. + if ((glyph_data[y * bits.pitch + (x / 8)] & (0x80 >> (x % 8))) != 0) + *row = 0xFFFF; + ++row; + } + image_data += image_pitch; + } + image->unlock(); + break; + } + + case FT_PIXEL_MODE_GRAY: + { + // Create our blank image. + texture_size = d.getOptimalSize(!driver->queryFeature(video::EVDF_TEXTURE_NPOT), !driver->queryFeature(video::EVDF_TEXTURE_NSQUARE), true, 0); + image = driver->createImage(video::ECF_A8R8G8B8, texture_size); + image->fill(video::SColor(0, 255, 255, 255)); + + // Load the grayscale data in. + const float gray_count = static_cast<float>(bits.num_grays); + const u32 image_pitch = image->getPitch() / sizeof(u32); + u32* image_data = (u32*)image->lock(); + u8* glyph_data = bits.buffer; + for (s32 y = 0; y < bits.rows; ++y) + { + u8* row = glyph_data; + for (s32 x = 0; x < bits.width; ++x) + { + image_data[y * image_pitch + x] |= static_cast<u32>(255.0f * (static_cast<float>(*row++) / gray_count)) << 24; + //data[y * image_pitch + x] |= ((u32)(*bitsdata++) << 24); + } + glyph_data += bits.pitch; + } + image->unlock(); + break; + } + default: + // TODO: error message? + return 0; + } + return image; +} + +void SGUITTGlyph::preload(u32 char_index, FT_Face face, video::IVideoDriver* driver, u32 font_size, const FT_Int32 loadFlags) +{ + if (isLoaded) return; + + // Set the size of the glyph. + FT_Set_Pixel_Sizes(face, 0, font_size); + + // Attempt to load the glyph. + if (FT_Load_Glyph(face, char_index, loadFlags) != FT_Err_Ok) + // TODO: error message? + return; + + FT_GlyphSlot glyph = face->glyph; + FT_Bitmap bits = glyph->bitmap; + + // Setup the glyph information here: + advance = glyph->advance; + offset = core::vector2di(glyph->bitmap_left, glyph->bitmap_top); + + // Try to get the last page with available slots. + CGUITTGlyphPage* page = parent->getLastGlyphPage(); + + // If we need to make a new page, do that now. + if (!page) + { + page = parent->createGlyphPage(bits.pixel_mode); + if (!page) + // TODO: add error message? + return; + } + + glyph_page = parent->getLastGlyphPageIndex(); + u32 texture_side_length = page->texture->getOriginalSize().Width; + core::vector2di page_position( + (page->used_slots % (texture_side_length / font_size)) * font_size, + (page->used_slots / (texture_side_length / font_size)) * font_size + ); + source_rect.UpperLeftCorner = page_position; + source_rect.LowerRightCorner = core::vector2di(page_position.X + bits.width, page_position.Y + bits.rows); + + page->dirty = true; + ++page->used_slots; + --page->available_slots; + + // We grab the glyph bitmap here so the data won't be removed when the next glyph is loaded. + surface = createGlyphImage(bits, driver); + + // Set our glyph as loaded. + isLoaded = true; +} + +void SGUITTGlyph::unload() +{ + if (surface) + { + surface->drop(); + surface = 0; + } + isLoaded = false; +} + +////////////////////// + +CGUITTFont* CGUITTFont::createTTFont(IGUIEnvironment *env, const io::path& filename, const u32 size, const bool antialias, const bool transparency) +{ + if (!c_libraryLoaded) + { + if (FT_Init_FreeType(&c_library)) + return 0; + c_libraryLoaded = true; + } + + CGUITTFont* font = new CGUITTFont(env); + bool ret = font->load(filename, size, antialias, transparency); + if (!ret) + { + font->drop(); + return 0; + } + + return font; +} + +CGUITTFont* CGUITTFont::createTTFont(IrrlichtDevice *device, const io::path& filename, const u32 size, const bool antialias, const bool transparency) +{ + if (!c_libraryLoaded) + { + if (FT_Init_FreeType(&c_library)) + return 0; + c_libraryLoaded = true; + } + + CGUITTFont* font = new CGUITTFont(device->getGUIEnvironment()); + font->Device = device; + bool ret = font->load(filename, size, antialias, transparency); + if (!ret) + { + font->drop(); + return 0; + } + + return font; +} + +CGUITTFont* CGUITTFont::create(IGUIEnvironment *env, const io::path& filename, const u32 size, const bool antialias, const bool transparency) +{ + return CGUITTFont::createTTFont(env, filename, size, antialias, transparency); +} + +CGUITTFont* CGUITTFont::create(IrrlichtDevice *device, const io::path& filename, const u32 size, const bool antialias, const bool transparency) +{ + return CGUITTFont::createTTFont(device, filename, size, antialias, transparency); +} + +////////////////////// + +//! Constructor. +CGUITTFont::CGUITTFont(IGUIEnvironment *env) +: use_monochrome(false), use_transparency(true), use_hinting(true), use_auto_hinting(true), +batch_load_size(1), Device(0), Environment(env), Driver(0), GlobalKerningWidth(0), GlobalKerningHeight(0) +{ + #ifdef _DEBUG + setDebugName("CGUITTFont"); + #endif + + if (Environment) + { + // don't grab environment, to avoid circular references + Driver = Environment->getVideoDriver(); + } + + if (Driver) + Driver->grab(); + + setInvisibleCharacters(L" "); + + // Glyphs aren't reference counted, so don't try to delete them when we free the array. + Glyphs.set_free_when_destroyed(false); +} + +bool CGUITTFont::load(const io::path& filename, const u32 size, const bool antialias, const bool transparency) +{ + // Some sanity checks. + if (Environment == 0 || Driver == 0) return false; + if (size == 0) return false; + if (filename.size() == 0) return false; + + io::IFileSystem* filesystem = Environment->getFileSystem(); + irr::ILogger* logger = (Device != 0 ? Device->getLogger() : 0); + this->size = size; + this->filename = filename; + + // Update the font loading flags when the font is first loaded. + this->use_monochrome = !antialias; + this->use_transparency = transparency; + update_load_flags(); + + // Log. + if (logger) + logger->log(L"CGUITTFont", core::stringw(core::stringw(L"Creating new font: ") + core::ustring(filename).toWCHAR_s() + L" " + core::stringc(size) + L"pt " + (antialias ? L"+antialias " : L"-antialias ") + (transparency ? L"+transparency" : L"-transparency")).c_str(), irr::ELL_INFORMATION); + + // Grab the face. + SGUITTFace* face = 0; + core::map<io::path, SGUITTFace*>::Node* node = c_faces.find(filename); + if (node == 0) + { + face = new SGUITTFace(); + c_faces.set(filename, face); + + if (filesystem) + { + // Read in the file data. + io::IReadFile* file = filesystem->createAndOpenFile(filename); + if (file == 0) + { + if (logger) logger->log(L"CGUITTFont", L"Failed to open the file.", irr::ELL_INFORMATION); + + c_faces.remove(filename); + delete face; + face = 0; + return false; + } + face->face_buffer = new FT_Byte[file->getSize()]; + file->read(face->face_buffer, file->getSize()); + face->face_buffer_size = file->getSize(); + file->drop(); + + // Create the face. + if (FT_New_Memory_Face(c_library, face->face_buffer, face->face_buffer_size, 0, &face->face)) + { + if (logger) logger->log(L"CGUITTFont", L"FT_New_Memory_Face failed.", irr::ELL_INFORMATION); + + c_faces.remove(filename); + delete face; + face = 0; + return false; + } + } + else + { + core::ustring converter(filename); + if (FT_New_Face(c_library, reinterpret_cast<const char*>(converter.toUTF8_s().c_str()), 0, &face->face)) + { + if (logger) logger->log(L"CGUITTFont", L"FT_New_Face failed.", irr::ELL_INFORMATION); + + c_faces.remove(filename); + delete face; + face = 0; + return false; + } + } + } + else + { + // Using another instance of this face. + face = node->getValue(); + face->grab(); + } + + // Store our face. + tt_face = face->face; + + // Store font metrics. + FT_Set_Pixel_Sizes(tt_face, size, 0); + font_metrics = tt_face->size->metrics; + + // Allocate our glyphs. + Glyphs.clear(); + Glyphs.reallocate(tt_face->num_glyphs); + Glyphs.set_used(tt_face->num_glyphs); + for (FT_Long i = 0; i < tt_face->num_glyphs; ++i) + { + Glyphs[i].isLoaded = false; + Glyphs[i].glyph_page = 0; + Glyphs[i].source_rect = core::recti(); + Glyphs[i].offset = core::vector2di(); + Glyphs[i].advance = FT_Vector(); + Glyphs[i].surface = 0; + Glyphs[i].parent = this; + } + + // Cache the first 127 ascii characters. + u32 old_size = batch_load_size; + batch_load_size = 127; + getGlyphIndexByChar((uchar32_t)0); + batch_load_size = old_size; + + return true; +} + +CGUITTFont::~CGUITTFont() +{ + // Delete the glyphs and glyph pages. + reset_images(); + CGUITTAssistDelete::Delete(Glyphs); + //Glyphs.clear(); + + // We aren't using this face anymore. + core::map<io::path, SGUITTFace*>::Node* n = c_faces.find(filename); + if (n) + { + SGUITTFace* f = n->getValue(); + + // Drop our face. If this was the last face, the destructor will clean up. + if (f->drop()) + c_faces.remove(filename); + + // If there are no more faces referenced by FreeType, clean up. + if (c_faces.size() == 0) + { + FT_Done_FreeType(c_library); + c_libraryLoaded = false; + } + } + + // Drop our driver now. + if (Driver) + Driver->drop(); +} + +void CGUITTFont::reset_images() +{ + // Delete the glyphs. + for (u32 i = 0; i != Glyphs.size(); ++i) + Glyphs[i].unload(); + + // Unload the glyph pages from video memory. + for (u32 i = 0; i != Glyph_Pages.size(); ++i) + delete Glyph_Pages[i]; + Glyph_Pages.clear(); + + // Always update the internal FreeType loading flags after resetting. + update_load_flags(); +} + +void CGUITTFont::update_glyph_pages() const +{ + for (u32 i = 0; i != Glyph_Pages.size(); ++i) + { + if (Glyph_Pages[i]->dirty) + Glyph_Pages[i]->updateTexture(); + } +} + +CGUITTGlyphPage* CGUITTFont::getLastGlyphPage() const +{ + CGUITTGlyphPage* page = 0; + if (Glyph_Pages.empty()) + return 0; + else + { + page = Glyph_Pages[getLastGlyphPageIndex()]; + if (page->available_slots == 0) + page = 0; + } + return page; +} + +CGUITTGlyphPage* CGUITTFont::createGlyphPage(const u8& pixel_mode) +{ + CGUITTGlyphPage* page = 0; + + // Name of our page. + io::path name("TTFontGlyphPage_"); + name += tt_face->family_name; + name += "."; + name += tt_face->style_name; + name += "."; + name += size; + name += "_"; + name += Glyph_Pages.size(); // The newly created page will be at the end of the collection. + + // Create the new page. + page = new CGUITTGlyphPage(Driver, name); + + // Determine our maximum texture size. + // If we keep getting 0, set it to 1024x1024, as that number is pretty safe. + core::dimension2du max_texture_size = max_page_texture_size; + if (max_texture_size.Width == 0 || max_texture_size.Height == 0) + max_texture_size = Driver->getMaxTextureSize(); + if (max_texture_size.Width == 0 || max_texture_size.Height == 0) + max_texture_size = core::dimension2du(1024, 1024); + + // We want to try to put at least 144 glyphs on a single texture. + core::dimension2du page_texture_size; + if (size <= 21) page_texture_size = core::dimension2du(256, 256); + else if (size <= 42) page_texture_size = core::dimension2du(512, 512); + else if (size <= 84) page_texture_size = core::dimension2du(1024, 1024); + else if (size <= 168) page_texture_size = core::dimension2du(2048, 2048); + else page_texture_size = core::dimension2du(4096, 4096); + + if (page_texture_size.Width > max_texture_size.Width || page_texture_size.Height > max_texture_size.Height) + page_texture_size = max_texture_size; + + if (!page->createPageTexture(pixel_mode, page_texture_size)) + // TODO: add error message? + return 0; + + if (page) + { + // Determine the number of glyph slots on the page and add it to the list of pages. + page->available_slots = (page_texture_size.Width / size) * (page_texture_size.Height / size); + Glyph_Pages.push_back(page); + } + return page; +} + +void CGUITTFont::setTransparency(const bool flag) +{ + use_transparency = flag; + reset_images(); +} + +void CGUITTFont::setMonochrome(const bool flag) +{ + use_monochrome = flag; + reset_images(); +} + +void CGUITTFont::setFontHinting(const bool enable, const bool enable_auto_hinting) +{ + use_hinting = enable; + use_auto_hinting = enable_auto_hinting; + reset_images(); +} + +void CGUITTFont::draw(const core::stringw& text, const core::rect<s32>& position, video::SColor color, bool hcenter, bool vcenter, const core::rect<s32>* clip) +{ + if (!Driver) + return; + + // Clear the glyph pages of their render information. + for (u32 i = 0; i < Glyph_Pages.size(); ++i) + { + Glyph_Pages[i]->render_positions.clear(); + Glyph_Pages[i]->render_source_rects.clear(); + } + + // Set up some variables. + core::dimension2d<s32> textDimension; + core::position2d<s32> offset = position.UpperLeftCorner; + + // Determine offset positions. + if (hcenter || vcenter) + { + textDimension = getDimension(text.c_str()); + + if (hcenter) + offset.X = ((position.getWidth() - textDimension.Width) >> 1) + offset.X; + + if (vcenter) + offset.Y = ((position.getHeight() - textDimension.Height) >> 1) + offset.Y; + } + + // Convert to a unicode string. + core::ustring utext(text); + + // Set up our render map. + core::map<u32, CGUITTGlyphPage*> Render_Map; + + // Start parsing characters. + u32 n; + uchar32_t previousChar = 0; + core::ustring::const_iterator iter(utext); + while (!iter.atEnd()) + { + uchar32_t currentChar = *iter; + n = getGlyphIndexByChar(currentChar); + bool visible = (Invisible.findFirst(currentChar) == -1); + if (n > 0 && visible) + { + bool lineBreak=false; + if (currentChar == L'\r') // Mac or Windows breaks + { + lineBreak = true; + if (*(iter + 1) == (uchar32_t)'\n') // Windows line breaks. + currentChar = *(++iter); + } + else if (currentChar == (uchar32_t)'\n') // Unix breaks + { + lineBreak = true; + } + + if (lineBreak) + { + previousChar = 0; + offset.Y += font_metrics.ascender / 64; + offset.X = position.UpperLeftCorner.X; + + if (hcenter) + offset.X += (position.getWidth() - textDimension.Width) >> 1; + ++iter; + continue; + } + + // Calculate the glyph offset. + s32 offx = Glyphs[n-1].offset.X; + s32 offy = (font_metrics.ascender / 64) - Glyphs[n-1].offset.Y; + + // Apply kerning. + core::vector2di k = getKerning(currentChar, previousChar); + offset.X += k.X; + offset.Y += k.Y; + + // Determine rendering information. + SGUITTGlyph& glyph = Glyphs[n-1]; + CGUITTGlyphPage* const page = Glyph_Pages[glyph.glyph_page]; + page->render_positions.push_back(core::position2di(offset.X + offx, offset.Y + offy)); + page->render_source_rects.push_back(glyph.source_rect); + Render_Map.set(glyph.glyph_page, page); + } + offset.X += getWidthFromCharacter(currentChar); + + previousChar = currentChar; + ++iter; + } + + // Draw now. + update_glyph_pages(); + core::map<u32, CGUITTGlyphPage*>::Iterator j = Render_Map.getIterator(); + while (!j.atEnd()) + { + core::map<u32, CGUITTGlyphPage*>::Node* n = j.getNode(); + j++; + if (n == 0) continue; + + CGUITTGlyphPage* page = n->getValue(); + + if (!use_transparency) color.color |= 0xff000000; + Driver->draw2DImageBatch(page->texture, page->render_positions, page->render_source_rects, clip, color, true); + } +} + +core::dimension2d<u32> CGUITTFont::getCharDimension(const wchar_t ch) const +{ + return core::dimension2d<u32>(getWidthFromCharacter(ch), getHeightFromCharacter(ch)); +} + +core::dimension2d<u32> CGUITTFont::getDimension(const wchar_t* text) const +{ + return getDimension(core::ustring(text)); +} + +core::dimension2d<u32> CGUITTFont::getDimension(const core::ustring& text) const +{ + // Get the maximum font height. Unfortunately, we have to do this hack as + // Irrlicht will draw things wrong. In FreeType, the font size is the + // maximum size for a single glyph, but that glyph may hang "under" the + // draw line, increasing the total font height to beyond the set size. + // Irrlicht does not understand this concept when drawing fonts. Also, I + // add +1 to give it a 1 pixel blank border. This makes things like + // tooltips look nicer. + s32 test1 = getHeightFromCharacter((uchar32_t)'g') + 1; + s32 test2 = getHeightFromCharacter((uchar32_t)'j') + 1; + s32 test3 = getHeightFromCharacter((uchar32_t)'_') + 1; + s32 max_font_height = core::max_(test1, core::max_(test2, test3)); + + core::dimension2d<u32> text_dimension(0, max_font_height); + core::dimension2d<u32> line(0, max_font_height); + + uchar32_t previousChar = 0; + core::ustring::const_iterator iter = text.begin(); + for (; !iter.atEnd(); ++iter) + { + uchar32_t p = *iter; + bool lineBreak = false; + if (p == '\r') // Mac or Windows line breaks. + { + lineBreak = true; + if (*(iter + 1) == '\n') + { + ++iter; + p = *iter; + } + } + else if (p == '\n') // Unix line breaks. + { + lineBreak = true; + } + + // Kerning. + core::vector2di k = getKerning(p, previousChar); + line.Width += k.X; + previousChar = p; + + // Check for linebreak. + if (lineBreak) + { + previousChar = 0; + text_dimension.Height += line.Height; + if (text_dimension.Width < line.Width) + text_dimension.Width = line.Width; + line.Width = 0; + line.Height = max_font_height; + continue; + } + line.Width += getWidthFromCharacter(p); + } + if (text_dimension.Width < line.Width) + text_dimension.Width = line.Width; + + return text_dimension; +} + +inline u32 CGUITTFont::getWidthFromCharacter(wchar_t c) const +{ + return getWidthFromCharacter((uchar32_t)c); +} + +inline u32 CGUITTFont::getWidthFromCharacter(uchar32_t c) const +{ + // Set the size of the face. + // This is because we cache faces and the face may have been set to a different size. + //FT_Set_Pixel_Sizes(tt_face, 0, size); + + u32 n = getGlyphIndexByChar(c); + if (n > 0) + { + int w = Glyphs[n-1].advance.x / 64; + return w; + } + if (c >= 0x2000) + return (font_metrics.ascender / 64); + else return (font_metrics.ascender / 64) / 2; +} + +inline u32 CGUITTFont::getHeightFromCharacter(wchar_t c) const +{ + return getHeightFromCharacter((uchar32_t)c); +} + +inline u32 CGUITTFont::getHeightFromCharacter(uchar32_t c) const +{ + // Set the size of the face. + // This is because we cache faces and the face may have been set to a different size. + //FT_Set_Pixel_Sizes(tt_face, 0, size); + + u32 n = getGlyphIndexByChar(c); + if (n > 0) + { + // Grab the true height of the character, taking into account underhanging glyphs. + s32 height = (font_metrics.ascender / 64) - Glyphs[n-1].offset.Y + Glyphs[n-1].source_rect.getHeight(); + return height; + } + if (c >= 0x2000) + return (font_metrics.ascender / 64); + else return (font_metrics.ascender / 64) / 2; +} + +u32 CGUITTFont::getGlyphIndexByChar(wchar_t c) const +{ + return getGlyphIndexByChar((uchar32_t)c); +} + +u32 CGUITTFont::getGlyphIndexByChar(uchar32_t c) const +{ + // Get the glyph. + u32 glyph = FT_Get_Char_Index(tt_face, c); + + // Check for a valid glyph. If it is invalid, attempt to use the replacement character. + if (glyph == 0) + glyph = FT_Get_Char_Index(tt_face, core::unicode::UTF_REPLACEMENT_CHARACTER); + + // If our glyph is already loaded, don't bother doing any batch loading code. + if (glyph != 0 && Glyphs[glyph - 1].isLoaded) + return glyph; + + // Determine our batch loading positions. + u32 half_size = (batch_load_size / 2); + u32 start_pos = 0; + if (c > half_size) start_pos = c - half_size; + u32 end_pos = start_pos + batch_load_size; + + // Load all our characters. + do + { + // Get the character we are going to load. + u32 char_index = FT_Get_Char_Index(tt_face, start_pos); + + // If the glyph hasn't been loaded yet, do it now. + if (char_index) + { + SGUITTGlyph& glyph = Glyphs[char_index - 1]; + if (!glyph.isLoaded) + { + glyph.preload(char_index, tt_face, Driver, size, load_flags); + Glyph_Pages[glyph.glyph_page]->pushGlyphToBePaged(&glyph); + } + } + } + while (++start_pos < end_pos); + + // Return our original character. + return glyph; +} + +s32 CGUITTFont::getCharacterFromPos(const wchar_t* text, s32 pixel_x) const +{ + return getCharacterFromPos(core::ustring(text), pixel_x); +} + +s32 CGUITTFont::getCharacterFromPos(const core::ustring& text, s32 pixel_x) const +{ + s32 x = 0; + //s32 idx = 0; + + u32 character = 0; + uchar32_t previousChar = 0; + core::ustring::const_iterator iter = text.begin(); + while (!iter.atEnd()) + { + uchar32_t c = *iter; + x += getWidthFromCharacter(c); + + // Kerning. + core::vector2di k = getKerning(c, previousChar); + x += k.X; + + if (x >= pixel_x) + return character; + + previousChar = c; + ++iter; + ++character; + } + + return -1; +} + +void CGUITTFont::setKerningWidth(s32 kerning) +{ + GlobalKerningWidth = kerning; +} + +void CGUITTFont::setKerningHeight(s32 kerning) +{ + GlobalKerningHeight = kerning; +} + +s32 CGUITTFont::getKerningWidth(const wchar_t* thisLetter, const wchar_t* previousLetter) const +{ + if (tt_face == 0) + return GlobalKerningWidth; + if (thisLetter == 0 || previousLetter == 0) + return 0; + + return getKerningWidth((uchar32_t)*thisLetter, (uchar32_t)*previousLetter); +} + +s32 CGUITTFont::getKerningWidth(const uchar32_t thisLetter, const uchar32_t previousLetter) const +{ + // Return only the kerning width. + return getKerning(thisLetter, previousLetter).X; +} + +s32 CGUITTFont::getKerningHeight() const +{ + // FreeType 2 currently doesn't return any height kerning information. + return GlobalKerningHeight; +} + +core::vector2di CGUITTFont::getKerning(const wchar_t thisLetter, const wchar_t previousLetter) const +{ + return getKerning((uchar32_t)thisLetter, (uchar32_t)previousLetter); +} + +core::vector2di CGUITTFont::getKerning(const uchar32_t thisLetter, const uchar32_t previousLetter) const +{ + if (tt_face == 0 || thisLetter == 0 || previousLetter == 0) + return core::vector2di(); + + // Set the size of the face. + // This is because we cache faces and the face may have been set to a different size. + FT_Set_Pixel_Sizes(tt_face, 0, size); + + core::vector2di ret(GlobalKerningWidth, GlobalKerningHeight); + + // If we don't have kerning, no point in continuing. + if (!FT_HAS_KERNING(tt_face)) + return ret; + + // Get the kerning information. + FT_Vector v; + FT_Get_Kerning(tt_face, getGlyphIndexByChar(previousLetter), getGlyphIndexByChar(thisLetter), FT_KERNING_DEFAULT, &v); + + // If we have a scalable font, the return value will be in font points. + if (FT_IS_SCALABLE(tt_face)) + { + // Font points, so divide by 64. + ret.X += (v.x / 64); + ret.Y += (v.y / 64); + } + else + { + // Pixel units. + ret.X += v.x; + ret.Y += v.y; + } + return ret; +} + +void CGUITTFont::setInvisibleCharacters(const wchar_t *s) +{ + core::ustring us(s); + Invisible = us; +} + +void CGUITTFont::setInvisibleCharacters(const core::ustring& s) +{ + Invisible = s; +} + +video::IImage* CGUITTFont::createTextureFromChar(const uchar32_t& ch) +{ + u32 n = getGlyphIndexByChar(ch); + const SGUITTGlyph& glyph = Glyphs[n-1]; + CGUITTGlyphPage* page = Glyph_Pages[glyph.glyph_page]; + + if (page->dirty) + page->updateTexture(); + + video::ITexture* tex = page->texture; + + // Acquire a read-only lock of the corresponding page texture. + #if IRRLICHT_VERSION_MAJOR==1 && IRRLICHT_VERSION_MINOR>=8 + void* ptr = tex->lock(video::ETLM_READ_ONLY); + #else + void* ptr = tex->lock(true); + #endif + + video::ECOLOR_FORMAT format = tex->getColorFormat(); + core::dimension2du tex_size = tex->getOriginalSize(); + video::IImage* pageholder = Driver->createImageFromData(format, tex_size, ptr, true, false); + + // Copy the image data out of the page texture. + core::dimension2du glyph_size(glyph.source_rect.getSize()); + video::IImage* image = Driver->createImage(format, glyph_size); + pageholder->copyTo(image, core::position2di(0, 0), glyph.source_rect); + + tex->unlock(); + return image; +} + +video::ITexture* CGUITTFont::getPageTextureByIndex(const u32& page_index) const +{ + if (page_index < Glyph_Pages.size()) + return Glyph_Pages[page_index]->texture; + else + return 0; +} + +void CGUITTFont::createSharedPlane() +{ + /* + 2___3 + | /| + | / | <-- plane mesh is like this, point 2 is (0,0), point 0 is (0, -1) + |/ | <-- the texture coords of point 2 is (0,0, point 0 is (0, 1) + 0---1 + */ + + using namespace core; + using namespace video; + using namespace scene; + S3DVertex vertices[4]; + u16 indices[6] = {0,2,3,3,1,0}; + vertices[0] = S3DVertex(vector3df(0,-1,0), vector3df(0,0,-1), SColor(255,255,255,255), vector2df(0,1)); + vertices[1] = S3DVertex(vector3df(1,-1,0), vector3df(0,0,-1), SColor(255,255,255,255), vector2df(1,1)); + vertices[2] = S3DVertex(vector3df(0, 0,0), vector3df(0,0,-1), SColor(255,255,255,255), vector2df(0,0)); + vertices[3] = S3DVertex(vector3df(1, 0,0), vector3df(0,0,-1), SColor(255,255,255,255), vector2df(1,0)); + + SMeshBuffer* buf = new SMeshBuffer(); + buf->append(vertices, 4, indices, 6); + + shared_plane_.addMeshBuffer( buf ); + + shared_plane_ptr_ = &shared_plane_; + buf->drop(); //the addMeshBuffer method will grab it, so we can drop this ptr. +} + +core::dimension2d<u32> CGUITTFont::getDimensionUntilEndOfLine(const wchar_t* p) const +{ + core::stringw s; + for (const wchar_t* temp = p; temp && *temp != '\0' && *temp != L'\r' && *temp != L'\n'; ++temp ) + s.append(*temp); + + return getDimension(s.c_str()); +} + +core::array<scene::ISceneNode*> CGUITTFont::addTextSceneNode(const wchar_t* text, scene::ISceneManager* smgr, scene::ISceneNode* parent, const video::SColor& color, bool center) +{ + using namespace core; + using namespace video; + using namespace scene; + + array<scene::ISceneNode*> container; + + if (!Driver || !smgr) return container; + if (!parent) + parent = smgr->addEmptySceneNode(smgr->getRootSceneNode(), -1); + // if you don't specify parent, then we add a empty node attached to the root node + // this is generally undesirable. + + if (!shared_plane_ptr_) //this points to a static mesh that contains the plane + createSharedPlane(); //if it's not initialized, we create one. + + dimension2d<s32> text_size(getDimension(text)); //convert from unsigned to signed. + vector3df start_point(0, 0, 0), offset; + + /** NOTICE: + Because we are considering adding texts into 3D world, all Y axis vectors are inverted. + **/ + + // There's currently no "vertical center" concept when you apply text scene node to the 3D world. + if (center) + { + offset.X = start_point.X = -text_size.Width / 2.f; + offset.Y = start_point.Y = +text_size.Height/ 2.f; + offset.X += (text_size.Width - getDimensionUntilEndOfLine(text).Width) >> 1; + } + + // the default font material + SMaterial mat; + mat.setFlag(video::EMF_LIGHTING, true); + mat.setFlag(video::EMF_ZWRITE_ENABLE, false); + mat.setFlag(video::EMF_NORMALIZE_NORMALS, true); + mat.ColorMaterial = video::ECM_NONE; + mat.MaterialType = use_transparency ? video::EMT_TRANSPARENT_ALPHA_CHANNEL : video::EMT_SOLID; + mat.MaterialTypeParam = 0.01f; + mat.DiffuseColor = color; + + wchar_t current_char = 0, previous_char = 0; + u32 n = 0; + + array<u32> glyph_indices; + + while (*text) + { + current_char = *text; + bool line_break=false; + if (current_char == L'\r') // Mac or Windows breaks + { + line_break = true; + if (*(text + 1) == L'\n') // Windows line breaks. + current_char = *(++text); + } + else if (current_char == L'\n') // Unix breaks + { + line_break = true; + } + + if (line_break) + { + previous_char = 0; + offset.Y -= tt_face->size->metrics.ascender / 64; + offset.X = start_point.X; + if (center) + offset.X += (text_size.Width - getDimensionUntilEndOfLine(text+1).Width) >> 1; + ++text; + } + else + { + n = getGlyphIndexByChar(current_char); + if (n > 0) + { + glyph_indices.push_back( n ); + + // Store glyph size and offset informations. + SGUITTGlyph const& glyph = Glyphs[n-1]; + u32 texw = glyph.source_rect.getWidth(); + u32 texh = glyph.source_rect.getHeight(); + s32 offx = glyph.offset.X; + s32 offy = (font_metrics.ascender / 64) - glyph.offset.Y; + + // Apply kerning. + vector2di k = getKerning(current_char, previous_char); + offset.X += k.X; + offset.Y += k.Y; + + vector3df current_pos(offset.X + offx, offset.Y - offy, 0); + dimension2d<u32> letter_size = dimension2d<u32>(texw, texh); + + // Now we copy planes corresponding to the letter size. + IMeshManipulator* mani = smgr->getMeshManipulator(); + IMesh* meshcopy = mani->createMeshCopy(shared_plane_ptr_); + #if IRRLICHT_VERSION_MAJOR==1 && IRRLICHT_VERSION_MINOR>=8 + mani->scale(meshcopy, vector3df((f32)letter_size.Width, (f32)letter_size.Height, 1)); + #else + mani->scaleMesh(meshcopy, vector3df((f32)letter_size.Width, (f32)letter_size.Height, 1)); + #endif + + ISceneNode* current_node = smgr->addMeshSceneNode(meshcopy, parent, -1, current_pos); + meshcopy->drop(); + + current_node->getMaterial(0) = mat; + current_node->setAutomaticCulling(EAC_OFF); + current_node->setIsDebugObject(true); //so the picking won't have any effect on individual letter + //current_node->setDebugDataVisible(EDS_BBOX); //de-comment this when debugging + + container.push_back(current_node); + } + offset.X += getWidthFromCharacter(current_char); + previous_char = current_char; + ++text; + } + } + + update_glyph_pages(); + //only after we update the textures can we use the glyph page textures. + + for (u32 i = 0; i < glyph_indices.size(); ++i) + { + u32 n = glyph_indices[i]; + SGUITTGlyph const& glyph = Glyphs[n-1]; + ITexture* current_tex = Glyph_Pages[glyph.glyph_page]->texture; + f32 page_texture_size = (f32)current_tex->getSize().Width; + //Now we calculate the UV position according to the texture size and the source rect. + // + // 2___3 + // | /| + // | / | <-- plane mesh is like this, point 2 is (0,0), point 0 is (0, -1) + // |/ | <-- the texture coords of point 2 is (0,0, point 0 is (0, 1) + // 0---1 + // + f32 u1 = glyph.source_rect.UpperLeftCorner.X / page_texture_size; + f32 u2 = u1 + (glyph.source_rect.getWidth() / page_texture_size); + f32 v1 = glyph.source_rect.UpperLeftCorner.Y / page_texture_size; + f32 v2 = v1 + (glyph.source_rect.getHeight() / page_texture_size); + + //we can be quite sure that this is IMeshSceneNode, because we just added them in the above loop. + IMeshSceneNode* node = static_cast<IMeshSceneNode*>(container[i]); + + S3DVertex* pv = static_cast<S3DVertex*>(node->getMesh()->getMeshBuffer(0)->getVertices()); + //pv[0].TCoords.Y = pv[1].TCoords.Y = (letter_size.Height - 1) / static_cast<f32>(letter_size.Height); + //pv[1].TCoords.X = pv[3].TCoords.X = (letter_size.Width - 1) / static_cast<f32>(letter_size.Width); + pv[0].TCoords = vector2df(u1, v2); + pv[1].TCoords = vector2df(u2, v2); + pv[2].TCoords = vector2df(u1, v1); + pv[3].TCoords = vector2df(u2, v1); + + container[i]->getMaterial(0).setTexture(0, current_tex); + } + + return container; +} + +} // end namespace gui +} // end namespace irr diff --git a/src/cguittfont/CGUITTFont.h b/src/cguittfont/CGUITTFont.h new file mode 100644 index 000000000..12e25e0f3 --- /dev/null +++ b/src/cguittfont/CGUITTFont.h @@ -0,0 +1,377 @@ +/* + CGUITTFont FreeType class for Irrlicht + Copyright (c) 2009-2010 John Norman + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any + damages arising from the use of this software. + + Permission is granted to anyone to use this software for any + purpose, including commercial applications, and to alter it and + redistribute it freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you + must not claim that you wrote the original software. If you use + this software in a product, an acknowledgment in the product + documentation would be appreciated but is not required. + + 2. Altered source versions must be plainly marked as such, and + must not be misrepresented as being the original software. + + 3. This notice may not be removed or altered from any source + distribution. + + The original version of this class can be located at: + http://irrlicht.suckerfreegames.com/ + + John Norman + john@suckerfreegames.com +*/ + +#ifndef __C_GUI_TTFONT_H_INCLUDED__ +#define __C_GUI_TTFONT_H_INCLUDED__ + +#include <irrlicht.h> +#include <ft2build.h> +#include FT_FREETYPE_H + +namespace irr +{ +namespace gui +{ + struct SGUITTFace; + class CGUITTFont; + + //! Class to assist in deleting glyphs. + class CGUITTAssistDelete + { + public: + template <class T, typename TAlloc> + static void Delete(core::array<T, TAlloc>& a) + { + TAlloc allocator; + allocator.deallocate(a.pointer()); + } + }; + + //! Structure representing a single TrueType glyph. + struct SGUITTGlyph + { + //! Constructor. + SGUITTGlyph() : isLoaded(false), glyph_page(0), surface(0), parent(0) {} + + //! Destructor. + ~SGUITTGlyph() { unload(); } + + //! Preload the glyph. + //! The preload process occurs when the program tries to cache the glyph from FT_Library. + //! However, it simply defines the SGUITTGlyph's properties and will only create the page + //! textures if necessary. The actual creation of the textures should only occur right + //! before the batch draw call. + void preload(u32 char_index, FT_Face face, video::IVideoDriver* driver, u32 font_size, const FT_Int32 loadFlags); + + //! Unloads the glyph. + void unload(); + + //! Creates the IImage object from the FT_Bitmap. + video::IImage* createGlyphImage(const FT_Bitmap& bits, video::IVideoDriver* driver) const; + + //! If true, the glyph has been loaded. + bool isLoaded; + + //! The page the glyph is on. + u32 glyph_page; + + //! The source rectangle for the glyph. + core::recti source_rect; + + //! The offset of glyph when drawn. + core::vector2di offset; + + //! Glyph advance information. + FT_Vector advance; + + //! This is just the temporary image holder. After this glyph is paged, + //! it will be dropped. + mutable video::IImage* surface; + + //! The pointer pointing to the parent (CGUITTFont) + CGUITTFont* parent; + }; + + //! Holds a sheet of glyphs. + class CGUITTGlyphPage + { + public: + CGUITTGlyphPage(video::IVideoDriver* Driver, const io::path& texture_name) :texture(0), available_slots(0), used_slots(0), dirty(false), driver(Driver), name(texture_name) {} + ~CGUITTGlyphPage() + { + if (texture) + { + if (driver) + driver->removeTexture(texture); + else texture->drop(); + } + } + + //! Create the actual page texture, + bool createPageTexture(const u8& pixel_mode, const core::dimension2du& texture_size) + { + if( texture ) + return false; + + bool flgmip = driver->getTextureCreationFlag(video::ETCF_CREATE_MIP_MAPS); + driver->setTextureCreationFlag(video::ETCF_CREATE_MIP_MAPS, false); + + // Set the texture color format. + switch (pixel_mode) + { + case FT_PIXEL_MODE_MONO: + texture = driver->addTexture(texture_size, name, video::ECF_A1R5G5B5); + break; + case FT_PIXEL_MODE_GRAY: + default: + texture = driver->addTexture(texture_size, name, video::ECF_A8R8G8B8); + break; + } + + // Restore our texture creation flags. + driver->setTextureCreationFlag(video::ETCF_CREATE_MIP_MAPS, flgmip); + return texture ? true : false; + } + + //! Add the glyph to a list of glyphs to be paged. + //! This collection will be cleared after updateTexture is called. + void pushGlyphToBePaged(const SGUITTGlyph* glyph) + { + glyph_to_be_paged.push_back(glyph); + } + + //! Updates the texture atlas with new glyphs. + void updateTexture() + { + if (!dirty) return; + + void* ptr = texture->lock(); + video::ECOLOR_FORMAT format = texture->getColorFormat(); + core::dimension2du size = texture->getOriginalSize(); + video::IImage* pageholder = driver->createImageFromData(format, size, ptr, true, false); + + for (u32 i = 0; i < glyph_to_be_paged.size(); ++i) + { + const SGUITTGlyph* glyph = glyph_to_be_paged[i]; + if (glyph && glyph->isLoaded) + { + if (glyph->surface) + { + glyph->surface->copyTo(pageholder, glyph->source_rect.UpperLeftCorner); + glyph->surface->drop(); + glyph->surface = 0; + } + else + { + ; // TODO: add error message? + //currently, if we failed to create the image, just ignore this operation. + } + } + } + + pageholder->drop(); + texture->unlock(); + glyph_to_be_paged.clear(); + dirty = false; + } + + video::ITexture* texture; + u32 available_slots; + u32 used_slots; + bool dirty; + + core::array<core::vector2di> render_positions; + core::array<core::recti> render_source_rects; + + private: + core::array<const SGUITTGlyph*> glyph_to_be_paged; + video::IVideoDriver* driver; + io::path name; + }; + + //! Class representing a TrueType font. + class CGUITTFont : public IGUIFont + { + public: + //! Creates a new TrueType font and returns a pointer to it. The pointer must be drop()'ed when finished. + //! \param env The IGUIEnvironment the font loads out of. + //! \param filename The filename of the font. + //! \param size The size of the font glyphs in pixels. Since this is the size of the individual glyphs, the true height of the font may change depending on the characters used. + //! \param antialias set the use_monochrome (opposite to antialias) flag + //! \param transparency set the use_transparency flag + //! \return Returns a pointer to a CGUITTFont. Will return 0 if the font failed to load. + static CGUITTFont* createTTFont(IGUIEnvironment *env, const io::path& filename, const u32 size, const bool antialias = true, const bool transparency = true); + static CGUITTFont* createTTFont(IrrlichtDevice *device, const io::path& filename, const u32 size, const bool antialias = true, const bool transparency = true); + static CGUITTFont* create(IGUIEnvironment *env, const io::path& filename, const u32 size, const bool antialias = true, const bool transparency = true); + static CGUITTFont* create(IrrlichtDevice *device, const io::path& filename, const u32 size, const bool antialias = true, const bool transparency = true); + + //! Destructor + virtual ~CGUITTFont(); + + //! Sets the amount of glyphs to batch load. + virtual void setBatchLoadSize(u32 batch_size) { batch_load_size = batch_size; } + + //! Sets the maximum texture size for a page of glyphs. + virtual void setMaxPageTextureSize(const core::dimension2du& texture_size) { max_page_texture_size = texture_size; } + + //! Get the font size. + virtual u32 getFontSize() const { return size; } + + //! Check the font's transparency. + virtual bool isTransparent() const { return use_transparency; } + + //! Check if the font auto-hinting is enabled. + //! Auto-hinting is FreeType's built-in font hinting engine. + virtual bool useAutoHinting() const { return use_auto_hinting; } + + //! Check if the font hinting is enabled. + virtual bool useHinting() const { return use_hinting; } + + //! Check if the font is being loaded as a monochrome font. + //! The font can either be a 256 color grayscale font, or a 2 color monochrome font. + virtual bool useMonochrome() const { return use_monochrome; } + + //! Tells the font to allow transparency when rendering. + //! Default: true. + //! \param flag If true, the font draws using transparency. + virtual void setTransparency(const bool flag); + + //! Tells the font to use monochrome rendering. + //! Default: false. + //! \param flag If true, the font draws using a monochrome image. If false, the font uses a grayscale image. + virtual void setMonochrome(const bool flag); + + //! Enables or disables font hinting. + //! Default: Hinting and auto-hinting true. + //! \param enable If false, font hinting is turned off. If true, font hinting is turned on. + //! \param enable_auto_hinting If true, FreeType uses its own auto-hinting algorithm. If false, it tries to use the algorithm specified by the font. + virtual void setFontHinting(const bool enable, const bool enable_auto_hinting = true); + + //! Draws some text and clips it to the specified rectangle if wanted. + virtual void draw(const core::stringw& text, const core::rect<s32>& position, + video::SColor color, bool hcenter=false, bool vcenter=false, + const core::rect<s32>* clip=0); + + //! Returns the dimension of a character produced by this font. + virtual core::dimension2d<u32> getCharDimension(const wchar_t ch) const; + + //! Returns the dimension of a text string. + virtual core::dimension2d<u32> getDimension(const wchar_t* text) const; + virtual core::dimension2d<u32> getDimension(const core::ustring& text) const; + + //! Calculates the index of the character in the text which is on a specific position. + virtual s32 getCharacterFromPos(const wchar_t* text, s32 pixel_x) const; + virtual s32 getCharacterFromPos(const core::ustring& text, s32 pixel_x) const; + + //! Sets global kerning width for the font. + virtual void setKerningWidth(s32 kerning); + + //! Sets global kerning height for the font. + virtual void setKerningHeight(s32 kerning); + + //! Gets kerning values (distance between letters) for the font. If no parameters are provided, + virtual s32 getKerningWidth(const wchar_t* thisLetter=0, const wchar_t* previousLetter=0) const; + virtual s32 getKerningWidth(const uchar32_t thisLetter=0, const uchar32_t previousLetter=0) const; + + //! Returns the distance between letters + virtual s32 getKerningHeight() const; + + //! Define which characters should not be drawn by the font. + virtual void setInvisibleCharacters(const wchar_t *s); + virtual void setInvisibleCharacters(const core::ustring& s); + + //! Get the last glyph page if there's still available slots. + //! If not, it will return zero. + CGUITTGlyphPage* getLastGlyphPage() const; + + //! Create a new glyph page texture. + //! \param pixel_mode the pixel mode defined by FT_Pixel_Mode + //should be better typed. fix later. + CGUITTGlyphPage* createGlyphPage(const u8& pixel_mode); + + //! Get the last glyph page's index. + u32 getLastGlyphPageIndex() const { return Glyph_Pages.size() - 1; } + + //! Create corresponding character's software image copy from the font, + //! so you can use this data just like any ordinary video::IImage. + //! \param ch The character you need + virtual video::IImage* createTextureFromChar(const uchar32_t& ch); + + //! This function is for debugging mostly. If the page doesn't exist it returns zero. + //! \param page_index Simply return the texture handle of a given page index. + virtual video::ITexture* getPageTextureByIndex(const u32& page_index) const; + + //! Add a list of scene nodes generated by putting font textures on the 3D planes. + virtual core::array<scene::ISceneNode*> addTextSceneNode + (const wchar_t* text, scene::ISceneManager* smgr, scene::ISceneNode* parent = 0, + const video::SColor& color = video::SColor(255, 0, 0, 0), bool center = false ); + + protected: + bool use_monochrome; + bool use_transparency; + bool use_hinting; + bool use_auto_hinting; + u32 size; + u32 batch_load_size; + core::dimension2du max_page_texture_size; + + private: + // Manages the FreeType library. + static FT_Library c_library; + static core::map<io::path, SGUITTFace*> c_faces; + static bool c_libraryLoaded; + static scene::IMesh* shared_plane_ptr_; + static scene::SMesh shared_plane_; + + CGUITTFont(IGUIEnvironment *env); + bool load(const io::path& filename, const u32 size, const bool antialias, const bool transparency); + void reset_images(); + void update_glyph_pages() const; + void update_load_flags() + { + // Set up our loading flags. + load_flags = FT_LOAD_DEFAULT | FT_LOAD_RENDER; + if (!useHinting()) load_flags |= FT_LOAD_NO_HINTING; + if (!useAutoHinting()) load_flags |= FT_LOAD_NO_AUTOHINT; + if (useMonochrome()) load_flags |= FT_LOAD_MONOCHROME | FT_LOAD_TARGET_MONO | FT_RENDER_MODE_MONO; + else load_flags |= FT_LOAD_TARGET_NORMAL; + } + u32 getWidthFromCharacter(wchar_t c) const; + u32 getWidthFromCharacter(uchar32_t c) const; + u32 getHeightFromCharacter(wchar_t c) const; + u32 getHeightFromCharacter(uchar32_t c) const; + u32 getGlyphIndexByChar(wchar_t c) const; + u32 getGlyphIndexByChar(uchar32_t c) const; + core::vector2di getKerning(const wchar_t thisLetter, const wchar_t previousLetter) const; + core::vector2di getKerning(const uchar32_t thisLetter, const uchar32_t previousLetter) const; + core::dimension2d<u32> getDimensionUntilEndOfLine(const wchar_t* p) const; + + void createSharedPlane(); + + irr::IrrlichtDevice* Device; + gui::IGUIEnvironment* Environment; + video::IVideoDriver* Driver; + io::path filename; + FT_Face tt_face; + FT_Size_Metrics font_metrics; + FT_Int32 load_flags; + + mutable core::array<CGUITTGlyphPage*> Glyph_Pages; + mutable core::array<SGUITTGlyph> Glyphs; + + s32 GlobalKerningWidth; + s32 GlobalKerningHeight; + core::ustring Invisible; + }; + +} // end namespace gui +} // end namespace irr + +#endif // __C_GUI_TTFONT_H_INCLUDED__ diff --git a/src/cguittfont/CMakeLists.txt b/src/cguittfont/CMakeLists.txt new file mode 100644 index 000000000..94d061462 --- /dev/null +++ b/src/cguittfont/CMakeLists.txt @@ -0,0 +1,17 @@ +include_directories( + ${IRRLICHT_INCLUDE_DIR} + ${FREETYPE_INCLUDE_DIRS} +) + +# CGUITTFont authors, y u no include headers you use? +# Do not add CGUITTFont.cpp to the line below. +# xCGUITTFont.cpp is a wrapper file that includes +# additional required headers. +add_library(cguittfont xCGUITTFont.cpp) + +target_link_libraries( + cguittfont + ${IRRLICHT_LIBRARY} + ${FREETYPE_LIBRARY} + ${ZLIB_LIBRARIES} # needed by freetype, repeated here for safety +) diff --git a/src/cguittfont/irrUString.h b/src/cguittfont/irrUString.h new file mode 100644 index 000000000..f41fa1f7b --- /dev/null +++ b/src/cguittfont/irrUString.h @@ -0,0 +1,3877 @@ +/* + Basic Unicode string class for Irrlicht. + Copyright (c) 2009-2011 John Norman + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any + damages arising from the use of this software. + + Permission is granted to anyone to use this software for any + purpose, including commercial applications, and to alter it and + redistribute it freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you + must not claim that you wrote the original software. If you use + this software in a product, an acknowledgment in the product + documentation would be appreciated but is not required. + + 2. Altered source versions must be plainly marked as such, and + must not be misrepresented as being the original software. + + 3. This notice may not be removed or altered from any source + distribution. + + The original version of this class can be located at: + http://irrlicht.suckerfreegames.com/ + + John Norman + john@suckerfreegames.com +*/ + +#ifndef __IRR_USTRING_H_INCLUDED__ +#define __IRR_USTRING_H_INCLUDED__ + +#if (__cplusplus > 199711L) || (_MSC_VER >= 1600) || defined(__GXX_EXPERIMENTAL_CXX0X__) +# define USTRING_CPP0X +# if defined(__GXX_EXPERIMENTAL_CXX0X__) && ((__GNUC__ > 4) || ((__GNUC__ == 4) && (__GNUC_MINOR__ >= 5))) +# define USTRING_CPP0X_NEWLITERALS +# endif +#endif + +#include <stdio.h> +#include <string.h> +#include <stdlib.h> + +#ifdef USTRING_CPP0X +# include <utility> +#endif + +#ifndef USTRING_NO_STL +# include <string> +# include <iterator> +# include <ostream> +#endif + +#include "irrTypes.h" +#include "irrAllocator.h" +#include "irrArray.h" +#include "irrMath.h" +#include "irrString.h" +#include "path.h" + +//! UTF-16 surrogate start values. +static const irr::u16 UTF16_HI_SURROGATE = 0xD800; +static const irr::u16 UTF16_LO_SURROGATE = 0xDC00; + +//! Is a UTF-16 code point a surrogate? +#define UTF16_IS_SURROGATE(c) (((c) & 0xF800) == 0xD800) +#define UTF16_IS_SURROGATE_HI(c) (((c) & 0xFC00) == 0xD800) +#define UTF16_IS_SURROGATE_LO(c) (((c) & 0xFC00) == 0xDC00) + + +namespace irr +{ + + // Define our character types. +#ifdef USTRING_CPP0X_NEWLITERALS // C++0x + typedef char32_t uchar32_t; + typedef char16_t uchar16_t; + typedef char uchar8_t; +#else + typedef u32 uchar32_t; + typedef u16 uchar16_t; + typedef u8 uchar8_t; +#endif + +namespace core +{ + +namespace unicode +{ + +//! The unicode replacement character. Used to replace invalid characters. +const irr::u16 UTF_REPLACEMENT_CHARACTER = 0xFFFD; + +//! Convert a UTF-16 surrogate pair into a UTF-32 character. +//! \param high The high value of the pair. +//! \param low The low value of the pair. +//! \return The UTF-32 character expressed by the surrogate pair. +inline uchar32_t toUTF32(uchar16_t high, uchar16_t low) +{ + // Convert the surrogate pair into a single UTF-32 character. + uchar32_t x = ((high & ((1 << 6) -1)) << 10) | (low & ((1 << 10) -1)); + uchar32_t wu = ((high >> 6) & ((1 << 5) - 1)) + 1; + return (wu << 16) | x; +} + +//! Swaps the endianness of a 16-bit value. +//! \return The new value. +inline uchar16_t swapEndian16(const uchar16_t& c) +{ + return ((c >> 8) & 0x00FF) | ((c << 8) & 0xFF00); +} + +//! Swaps the endianness of a 32-bit value. +//! \return The new value. +inline uchar32_t swapEndian32(const uchar32_t& c) +{ + return ((c >> 24) & 0x000000FF) | + ((c >> 8) & 0x0000FF00) | + ((c << 8) & 0x00FF0000) | + ((c << 24) & 0xFF000000); +} + +//! The Unicode byte order mark. +const u16 BOM = 0xFEFF; + +//! The size of the Unicode byte order mark in terms of the Unicode character size. +const u8 BOM_UTF8_LEN = 3; +const u8 BOM_UTF16_LEN = 1; +const u8 BOM_UTF32_LEN = 1; + +//! Unicode byte order marks for file operations. +const u8 BOM_ENCODE_UTF8[3] = { 0xEF, 0xBB, 0xBF }; +const u8 BOM_ENCODE_UTF16_BE[2] = { 0xFE, 0xFF }; +const u8 BOM_ENCODE_UTF16_LE[2] = { 0xFF, 0xFE }; +const u8 BOM_ENCODE_UTF32_BE[4] = { 0x00, 0x00, 0xFE, 0xFF }; +const u8 BOM_ENCODE_UTF32_LE[4] = { 0xFF, 0xFE, 0x00, 0x00 }; + +//! The size in bytes of the Unicode byte marks for file operations. +const u8 BOM_ENCODE_UTF8_LEN = 3; +const u8 BOM_ENCODE_UTF16_LEN = 2; +const u8 BOM_ENCODE_UTF32_LEN = 4; + +//! Unicode encoding type. +enum EUTF_ENCODE +{ + EUTFE_NONE = 0, + EUTFE_UTF8, + EUTFE_UTF16, + EUTFE_UTF16_LE, + EUTFE_UTF16_BE, + EUTFE_UTF32, + EUTFE_UTF32_LE, + EUTFE_UTF32_BE +}; + +//! Unicode endianness. +enum EUTF_ENDIAN +{ + EUTFEE_NATIVE = 0, + EUTFEE_LITTLE, + EUTFEE_BIG +}; + +//! Returns the specified unicode byte order mark in a byte array. +//! The byte order mark is the first few bytes in a text file that signifies its encoding. +/** \param mode The Unicode encoding method that we want to get the byte order mark for. + If EUTFE_UTF16 or EUTFE_UTF32 is passed, it uses the native system endianness. **/ +//! \return An array that contains a byte order mark. +inline core::array<u8> getUnicodeBOM(EUTF_ENCODE mode) +{ +#define COPY_ARRAY(source, size) \ + memcpy(ret.pointer(), source, size); \ + ret.set_used(size) + + core::array<u8> ret(4); + switch (mode) + { + case EUTFE_UTF8: + COPY_ARRAY(BOM_ENCODE_UTF8, BOM_ENCODE_UTF8_LEN); + break; + case EUTFE_UTF16: + #ifdef __BIG_ENDIAN__ + COPY_ARRAY(BOM_ENCODE_UTF16_BE, BOM_ENCODE_UTF16_LEN); + #else + COPY_ARRAY(BOM_ENCODE_UTF16_LE, BOM_ENCODE_UTF16_LEN); + #endif + break; + case EUTFE_UTF16_BE: + COPY_ARRAY(BOM_ENCODE_UTF16_BE, BOM_ENCODE_UTF16_LEN); + break; + case EUTFE_UTF16_LE: + COPY_ARRAY(BOM_ENCODE_UTF16_LE, BOM_ENCODE_UTF16_LEN); + break; + case EUTFE_UTF32: + #ifdef __BIG_ENDIAN__ + COPY_ARRAY(BOM_ENCODE_UTF32_BE, BOM_ENCODE_UTF32_LEN); + #else + COPY_ARRAY(BOM_ENCODE_UTF32_LE, BOM_ENCODE_UTF32_LEN); + #endif + break; + case EUTFE_UTF32_BE: + COPY_ARRAY(BOM_ENCODE_UTF32_BE, BOM_ENCODE_UTF32_LEN); + break; + case EUTFE_UTF32_LE: + COPY_ARRAY(BOM_ENCODE_UTF32_LE, BOM_ENCODE_UTF32_LEN); + break; + } + return ret; + +#undef COPY_ARRAY +} + +//! Detects if the given data stream starts with a unicode BOM. +//! \param data The data stream to check. +//! \return The unicode BOM associated with the data stream, or EUTFE_NONE if none was found. +inline EUTF_ENCODE determineUnicodeBOM(const char* data) +{ + if (memcmp(data, BOM_ENCODE_UTF8, 3) == 0) return EUTFE_UTF8; + if (memcmp(data, BOM_ENCODE_UTF16_BE, 2) == 0) return EUTFE_UTF16_BE; + if (memcmp(data, BOM_ENCODE_UTF16_LE, 2) == 0) return EUTFE_UTF16_LE; + if (memcmp(data, BOM_ENCODE_UTF32_BE, 4) == 0) return EUTFE_UTF32_BE; + if (memcmp(data, BOM_ENCODE_UTF32_LE, 4) == 0) return EUTFE_UTF32_LE; + return EUTFE_NONE; +} + +} // end namespace unicode + + +//! UTF-16 string class. +template <typename TAlloc = irrAllocator<uchar16_t> > +class ustring16 +{ +public: + + ///------------------/// + /// iterator classes /// + ///------------------/// + + //! Access an element in a unicode string, allowing one to change it. + class _ustring16_iterator_access + { + public: + _ustring16_iterator_access(const ustring16<TAlloc>* s, u32 p) : ref(s), pos(p) {} + + //! Allow the class to be interpreted as a single UTF-32 character. + operator uchar32_t() const + { + return _get(); + } + + //! Allow one to change the character in the unicode string. + //! \param c The new character to use. + //! \return Myself. + _ustring16_iterator_access& operator=(const uchar32_t c) + { + _set(c); + return *this; + } + + //! Increments the value by 1. + //! \return Myself. + _ustring16_iterator_access& operator++() + { + _set(_get() + 1); + return *this; + } + + //! Increments the value by 1, returning the old value. + //! \return A unicode character. + uchar32_t operator++(int) + { + uchar32_t old = _get(); + _set(old + 1); + return old; + } + + //! Decrements the value by 1. + //! \return Myself. + _ustring16_iterator_access& operator--() + { + _set(_get() - 1); + return *this; + } + + //! Decrements the value by 1, returning the old value. + //! \return A unicode character. + uchar32_t operator--(int) + { + uchar32_t old = _get(); + _set(old - 1); + return old; + } + + //! Adds to the value by a specified amount. + //! \param val The amount to add to this character. + //! \return Myself. + _ustring16_iterator_access& operator+=(int val) + { + _set(_get() + val); + return *this; + } + + //! Subtracts from the value by a specified amount. + //! \param val The amount to subtract from this character. + //! \return Myself. + _ustring16_iterator_access& operator-=(int val) + { + _set(_get() - val); + return *this; + } + + //! Multiples the value by a specified amount. + //! \param val The amount to multiply this character by. + //! \return Myself. + _ustring16_iterator_access& operator*=(int val) + { + _set(_get() * val); + return *this; + } + + //! Divides the value by a specified amount. + //! \param val The amount to divide this character by. + //! \return Myself. + _ustring16_iterator_access& operator/=(int val) + { + _set(_get() / val); + return *this; + } + + //! Modulos the value by a specified amount. + //! \param val The amount to modulo this character by. + //! \return Myself. + _ustring16_iterator_access& operator%=(int val) + { + _set(_get() % val); + return *this; + } + + //! Adds to the value by a specified amount. + //! \param val The amount to add to this character. + //! \return A unicode character. + uchar32_t operator+(int val) const + { + return _get() + val; + } + + //! Subtracts from the value by a specified amount. + //! \param val The amount to subtract from this character. + //! \return A unicode character. + uchar32_t operator-(int val) const + { + return _get() - val; + } + + //! Multiplies the value by a specified amount. + //! \param val The amount to multiply this character by. + //! \return A unicode character. + uchar32_t operator*(int val) const + { + return _get() * val; + } + + //! Divides the value by a specified amount. + //! \param val The amount to divide this character by. + //! \return A unicode character. + uchar32_t operator/(int val) const + { + return _get() / val; + } + + //! Modulos the value by a specified amount. + //! \param val The amount to modulo this character by. + //! \return A unicode character. + uchar32_t operator%(int val) const + { + return _get() % val; + } + + private: + //! Gets a uchar32_t from our current position. + uchar32_t _get() const + { + const uchar16_t* a = ref->c_str(); + if (!UTF16_IS_SURROGATE(a[pos])) + return static_cast<uchar32_t>(a[pos]); + else + { + if (pos + 1 >= ref->size_raw()) + return 0; + + return unicode::toUTF32(a[pos], a[pos + 1]); + } + } + + //! Sets a uchar32_t at our current position. + void _set(uchar32_t c) + { + ustring16<TAlloc>* ref2 = const_cast<ustring16<TAlloc>*>(ref); + const uchar16_t* a = ref2->c_str(); + if (c > 0xFFFF) + { + // c will be multibyte, so split it up into the high and low surrogate pairs. + uchar16_t x = static_cast<uchar16_t>(c); + uchar16_t vh = UTF16_HI_SURROGATE | ((((c >> 16) & ((1 << 5) - 1)) - 1) << 6) | (x >> 10); + uchar16_t vl = UTF16_LO_SURROGATE | (x & ((1 << 10) - 1)); + + // If the previous position was a surrogate pair, just replace them. Else, insert the low pair. + if (UTF16_IS_SURROGATE_HI(a[pos]) && pos + 1 != ref2->size_raw()) + ref2->replace_raw(vl, static_cast<u32>(pos) + 1); + else ref2->insert_raw(vl, static_cast<u32>(pos) + 1); + + ref2->replace_raw(vh, static_cast<u32>(pos)); + } + else + { + // c will be a single byte. + uchar16_t vh = static_cast<uchar16_t>(c); + + // If the previous position was a surrogate pair, remove the extra byte. + if (UTF16_IS_SURROGATE_HI(a[pos])) + ref2->erase_raw(static_cast<u32>(pos) + 1); + + ref2->replace_raw(vh, static_cast<u32>(pos)); + } + } + + const ustring16<TAlloc>* ref; + u32 pos; + }; + typedef typename ustring16<TAlloc>::_ustring16_iterator_access access; + + + //! Iterator to iterate through a UTF-16 string. +#ifndef USTRING_NO_STL + class _ustring16_const_iterator : public std::iterator< + std::bidirectional_iterator_tag, // iterator_category + access, // value_type + ptrdiff_t, // difference_type + const access, // pointer + const access // reference + > +#else + class _ustring16_const_iterator +#endif + { + public: + typedef _ustring16_const_iterator _Iter; + typedef std::iterator<std::bidirectional_iterator_tag, access, ptrdiff_t, const access, const access> _Base; + typedef const access const_pointer; + typedef const access const_reference; + +#ifndef USTRING_NO_STL + typedef typename _Base::value_type value_type; + typedef typename _Base::difference_type difference_type; + typedef typename _Base::difference_type distance_type; + typedef typename _Base::pointer pointer; + typedef const_reference reference; +#else + typedef access value_type; + typedef u32 difference_type; + typedef u32 distance_type; + typedef const_pointer pointer; + typedef const_reference reference; +#endif + + //! Constructors. + _ustring16_const_iterator(const _Iter& i) : ref(i.ref), pos(i.pos) {} + _ustring16_const_iterator(const ustring16<TAlloc>& s) : ref(&s), pos(0) {} + _ustring16_const_iterator(const ustring16<TAlloc>& s, const u32 p) : ref(&s), pos(0) + { + if (ref->size_raw() == 0 || p == 0) + return; + + // Go to the appropriate position. + u32 i = p; + u32 sr = ref->size_raw(); + const uchar16_t* a = ref->c_str(); + while (i != 0 && pos < sr) + { + if (UTF16_IS_SURROGATE_HI(a[pos])) + pos += 2; + else ++pos; + --i; + } + } + + //! Test for equalness. + bool operator==(const _Iter& iter) const + { + if (ref == iter.ref && pos == iter.pos) + return true; + return false; + } + + //! Test for unequalness. + bool operator!=(const _Iter& iter) const + { + if (ref != iter.ref || pos != iter.pos) + return true; + return false; + } + + //! Switch to the next full character in the string. + _Iter& operator++() + { // ++iterator + if (pos == ref->size_raw()) return *this; + const uchar16_t* a = ref->c_str(); + if (UTF16_IS_SURROGATE_HI(a[pos])) + pos += 2; // TODO: check for valid low surrogate? + else ++pos; + if (pos > ref->size_raw()) pos = ref->size_raw(); + return *this; + } + + //! Switch to the next full character in the string, returning the previous position. + _Iter operator++(int) + { // iterator++ + _Iter _tmp(*this); + ++*this; + return _tmp; + } + + //! Switch to the previous full character in the string. + _Iter& operator--() + { // --iterator + if (pos == 0) return *this; + const uchar16_t* a = ref->c_str(); + --pos; + if (UTF16_IS_SURROGATE_LO(a[pos]) && pos != 0) // low surrogate, go back one more. + --pos; + return *this; + } + + //! Switch to the previous full character in the string, returning the previous position. + _Iter operator--(int) + { // iterator-- + _Iter _tmp(*this); + --*this; + return _tmp; + } + + //! Advance a specified number of full characters in the string. + //! \return Myself. + _Iter& operator+=(const difference_type v) + { + if (v == 0) return *this; + if (v < 0) return operator-=(v * -1); + + if (pos >= ref->size_raw()) + return *this; + + // Go to the appropriate position. + // TODO: Don't force u32 on an x64 OS. Make it agnostic. + u32 i = (u32)v; + u32 sr = ref->size_raw(); + const uchar16_t* a = ref->c_str(); + while (i != 0 && pos < sr) + { + if (UTF16_IS_SURROGATE_HI(a[pos])) + pos += 2; + else ++pos; + --i; + } + if (pos > sr) + pos = sr; + + return *this; + } + + //! Go back a specified number of full characters in the string. + //! \return Myself. + _Iter& operator-=(const difference_type v) + { + if (v == 0) return *this; + if (v > 0) return operator+=(v * -1); + + if (pos == 0) + return *this; + + // Go to the appropriate position. + // TODO: Don't force u32 on an x64 OS. Make it agnostic. + u32 i = (u32)v; + const uchar16_t* a = ref->c_str(); + while (i != 0 && pos != 0) + { + --pos; + if (UTF16_IS_SURROGATE_LO(a[pos]) != 0 && pos != 0) + --pos; + --i; + } + + return *this; + } + + //! Return a new iterator that is a variable number of full characters forward from the current position. + _Iter operator+(const difference_type v) const + { + _Iter ret(*this); + ret += v; + return ret; + } + + //! Return a new iterator that is a variable number of full characters backward from the current position. + _Iter operator-(const difference_type v) const + { + _Iter ret(*this); + ret -= v; + return ret; + } + + //! Returns the distance between two iterators. + difference_type operator-(const _Iter& iter) const + { + // Make sure we reference the same object! + if (ref != iter.ref) + return difference_type(); + + _Iter i = iter; + difference_type ret; + + // Walk up. + if (pos > i.pos) + { + while (pos > i.pos) + { + ++i; + ++ret; + } + return ret; + } + + // Walk down. + while (pos < i.pos) + { + --i; + --ret; + } + return ret; + } + + //! Accesses the full character at the iterator's position. + const_reference operator*() const + { + if (pos >= ref->size_raw()) + { + const uchar16_t* a = ref->c_str(); + u32 p = ref->size_raw(); + if (UTF16_IS_SURROGATE_LO(a[p])) + --p; + reference ret(ref, p); + return ret; + } + const_reference ret(ref, pos); + return ret; + } + + //! Accesses the full character at the iterator's position. + reference operator*() + { + if (pos >= ref->size_raw()) + { + const uchar16_t* a = ref->c_str(); + u32 p = ref->size_raw(); + if (UTF16_IS_SURROGATE_LO(a[p])) + --p; + reference ret(ref, p); + return ret; + } + reference ret(ref, pos); + return ret; + } + + //! Accesses the full character at the iterator's position. + const_pointer operator->() const + { + return operator*(); + } + + //! Accesses the full character at the iterator's position. + pointer operator->() + { + return operator*(); + } + + //! Is the iterator at the start of the string? + bool atStart() const + { + return pos == 0; + } + + //! Is the iterator at the end of the string? + bool atEnd() const + { + const uchar16_t* a = ref->c_str(); + if (UTF16_IS_SURROGATE(a[pos])) + return (pos + 1) >= ref->size_raw(); + else return pos >= ref->size_raw(); + } + + //! Moves the iterator to the start of the string. + void toStart() + { + pos = 0; + } + + //! Moves the iterator to the end of the string. + void toEnd() + { + const uchar16_t* a = ref->c_str(); + pos = ref->size_raw(); + } + + //! Returns the iterator's position. + //! \return The iterator's position. + u32 getPos() const + { + return pos; + } + + protected: + const ustring16<TAlloc>* ref; + u32 pos; + }; + + //! Iterator to iterate through a UTF-16 string. + class _ustring16_iterator : public _ustring16_const_iterator + { + public: + typedef _ustring16_iterator _Iter; + typedef _ustring16_const_iterator _Base; + typedef typename _Base::const_pointer const_pointer; + typedef typename _Base::const_reference const_reference; + + typedef typename _Base::value_type value_type; + typedef typename _Base::difference_type difference_type; + typedef typename _Base::distance_type distance_type; + typedef access pointer; + typedef access reference; + + using _Base::pos; + using _Base::ref; + + //! Constructors. + _ustring16_iterator(const _Iter& i) : _ustring16_const_iterator(i) {} + _ustring16_iterator(const ustring16<TAlloc>& s) : _ustring16_const_iterator(s) {} + _ustring16_iterator(const ustring16<TAlloc>& s, const u32 p) : _ustring16_const_iterator(s, p) {} + + //! Accesses the full character at the iterator's position. + reference operator*() const + { + if (pos >= ref->size_raw()) + { + const uchar16_t* a = ref->c_str(); + u32 p = ref->size_raw(); + if (UTF16_IS_SURROGATE_LO(a[p])) + --p; + reference ret(ref, p); + return ret; + } + reference ret(ref, pos); + return ret; + } + + //! Accesses the full character at the iterator's position. + reference operator*() + { + if (pos >= ref->size_raw()) + { + const uchar16_t* a = ref->c_str(); + u32 p = ref->size_raw(); + if (UTF16_IS_SURROGATE_LO(a[p])) + --p; + reference ret(ref, p); + return ret; + } + reference ret(ref, pos); + return ret; + } + + //! Accesses the full character at the iterator's position. + pointer operator->() const + { + return operator*(); + } + + //! Accesses the full character at the iterator's position. + pointer operator->() + { + return operator*(); + } + }; + + typedef typename ustring16<TAlloc>::_ustring16_iterator iterator; + typedef typename ustring16<TAlloc>::_ustring16_const_iterator const_iterator; + + ///----------------------/// + /// end iterator classes /// + ///----------------------/// + + //! Default constructor + ustring16() + : array(0), allocated(1), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + array = allocator.allocate(1); // new u16[1]; + array[0] = 0x0; + } + + + //! Constructor + ustring16(const ustring16<TAlloc>& other) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + *this = other; + } + + + //! Constructor from other string types + template <class B, class A> + ustring16(const string<B, A>& other) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + *this = other; + } + + +#ifndef USTRING_NO_STL + //! Constructor from std::string + template <class B, class A, typename Alloc> + ustring16(const std::basic_string<B, A, Alloc>& other) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + *this = other.c_str(); + } + + + //! Constructor from iterator. + template <typename Itr> + ustring16(Itr first, Itr last) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + reserve(std::distance(first, last)); + array[used] = 0; + + for (; first != last; ++first) + append((uchar32_t)*first); + } +#endif + + +#ifndef USTRING_CPP0X_NEWLITERALS + //! Constructor for copying a character string from a pointer. + ustring16(const char* const c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + loadDataStream(c, strlen(c)); + //append((uchar8_t*)c); + } + + + //! Constructor for copying a character string from a pointer with a given length. + ustring16(const char* const c, u32 length) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + loadDataStream(c, length); + } +#endif + + + //! Constructor for copying a UTF-8 string from a pointer. + ustring16(const uchar8_t* const c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c); + } + + + //! Constructor for copying a UTF-8 string from a single char. + ustring16(const char c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append((uchar32_t)c); + } + + + //! Constructor for copying a UTF-8 string from a pointer with a given length. + ustring16(const uchar8_t* const c, u32 length) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c, length); + } + + + //! Constructor for copying a UTF-16 string from a pointer. + ustring16(const uchar16_t* const c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c); + } + + + //! Constructor for copying a UTF-16 string from a pointer with a given length + ustring16(const uchar16_t* const c, u32 length) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c, length); + } + + + //! Constructor for copying a UTF-32 string from a pointer. + ustring16(const uchar32_t* const c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c); + } + + + //! Constructor for copying a UTF-32 from a pointer with a given length. + ustring16(const uchar32_t* const c, u32 length) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + append(c, length); + } + + + //! Constructor for copying a wchar_t string from a pointer. + ustring16(const wchar_t* const c) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + if (sizeof(wchar_t) == 4) + append(reinterpret_cast<const uchar32_t* const>(c)); + else if (sizeof(wchar_t) == 2) + append(reinterpret_cast<const uchar16_t* const>(c)); + else if (sizeof(wchar_t) == 1) + append(reinterpret_cast<const uchar8_t* const>(c)); + } + + + //! Constructor for copying a wchar_t string from a pointer with a given length. + ustring16(const wchar_t* const c, u32 length) + : array(0), allocated(0), used(0) + { +#if __BIG_ENDIAN__ + encoding = unicode::EUTFE_UTF16_BE; +#else + encoding = unicode::EUTFE_UTF16_LE; +#endif + + if (sizeof(wchar_t) == 4) + append(reinterpret_cast<const uchar32_t* const>(c), length); + else if (sizeof(wchar_t) == 2) + append(reinterpret_cast<const uchar16_t* const>(c), length); + else if (sizeof(wchar_t) == 1) + append(reinterpret_cast<const uchar8_t* const>(c), length); + } + + +#ifdef USTRING_CPP0X + //! Constructor for moving a ustring16 + ustring16(ustring16<TAlloc>&& other) + : array(other.array), encoding(other.encoding), allocated(other.allocated), used(other.used) + { + //std::cout << "MOVE constructor" << std::endl; + other.array = 0; + other.allocated = 0; + other.used = 0; + } +#endif + + + //! Destructor + ~ustring16() + { + allocator.deallocate(array); // delete [] array; + } + + + //! Assignment operator + ustring16& operator=(const ustring16<TAlloc>& other) + { + if (this == &other) + return *this; + + used = other.size_raw(); + if (used >= allocated) + { + allocator.deallocate(array); // delete [] array; + allocated = used + 1; + array = allocator.allocate(used + 1); //new u16[used]; + } + + const uchar16_t* p = other.c_str(); + for (u32 i=0; i<=used; ++i, ++p) + array[i] = *p; + + array[used] = 0; + + // Validate our new UTF-16 string. + validate(); + + return *this; + } + + +#ifdef USTRING_CPP0X + //! Move assignment operator + ustring16& operator=(ustring16<TAlloc>&& other) + { + if (this != &other) + { + //std::cout << "MOVE operator=" << std::endl; + allocator.deallocate(array); + + array = other.array; + allocated = other.allocated; + encoding = other.encoding; + used = other.used; + other.array = 0; + other.used = 0; + } + return *this; + } +#endif + + + //! Assignment operator for other string types + template <class B, class A> + ustring16<TAlloc>& operator=(const string<B, A>& other) + { + *this = other.c_str(); + return *this; + } + + + //! Assignment operator for UTF-8 strings + ustring16<TAlloc>& operator=(const uchar8_t* const c) + { + if (!array) + { + array = allocator.allocate(1); //new u16[1]; + allocated = 1; + } + used = 0; + array[used] = 0x0; + if (!c) return *this; + + //! Append our string now. + append(c); + return *this; + } + + + //! Assignment operator for UTF-16 strings + ustring16<TAlloc>& operator=(const uchar16_t* const c) + { + if (!array) + { + array = allocator.allocate(1); //new u16[1]; + allocated = 1; + } + used = 0; + array[used] = 0x0; + if (!c) return *this; + + //! Append our string now. + append(c); + return *this; + } + + + //! Assignment operator for UTF-32 strings + ustring16<TAlloc>& operator=(const uchar32_t* const c) + { + if (!array) + { + array = allocator.allocate(1); //new u16[1]; + allocated = 1; + } + used = 0; + array[used] = 0x0; + if (!c) return *this; + + //! Append our string now. + append(c); + return *this; + } + + + //! Assignment operator for wchar_t strings. + /** Note that this assumes that a correct unicode string is stored in the wchar_t string. + Since wchar_t changes depending on its platform, it could either be a UTF-8, -16, or -32 string. + This function assumes you are storing the correct unicode encoding inside the wchar_t string. **/ + ustring16<TAlloc>& operator=(const wchar_t* const c) + { + if (sizeof(wchar_t) == 4) + *this = reinterpret_cast<const uchar32_t* const>(c); + else if (sizeof(wchar_t) == 2) + *this = reinterpret_cast<const uchar16_t* const>(c); + else if (sizeof(wchar_t) == 1) + *this = reinterpret_cast<const uchar8_t* const>(c); + + return *this; + } + + + //! Assignment operator for other strings. + /** Note that this assumes that a correct unicode string is stored in the string. **/ + template <class B> + ustring16<TAlloc>& operator=(const B* const c) + { + if (sizeof(B) == 4) + *this = reinterpret_cast<const uchar32_t* const>(c); + else if (sizeof(B) == 2) + *this = reinterpret_cast<const uchar16_t* const>(c); + else if (sizeof(B) == 1) + *this = reinterpret_cast<const uchar8_t* const>(c); + + return *this; + } + + + //! Direct access operator + access operator [](const u32 index) + { + _IRR_DEBUG_BREAK_IF(index>=size()) // bad index + iterator iter(*this, index); + return iter.operator*(); + } + + + //! Direct access operator + const access operator [](const u32 index) const + { + _IRR_DEBUG_BREAK_IF(index>=size()) // bad index + const_iterator iter(*this, index); + return iter.operator*(); + } + + + //! Equality operator + bool operator ==(const uchar16_t* const str) const + { + if (!str) + return false; + + u32 i; + for(i=0; array[i] && str[i]; ++i) + if (array[i] != str[i]) + return false; + + return !array[i] && !str[i]; + } + + + //! Equality operator + bool operator ==(const ustring16<TAlloc>& other) const + { + for(u32 i=0; array[i] && other.array[i]; ++i) + if (array[i] != other.array[i]) + return false; + + return used == other.used; + } + + + //! Is smaller comparator + bool operator <(const ustring16<TAlloc>& other) const + { + for(u32 i=0; array[i] && other.array[i]; ++i) + { + s32 diff = array[i] - other.array[i]; + if ( diff ) + return diff < 0; + } + + return used < other.used; + } + + + //! Inequality operator + bool operator !=(const uchar16_t* const str) const + { + return !(*this == str); + } + + + //! Inequality operator + bool operator !=(const ustring16<TAlloc>& other) const + { + return !(*this == other); + } + + + //! Returns the length of a ustring16 in full characters. + //! \return Length of a ustring16 in full characters. + u32 size() const + { + const_iterator i(*this, 0); + u32 pos = 0; + while (!i.atEnd()) + { + ++i; + ++pos; + } + return pos; + } + + + //! Informs if the ustring is empty or not. + //! \return True if the ustring is empty, false if not. + bool empty() const + { + return (size_raw() == 0); + } + + + //! Returns a pointer to the raw UTF-16 string data. + //! \return pointer to C-style NUL terminated array of UTF-16 code points. + const uchar16_t* c_str() const + { + return array; + } + + + //! Compares the first n characters of this string with another. + //! \param other Other string to compare to. + //! \param n Number of characters to compare. + //! \return True if the n first characters of both strings are equal. + bool equalsn(const ustring16<TAlloc>& other, u32 n) const + { + u32 i; + const uchar16_t* oa = other.c_str(); + for(i=0; array[i] && oa[i] && i < n; ++i) + if (array[i] != oa[i]) + return false; + + // if one (or both) of the strings was smaller then they + // are only equal if they have the same length + return (i == n) || (used == other.used); + } + + + //! Compares the first n characters of this string with another. + //! \param str Other string to compare to. + //! \param n Number of characters to compare. + //! \return True if the n first characters of both strings are equal. + bool equalsn(const uchar16_t* const str, u32 n) const + { + if (!str) + return false; + u32 i; + for(i=0; array[i] && str[i] && i < n; ++i) + if (array[i] != str[i]) + return false; + + // if one (or both) of the strings was smaller then they + // are only equal if they have the same length + return (i == n) || (array[i] == 0 && str[i] == 0); + } + + + //! Appends a character to this ustring16 + //! \param character The character to append. + //! \return A reference to our current string. + ustring16<TAlloc>& append(uchar32_t character) + { + if (used + 2 >= allocated) + reallocate(used + 2); + + if (character > 0xFFFF) + { + used += 2; + + // character will be multibyte, so split it up into a surrogate pair. + uchar16_t x = static_cast<uchar16_t>(character); + uchar16_t vh = UTF16_HI_SURROGATE | ((((character >> 16) & ((1 << 5) - 1)) - 1) << 6) | (x >> 10); + uchar16_t vl = UTF16_LO_SURROGATE | (x & ((1 << 10) - 1)); + array[used-2] = vh; + array[used-1] = vl; + } + else + { + ++used; + array[used-1] = character; + } + array[used] = 0; + + return *this; + } + + + //! Appends a UTF-8 string to this ustring16 + //! \param other The UTF-8 string to append. + //! \param length The length of the string to append. + //! \return A reference to our current string. + ustring16<TAlloc>& append(const uchar8_t* const other, u32 length=0xffffffff) + { + if (!other) + return *this; + + // Determine if the string is long enough for a BOM. + u32 len = 0; + const uchar8_t* p = other; + do + { + ++len; + } while (*p++ && len < unicode::BOM_ENCODE_UTF8_LEN); + + // Check for BOM. + unicode::EUTF_ENCODE c_bom = unicode::EUTFE_NONE; + if (len == unicode::BOM_ENCODE_UTF8_LEN) + { + if (memcmp(other, unicode::BOM_ENCODE_UTF8, unicode::BOM_ENCODE_UTF8_LEN) == 0) + c_bom = unicode::EUTFE_UTF8; + } + + // If a BOM was found, don't include it in the string. + const uchar8_t* c2 = other; + if (c_bom != unicode::EUTFE_NONE) + { + c2 = other + unicode::BOM_UTF8_LEN; + length -= unicode::BOM_UTF8_LEN; + } + + // Calculate the size of the string to read in. + len = 0; + p = c2; + do + { + ++len; + } while(*p++ && len < length); + if (len > length) + len = length; + + // If we need to grow the array, do it now. + if (used + len >= allocated) + reallocate(used + (len * 2)); + u32 start = used; + + // Convert UTF-8 to UTF-16. + u32 pos = start; + for (u32 l = 0; l<len;) + { + ++used; + if (((c2[l] >> 6) & 0x03) == 0x02) + { // Invalid continuation byte. + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + ++l; + } + else if (c2[l] == 0xC0 || c2[l] == 0xC1) + { // Invalid byte - overlong encoding. + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + ++l; + } + else if ((c2[l] & 0xF8) == 0xF0) + { // 4 bytes UTF-8, 2 bytes UTF-16. + // Check for a full string. + if ((l + 3) >= len) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + l += 3; + break; + } + + // Validate. + bool valid = true; + u8 l2 = 0; + if (valid && (((c2[l+1] >> 6) & 0x03) == 0x02)) ++l2; else valid = false; + if (valid && (((c2[l+2] >> 6) & 0x03) == 0x02)) ++l2; else valid = false; + if (valid && (((c2[l+3] >> 6) & 0x03) == 0x02)) ++l2; else valid = false; + if (!valid) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + l += l2; + continue; + } + + // Decode. + uchar8_t b1 = ((c2[l] & 0x7) << 2) | ((c2[l+1] >> 4) & 0x3); + uchar8_t b2 = ((c2[l+1] & 0xF) << 4) | ((c2[l+2] >> 2) & 0xF); + uchar8_t b3 = ((c2[l+2] & 0x3) << 6) | (c2[l+3] & 0x3F); + uchar32_t v = b3 | ((uchar32_t)b2 << 8) | ((uchar32_t)b1 << 16); + + // Split v up into a surrogate pair. + uchar16_t x = static_cast<uchar16_t>(v); + uchar16_t vh = UTF16_HI_SURROGATE | ((((v >> 16) & ((1 << 5) - 1)) - 1) << 6) | (x >> 10); + uchar16_t vl = UTF16_LO_SURROGATE | (x & ((1 << 10) - 1)); + + array[pos++] = vh; + array[pos++] = vl; + l += 4; + ++used; // Using two shorts this time, so increase used by 1. + } + else if ((c2[l] & 0xF0) == 0xE0) + { // 3 bytes UTF-8, 1 byte UTF-16. + // Check for a full string. + if ((l + 2) >= len) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + l += 2; + break; + } + + // Validate. + bool valid = true; + u8 l2 = 0; + if (valid && (((c2[l+1] >> 6) & 0x03) == 0x02)) ++l2; else valid = false; + if (valid && (((c2[l+2] >> 6) & 0x03) == 0x02)) ++l2; else valid = false; + if (!valid) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + l += l2; + continue; + } + + // Decode. + uchar8_t b1 = ((c2[l] & 0xF) << 4) | ((c2[l+1] >> 2) & 0xF); + uchar8_t b2 = ((c2[l+1] & 0x3) << 6) | (c2[l+2] & 0x3F); + uchar16_t ch = b2 | ((uchar16_t)b1 << 8); + array[pos++] = ch; + l += 3; + } + else if ((c2[l] & 0xE0) == 0xC0) + { // 2 bytes UTF-8, 1 byte UTF-16. + // Check for a full string. + if ((l + 1) >= len) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + l += 1; + break; + } + + // Validate. + if (((c2[l+1] >> 6) & 0x03) != 0x02) + { + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + ++l; + continue; + } + + // Decode. + uchar8_t b1 = (c2[l] >> 2) & 0x7; + uchar8_t b2 = ((c2[l] & 0x3) << 6) | (c2[l+1] & 0x3F); + uchar16_t ch = b2 | ((uchar16_t)b1 << 8); + array[pos++] = ch; + l += 2; + } + else + { // 1 byte UTF-8, 1 byte UTF-16. + // Validate. + if (c2[l] > 0x7F) + { // Values above 0xF4 are restricted and aren't used. By now, anything above 0x7F is invalid. + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + } + else array[pos++] = static_cast<uchar16_t>(c2[l]); + ++l; + } + } + array[used] = 0; + + // Validate our new UTF-16 string. + validate(); + + return *this; + } + + + //! Appends a UTF-16 string to this ustring16 + //! \param other The UTF-16 string to append. + //! \param length The length of the string to append. + //! \return A reference to our current string. + ustring16<TAlloc>& append(const uchar16_t* const other, u32 length=0xffffffff) + { + if (!other) + return *this; + + // Determine if the string is long enough for a BOM. + u32 len = 0; + const uchar16_t* p = other; + do + { + ++len; + } while (*p++ && len < unicode::BOM_ENCODE_UTF16_LEN); + + // Check for the BOM to determine the string's endianness. + unicode::EUTF_ENDIAN c_end = unicode::EUTFEE_NATIVE; + if (memcmp(other, unicode::BOM_ENCODE_UTF16_LE, unicode::BOM_ENCODE_UTF16_LEN) == 0) + c_end = unicode::EUTFEE_LITTLE; + else if (memcmp(other, unicode::BOM_ENCODE_UTF16_BE, unicode::BOM_ENCODE_UTF16_LEN) == 0) + c_end = unicode::EUTFEE_BIG; + + // If a BOM was found, don't include it in the string. + const uchar16_t* c2 = other; + if (c_end != unicode::EUTFEE_NATIVE) + { + c2 = other + unicode::BOM_UTF16_LEN; + length -= unicode::BOM_UTF16_LEN; + } + + // Calculate the size of the string to read in. + len = 0; + p = c2; + do + { + ++len; + } while(*p++ && len < length); + if (len > length) + len = length; + + // If we need to grow the size of the array, do it now. + if (used + len >= allocated) + reallocate(used + (len * 2)); + u32 start = used; + used += len; + + // Copy the string now. + unicode::EUTF_ENDIAN m_end = getEndianness(); + for (u32 l = start; l < start + len; ++l) + { + array[l] = (uchar16_t)c2[l]; + if (c_end != unicode::EUTFEE_NATIVE && c_end != m_end) + array[l] = unicode::swapEndian16(array[l]); + } + + array[used] = 0; + + // Validate our new UTF-16 string. + validate(); + return *this; + } + + + //! Appends a UTF-32 string to this ustring16 + //! \param other The UTF-32 string to append. + //! \param length The length of the string to append. + //! \return A reference to our current string. + ustring16<TAlloc>& append(const uchar32_t* const other, u32 length=0xffffffff) + { + if (!other) + return *this; + + // Check for the BOM to determine the string's endianness. + unicode::EUTF_ENDIAN c_end = unicode::EUTFEE_NATIVE; + if (memcmp(other, unicode::BOM_ENCODE_UTF32_LE, unicode::BOM_ENCODE_UTF32_LEN) == 0) + c_end = unicode::EUTFEE_LITTLE; + else if (memcmp(other, unicode::BOM_ENCODE_UTF32_BE, unicode::BOM_ENCODE_UTF32_LEN) == 0) + c_end = unicode::EUTFEE_BIG; + + // If a BOM was found, don't include it in the string. + const uchar32_t* c2 = other; + if (c_end != unicode::EUTFEE_NATIVE) + { + c2 = other + unicode::BOM_UTF32_LEN; + length -= unicode::BOM_UTF32_LEN; + } + + // Calculate the size of the string to read in. + u32 len = 0; + const uchar32_t* p = c2; + do + { + ++len; + } while(*p++ && len < length); + if (len > length) + len = length; + + // If we need to grow the size of the array, do it now. + // In case all of the UTF-32 string is split into surrogate pairs, do len * 2. + if (used + (len * 2) >= allocated) + reallocate(used + ((len * 2) * 2)); + u32 start = used; + + // Convert UTF-32 to UTF-16. + unicode::EUTF_ENDIAN m_end = getEndianness(); + u32 pos = start; + for (u32 l = 0; l<len; ++l) + { + ++used; + + uchar32_t ch = c2[l]; + if (c_end != unicode::EUTFEE_NATIVE && c_end != m_end) + ch = unicode::swapEndian32(ch); + + if (ch > 0xFFFF) + { + // Split ch up into a surrogate pair as it is over 16 bits long. + uchar16_t x = static_cast<uchar16_t>(ch); + uchar16_t vh = UTF16_HI_SURROGATE | ((((ch >> 16) & ((1 << 5) - 1)) - 1) << 6) | (x >> 10); + uchar16_t vl = UTF16_LO_SURROGATE | (x & ((1 << 10) - 1)); + array[pos++] = vh; + array[pos++] = vl; + ++used; // Using two shorts, so increased used again. + } + else if (ch >= 0xD800 && ch <= 0xDFFF) + { + // Between possible UTF-16 surrogates (invalid!) + array[pos++] = unicode::UTF_REPLACEMENT_CHARACTER; + } + else array[pos++] = static_cast<uchar16_t>(ch); + } + array[used] = 0; + + // Validate our new UTF-16 string. + validate(); + + return *this; + } + + + //! Appends a ustring16 to this ustring16 + //! \param other The string to append to this one. + //! \return A reference to our current string. + ustring16<TAlloc>& append(const ustring16<TAlloc>& other) + { + const uchar16_t* oa = other.c_str(); + + u32 len = other.size_raw(); + + if (used + len >= allocated) + reallocate(used + len); + + for (u32 l=0; l<len; ++l) + array[used+l] = oa[l]; + + used += len; + array[used] = 0; + + return *this; + } + + + //! Appends a certain amount of characters of a ustring16 to this ustring16. + //! \param other The string to append to this one. + //! \param length How many characters of the other string to add to this one. + //! \return A reference to our current string. + ustring16<TAlloc>& append(const ustring16<TAlloc>& other, u32 length) + { + if (other.size() == 0) + return *this; + + if (other.size() < length) + { + append(other); + return *this; + } + + if (used + length * 2 >= allocated) + reallocate(used + length * 2); + + const_iterator iter(other, 0); + u32 l = length; + while (!iter.atEnd() && l) + { + uchar32_t c = *iter; + append(c); + ++iter; + --l; + } + + return *this; + } + + + //! Reserves some memory. + //! \param count The amount of characters to reserve. + void reserve(u32 count) + { + if (count < allocated) + return; + + reallocate(count); + } + + + //! Finds first occurrence of character. + //! \param c The character to search for. + //! \return Position where the character has been found, or -1 if not found. + s32 findFirst(uchar32_t c) const + { + const_iterator i(*this, 0); + + s32 pos = 0; + while (!i.atEnd()) + { + uchar32_t t = *i; + if (c == t) + return pos; + ++pos; + ++i; + } + + return -1; + } + + //! Finds first occurrence of a character of a list. + //! \param c A list of characters to find. For example if the method should find the first occurrence of 'a' or 'b', this parameter should be "ab". + //! \param count The amount of characters in the list. Usually, this should be strlen(c). + //! \return Position where one of the characters has been found, or -1 if not found. + s32 findFirstChar(const uchar32_t* const c, u32 count=1) const + { + if (!c || !count) + return -1; + + const_iterator i(*this, 0); + + s32 pos = 0; + while (!i.atEnd()) + { + uchar32_t t = *i; + for (u32 j=0; j<count; ++j) + if (t == c[j]) + return pos; + ++pos; + ++i; + } + + return -1; + } + + + //! Finds first position of a character not in a given list. + //! \param c A list of characters to NOT find. For example if the method should find the first occurrence of a character not 'a' or 'b', this parameter should be "ab". + //! \param count The amount of characters in the list. Usually, this should be strlen(c). + //! \return Position where the character has been found, or -1 if not found. + s32 findFirstCharNotInList(const uchar32_t* const c, u32 count=1) const + { + if (!c || !count) + return -1; + + const_iterator i(*this, 0); + + s32 pos = 0; + while (!i.atEnd()) + { + uchar32_t t = *i; + u32 j; + for (j=0; j<count; ++j) + if (t == c[j]) + break; + + if (j==count) + return pos; + ++pos; + ++i; + } + + return -1; + } + + //! Finds last position of a character not in a given list. + //! \param c A list of characters to NOT find. For example if the method should find the first occurrence of a character not 'a' or 'b', this parameter should be "ab". + //! \param count The amount of characters in the list. Usually, this should be strlen(c). + //! \return Position where the character has been found, or -1 if not found. + s32 findLastCharNotInList(const uchar32_t* const c, u32 count=1) const + { + if (!c || !count) + return -1; + + const_iterator i(end()); + --i; + + s32 pos = size() - 1; + while (!i.atStart()) + { + uchar32_t t = *i; + u32 j; + for (j=0; j<count; ++j) + if (t == c[j]) + break; + + if (j==count) + return pos; + --pos; + --i; + } + + return -1; + } + + //! Finds next occurrence of character. + //! \param c The character to search for. + //! \param startPos The position in the string to start searching. + //! \return Position where the character has been found, or -1 if not found. + s32 findNext(uchar32_t c, u32 startPos) const + { + const_iterator i(*this, startPos); + + s32 pos = startPos; + while (!i.atEnd()) + { + uchar32_t t = *i; + if (t == c) + return pos; + ++pos; + ++i; + } + + return -1; + } + + + //! Finds last occurrence of character. + //! \param c The character to search for. + //! \param start The start position of the reverse search ( default = -1, on end ). + //! \return Position where the character has been found, or -1 if not found. + s32 findLast(uchar32_t c, s32 start = -1) const + { + u32 s = size(); + start = core::clamp ( start < 0 ? (s32)s : start, 0, (s32)s ) - 1; + + const_iterator i(*this, start); + u32 pos = start; + while (!i.atStart()) + { + uchar32_t t = *i; + if (t == c) + return pos; + --pos; + --i; + } + + return -1; + } + + //! Finds last occurrence of a character in a list. + //! \param c A list of strings to find. For example if the method should find the last occurrence of 'a' or 'b', this parameter should be "ab". + //! \param count The amount of characters in the list. Usually, this should be strlen(c). + //! \return Position where one of the characters has been found, or -1 if not found. + s32 findLastChar(const uchar32_t* const c, u32 count=1) const + { + if (!c || !count) + return -1; + + const_iterator i(end()); + --i; + + s32 pos = size(); + while (!i.atStart()) + { + uchar32_t t = *i; + for (u32 j=0; j<count; ++j) + if (t == c[j]) + return pos; + --pos; + --i; + } + + return -1; + } + + + //! Finds another ustring16 in this ustring16. + //! \param str The string to find. + //! \param start The start position of the search. + //! \return Positions where the ustring16 has been found, or -1 if not found. + s32 find(const ustring16<TAlloc>& str, const u32 start = 0) const + { + u32 my_size = size(); + u32 their_size = str.size(); + + if (their_size == 0 || my_size - start < their_size) + return -1; + + const_iterator i(*this, start); + + s32 pos = start; + while (!i.atEnd()) + { + const_iterator i2(i); + const_iterator j(str, 0); + uchar32_t t1 = (uchar32_t)*i2; + uchar32_t t2 = (uchar32_t)*j; + while (t1 == t2) + { + ++i2; + ++j; + if (j.atEnd()) + return pos; + t1 = (uchar32_t)*i2; + t2 = (uchar32_t)*j; + } + ++i; + ++pos; + } + + return -1; + } + + + //! Finds another ustring16 in this ustring16. + //! \param str The string to find. + //! \param start The start position of the search. + //! \return Positions where the string has been found, or -1 if not found. + s32 find_raw(const ustring16<TAlloc>& str, const u32 start = 0) const + { + const uchar16_t* data = str.c_str(); + if (data && *data) + { + u32 len = 0; + + while (data[len]) + ++len; + + if (len > used) + return -1; + + for (u32 i=start; i<=used-len; ++i) + { + u32 j=0; + + while(data[j] && array[i+j] == data[j]) + ++j; + + if (!data[j]) + return i; + } + } + + return -1; + } + + + //! Returns a substring. + //! \param begin: Start of substring. + //! \param length: Length of substring. + //! \return A reference to our current string. + ustring16<TAlloc> subString(u32 begin, s32 length) const + { + u32 len = size(); + // if start after ustring16 + // or no proper substring length + if ((length <= 0) || (begin>=len)) + return ustring16<TAlloc>(""); + // clamp length to maximal value + if ((length+begin) > len) + length = len-begin; + + ustring16<TAlloc> o; + o.reserve((length+1) * 2); + + const_iterator i(*this, begin); + while (!i.atEnd() && length) + { + o.append(*i); + ++i; + --length; + } + + return o; + } + + + //! Appends a character to this ustring16. + //! \param c Character to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (char c) + { + append((uchar32_t)c); + return *this; + } + + + //! Appends a character to this ustring16. + //! \param c Character to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (uchar32_t c) + { + append(c); + return *this; + } + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (short c) + { + append(core::stringc(c)); + return *this; + } + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (unsigned short c) + { + append(core::stringc(c)); + return *this; + } + + +#ifdef USTRING_CPP0X_NEWLITERALS + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (int c) + { + append(core::stringc(c)); + return *this; + } + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (unsigned int c) + { + append(core::stringc(c)); + return *this; + } +#endif + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (long c) + { + append(core::stringc(c)); + return *this; + } + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (unsigned long c) + { + append(core::stringc(c)); + return *this; + } + + + //! Appends a number to this ustring16. + //! \param c Number to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (double c) + { + append(core::stringc(c)); + return *this; + } + + + //! Appends a char ustring16 to this ustring16. + //! \param c Char ustring16 to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (const uchar16_t* const c) + { + append(c); + return *this; + } + + + //! Appends a ustring16 to this ustring16. + //! \param other ustring16 to append. + //! \return A reference to our current string. + ustring16<TAlloc>& operator += (const ustring16<TAlloc>& other) + { + append(other); + return *this; + } + + + //! Replaces all characters of a given type with another one. + //! \param toReplace Character to replace. + //! \param replaceWith Character replacing the old one. + //! \return A reference to our current string. + ustring16<TAlloc>& replace(uchar32_t toReplace, uchar32_t replaceWith) + { + iterator i(*this, 0); + while (!i.atEnd()) + { + typename ustring16<TAlloc>::access a = *i; + if ((uchar32_t)a == toReplace) + a = replaceWith; + ++i; + } + return *this; + } + + + //! Replaces all instances of a string with another one. + //! \param toReplace The string to replace. + //! \param replaceWith The string replacing the old one. + //! \return A reference to our current string. + ustring16<TAlloc>& replace(const ustring16<TAlloc>& toReplace, const ustring16<TAlloc>& replaceWith) + { + if (toReplace.size() == 0) + return *this; + + const uchar16_t* other = toReplace.c_str(); + const uchar16_t* replace = replaceWith.c_str(); + const u32 other_size = toReplace.size_raw(); + const u32 replace_size = replaceWith.size_raw(); + + // Determine the delta. The algorithm will change depending on the delta. + s32 delta = replace_size - other_size; + + // A character for character replace. The string will not shrink or grow. + if (delta == 0) + { + s32 pos = 0; + while ((pos = find_raw(other, pos)) != -1) + { + for (u32 i = 0; i < replace_size; ++i) + array[pos + i] = replace[i]; + ++pos; + } + return *this; + } + + // We are going to be removing some characters. The string will shrink. + if (delta < 0) + { + u32 i = 0; + for (u32 pos = 0; pos <= used; ++i, ++pos) + { + // Is this potentially a match? + if (array[pos] == *other) + { + // Check to see if we have a match. + u32 j; + for (j = 0; j < other_size; ++j) + { + if (array[pos + j] != other[j]) + break; + } + + // If we have a match, replace characters. + if (j == other_size) + { + for (j = 0; j < replace_size; ++j) + array[i + j] = replace[j]; + i += replace_size - 1; + pos += other_size - 1; + continue; + } + } + + // No match found, just copy characters. + array[i - 1] = array[pos]; + } + array[i] = 0; + used = i; + + return *this; + } + + // We are going to be adding characters, so the string size will increase. + // Count the number of times toReplace exists in the string so we can allocate the new size. + u32 find_count = 0; + s32 pos = 0; + while ((pos = find_raw(other, pos)) != -1) + { + ++find_count; + ++pos; + } + + // Re-allocate the string now, if needed. + u32 len = delta * find_count; + if (used + len >= allocated) + reallocate(used + len); + + // Start replacing. + pos = 0; + while ((pos = find_raw(other, pos)) != -1) + { + uchar16_t* start = array + pos + other_size - 1; + uchar16_t* ptr = array + used; + uchar16_t* end = array + used + delta; + + // Shift characters to make room for the string. + while (ptr != start) + { + *end = *ptr; + --ptr; + --end; + } + + // Add the new string now. + for (u32 i = 0; i < replace_size; ++i) + array[pos + i] = replace[i]; + + pos += replace_size; + used += delta; + } + + // Terminate the string and return ourself. + array[used] = 0; + return *this; + } + + + //! Removes characters from a ustring16.. + //! \param c The character to remove. + //! \return A reference to our current string. + ustring16<TAlloc>& remove(uchar32_t c) + { + u32 pos = 0; + u32 found = 0; + u32 len = (c > 0xFFFF ? 2 : 1); // Remove characters equal to the size of c as a UTF-16 character. + for (u32 i=0; i<=used; ++i) + { + uchar32_t uc32 = 0; + if (!UTF16_IS_SURROGATE_HI(array[i])) + uc32 |= array[i]; + else if (i + 1 <= used) + { + // Convert the surrogate pair into a single UTF-32 character. + uc32 = unicode::toUTF32(array[i], array[i + 1]); + } + u32 len2 = (uc32 > 0xFFFF ? 2 : 1); + + if (uc32 == c) + { + found += len; + continue; + } + + array[pos++] = array[i]; + if (len2 == 2) + array[pos++] = array[++i]; + } + used -= found; + array[used] = 0; + return *this; + } + + + //! Removes a ustring16 from the ustring16. + //! \param toRemove The string to remove. + //! \return A reference to our current string. + ustring16<TAlloc>& remove(const ustring16<TAlloc>& toRemove) + { + u32 size = toRemove.size_raw(); + if (size == 0) return *this; + + const uchar16_t* tra = toRemove.c_str(); + u32 pos = 0; + u32 found = 0; + for (u32 i=0; i<=used; ++i) + { + u32 j = 0; + while (j < size) + { + if (array[i + j] != tra[j]) + break; + ++j; + } + if (j == size) + { + found += size; + i += size - 1; + continue; + } + + array[pos++] = array[i]; + } + used -= found; + array[used] = 0; + return *this; + } + + + //! Removes characters from the ustring16. + //! \param characters The characters to remove. + //! \return A reference to our current string. + ustring16<TAlloc>& removeChars(const ustring16<TAlloc>& characters) + { + if (characters.size_raw() == 0) + return *this; + + u32 pos = 0; + u32 found = 0; + const_iterator iter(characters); + for (u32 i=0; i<=used; ++i) + { + uchar32_t uc32 = 0; + if (!UTF16_IS_SURROGATE_HI(array[i])) + uc32 |= array[i]; + else if (i + 1 <= used) + { + // Convert the surrogate pair into a single UTF-32 character. + uc32 = unicode::toUTF32(array[i], array[i+1]); + } + u32 len2 = (uc32 > 0xFFFF ? 2 : 1); + + bool cont = false; + iter.toStart(); + while (!iter.atEnd()) + { + uchar32_t c = *iter; + if (uc32 == c) + { + found += (c > 0xFFFF ? 2 : 1); // Remove characters equal to the size of c as a UTF-16 character. + ++i; + cont = true; + break; + } + ++iter; + } + if (cont) continue; + + array[pos++] = array[i]; + if (len2 == 2) + array[pos++] = array[++i]; + } + used -= found; + array[used] = 0; + return *this; + } + + + //! Trims the ustring16. + //! Removes the specified characters (by default, Latin-1 whitespace) from the begining and the end of the ustring16. + //! \param whitespace The characters that are to be considered as whitespace. + //! \return A reference to our current string. + ustring16<TAlloc>& trim(const ustring16<TAlloc>& whitespace = " \t\n\r") + { + core::array<uchar32_t> utf32white = whitespace.toUTF32(); + + // find start and end of the substring without the specified characters + const s32 begin = findFirstCharNotInList(utf32white.const_pointer(), whitespace.used + 1); + if (begin == -1) + return (*this=""); + + const s32 end = findLastCharNotInList(utf32white.const_pointer(), whitespace.used + 1); + + return (*this = subString(begin, (end +1) - begin)); + } + + + //! Erases a character from the ustring16. + //! May be slow, because all elements following after the erased element have to be copied. + //! \param index Index of element to be erased. + //! \return A reference to our current string. + ustring16<TAlloc>& erase(u32 index) + { + _IRR_DEBUG_BREAK_IF(index>used) // access violation + + iterator i(*this, index); + + uchar32_t t = *i; + u32 len = (t > 0xFFFF ? 2 : 1); + + for (u32 j = static_cast<u32>(i.getPos()) + len; j <= used; ++j) + array[j - len] = array[j]; + + used -= len; + array[used] = 0; + + return *this; + } + + + //! Validate the existing ustring16, checking for valid surrogate pairs and checking for proper termination. + //! \return A reference to our current string. + ustring16<TAlloc>& validate() + { + // Validate all unicode characters. + for (u32 i=0; i<allocated; ++i) + { + // Terminate on existing null. + if (array[i] == 0) + { + used = i; + return *this; + } + if (UTF16_IS_SURROGATE(array[i])) + { + if (((i+1) >= allocated) || UTF16_IS_SURROGATE_LO(array[i])) + array[i] = unicode::UTF_REPLACEMENT_CHARACTER; + else if (UTF16_IS_SURROGATE_HI(array[i]) && !UTF16_IS_SURROGATE_LO(array[i+1])) + array[i] = unicode::UTF_REPLACEMENT_CHARACTER; + ++i; + } + if (array[i] >= 0xFDD0 && array[i] <= 0xFDEF) + array[i] = unicode::UTF_REPLACEMENT_CHARACTER; + } + + // terminate + used = 0; + if (allocated > 0) + { + used = allocated - 1; + array[used] = 0; + } + return *this; + } + + + //! Gets the last char of the ustring16, or 0. + //! \return The last char of the ustring16, or 0. + uchar32_t lastChar() const + { + if (used < 1) + return 0; + + if (UTF16_IS_SURROGATE_LO(array[used-1])) + { + // Make sure we have a paired surrogate. + if (used < 2) + return 0; + + // Check for an invalid surrogate. + if (!UTF16_IS_SURROGATE_HI(array[used-2])) + return 0; + + // Convert the surrogate pair into a single UTF-32 character. + return unicode::toUTF32(array[used-2], array[used-1]); + } + else + { + return array[used-1]; + } + } + + + //! Split the ustring16 into parts. + /** This method will split a ustring16 at certain delimiter characters + into the container passed in as reference. The type of the container + has to be given as template parameter. It must provide a push_back and + a size method. + \param ret The result container + \param c C-style ustring16 of delimiter characters + \param count Number of delimiter characters + \param ignoreEmptyTokens Flag to avoid empty substrings in the result + container. If two delimiters occur without a character in between, an + empty substring would be placed in the result. If this flag is set, + only non-empty strings are stored. + \param keepSeparators Flag which allows to add the separator to the + result ustring16. If this flag is true, the concatenation of the + substrings results in the original ustring16. Otherwise, only the + characters between the delimiters are returned. + \return The number of resulting substrings + */ + template<class container> + u32 split(container& ret, const uchar32_t* const c, u32 count=1, bool ignoreEmptyTokens=true, bool keepSeparators=false) const + { + if (!c) + return 0; + + const_iterator i(*this); + const u32 oldSize=ret.size(); + u32 pos = 0; + u32 lastpos = 0; + u32 lastpospos = 0; + bool lastWasSeparator = false; + while (!i.atEnd()) + { + uchar32_t ch = *i; + bool foundSeparator = false; + for (u32 j=0; j<count; ++j) + { + if (ch == c[j]) + { + if ((!ignoreEmptyTokens || pos - lastpos != 0) && + !lastWasSeparator) + ret.push_back(ustring16<TAlloc>(&array[lastpospos], pos - lastpos)); + foundSeparator = true; + lastpos = (keepSeparators ? pos : pos + 1); + lastpospos = (keepSeparators ? i.getPos() : i.getPos() + 1); + break; + } + } + lastWasSeparator = foundSeparator; + ++pos; + ++i; + } + u32 s = size() + 1; + if (s > lastpos) + ret.push_back(ustring16<TAlloc>(&array[lastpospos], s - lastpos)); + return ret.size()-oldSize; + } + + + //! Split the ustring16 into parts. + /** This method will split a ustring16 at certain delimiter characters + into the container passed in as reference. The type of the container + has to be given as template parameter. It must provide a push_back and + a size method. + \param ret The result container + \param c A unicode string of delimiter characters + \param ignoreEmptyTokens Flag to avoid empty substrings in the result + container. If two delimiters occur without a character in between, an + empty substring would be placed in the result. If this flag is set, + only non-empty strings are stored. + \param keepSeparators Flag which allows to add the separator to the + result ustring16. If this flag is true, the concatenation of the + substrings results in the original ustring16. Otherwise, only the + characters between the delimiters are returned. + \return The number of resulting substrings + */ + template<class container> + u32 split(container& ret, const ustring16<TAlloc>& c, bool ignoreEmptyTokens=true, bool keepSeparators=false) const + { + core::array<uchar32_t> v = c.toUTF32(); + return split(ret, v.pointer(), v.size(), ignoreEmptyTokens, keepSeparators); + } + + + //! Gets the size of the allocated memory buffer for the string. + //! \return The size of the allocated memory buffer. + u32 capacity() const + { + return allocated; + } + + + //! Returns the raw number of UTF-16 code points in the string which includes the individual surrogates. + //! \return The raw number of UTF-16 code points, excluding the trialing NUL. + u32 size_raw() const + { + return used; + } + + + //! Inserts a character into the string. + //! \param c The character to insert. + //! \param pos The position to insert the character. + //! \return A reference to our current string. + ustring16<TAlloc>& insert(uchar32_t c, u32 pos) + { + u8 len = (c > 0xFFFF ? 2 : 1); + + if (used + len >= allocated) + reallocate(used + len); + + used += len; + + iterator iter(*this, pos); + for (u32 i = used - 2; i > iter.getPos(); --i) + array[i] = array[i - len]; + + if (c > 0xFFFF) + { + // c will be multibyte, so split it up into a surrogate pair. + uchar16_t x = static_cast<uchar16_t>(c); + uchar16_t vh = UTF16_HI_SURROGATE | ((((c >> 16) & ((1 << 5) - 1)) - 1) << 6) | (x >> 10); + uchar16_t vl = UTF16_LO_SURROGATE | (x & ((1 << 10) - 1)); + array[iter.getPos()] = vh; + array[iter.getPos()+1] = vl; + } + else + { + array[iter.getPos()] = static_cast<uchar16_t>(c); + } + array[used] = 0; + return *this; + } + + + //! Inserts a string into the string. + //! \param c The string to insert. + //! \param pos The position to insert the string. + //! \return A reference to our current string. + ustring16<TAlloc>& insert(const ustring16<TAlloc>& c, u32 pos) + { + u32 len = c.size_raw(); + if (len == 0) return *this; + + if (used + len >= allocated) + reallocate(used + len); + + used += len; + + iterator iter(*this, pos); + for (u32 i = used - 2; i > iter.getPos() + len; --i) + array[i] = array[i - len]; + + const uchar16_t* s = c.c_str(); + for (u32 i = 0; i < len; ++i) + { + array[pos++] = *s; + ++s; + } + + array[used] = 0; + return *this; + } + + + //! Inserts a character into the string. + //! \param c The character to insert. + //! \param pos The position to insert the character. + //! \return A reference to our current string. + ustring16<TAlloc>& insert_raw(uchar16_t c, u32 pos) + { + if (used + 1 >= allocated) + reallocate(used + 1); + + ++used; + + for (u32 i = used - 1; i > pos; --i) + array[i] = array[i - 1]; + + array[pos] = c; + array[used] = 0; + return *this; + } + + + //! Removes a character from string. + //! \param pos Position of the character to remove. + //! \return A reference to our current string. + ustring16<TAlloc>& erase_raw(u32 pos) + { + for (u32 i=pos; i<=used; ++i) + { + array[i] = array[i + 1]; + } + --used; + array[used] = 0; + return *this; + } + + + //! Replaces a character in the string. + //! \param c The new character. + //! \param pos The position of the character to replace. + //! \return A reference to our current string. + ustring16<TAlloc>& replace_raw(uchar16_t c, u32 pos) + { + array[pos] = c; + return *this; + } + + + //! Returns an iterator to the beginning of the string. + //! \return An iterator to the beginning of the string. + iterator begin() + { + iterator i(*this, 0); + return i; + } + + + //! Returns an iterator to the beginning of the string. + //! \return An iterator to the beginning of the string. + const_iterator begin() const + { + const_iterator i(*this, 0); + return i; + } + + + //! Returns an iterator to the beginning of the string. + //! \return An iterator to the beginning of the string. + const_iterator cbegin() const + { + const_iterator i(*this, 0); + return i; + } + + + //! Returns an iterator to the end of the string. + //! \return An iterator to the end of the string. + iterator end() + { + iterator i(*this, 0); + i.toEnd(); + return i; + } + + + //! Returns an iterator to the end of the string. + //! \return An iterator to the end of the string. + const_iterator end() const + { + const_iterator i(*this, 0); + i.toEnd(); + return i; + } + + + //! Returns an iterator to the end of the string. + //! \return An iterator to the end of the string. + const_iterator cend() const + { + const_iterator i(*this, 0); + i.toEnd(); + return i; + } + + + //! Converts the string to a UTF-8 encoded string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return A string containing the UTF-8 encoded string. + core::string<uchar8_t> toUTF8_s(const bool addBOM = false) const + { + core::string<uchar8_t> ret; + ret.reserve(used * 4 + (addBOM ? unicode::BOM_UTF8_LEN : 0) + 1); + const_iterator iter(*this, 0); + + // Add the byte order mark if the user wants it. + if (addBOM) + { + ret.append(unicode::BOM_ENCODE_UTF8[0]); + ret.append(unicode::BOM_ENCODE_UTF8[1]); + ret.append(unicode::BOM_ENCODE_UTF8[2]); + } + + while (!iter.atEnd()) + { + uchar32_t c = *iter; + if (c > 0xFFFF) + { // 4 bytes + uchar8_t b1 = (0x1E << 3) | ((c >> 18) & 0x7); + uchar8_t b2 = (0x2 << 6) | ((c >> 12) & 0x3F); + uchar8_t b3 = (0x2 << 6) | ((c >> 6) & 0x3F); + uchar8_t b4 = (0x2 << 6) | (c & 0x3F); + ret.append(b1); + ret.append(b2); + ret.append(b3); + ret.append(b4); + } + else if (c > 0x7FF) + { // 3 bytes + uchar8_t b1 = (0xE << 4) | ((c >> 12) & 0xF); + uchar8_t b2 = (0x2 << 6) | ((c >> 6) & 0x3F); + uchar8_t b3 = (0x2 << 6) | (c & 0x3F); + ret.append(b1); + ret.append(b2); + ret.append(b3); + } + else if (c > 0x7F) + { // 2 bytes + uchar8_t b1 = (0x6 << 5) | ((c >> 6) & 0x1F); + uchar8_t b2 = (0x2 << 6) | (c & 0x3F); + ret.append(b1); + ret.append(b2); + } + else + { // 1 byte + ret.append(static_cast<uchar8_t>(c)); + } + ++iter; + } + return ret; + } + + + //! Converts the string to a UTF-8 encoded string array. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return An array containing the UTF-8 encoded string. + core::array<uchar8_t> toUTF8(const bool addBOM = false) const + { + core::array<uchar8_t> ret(used * 4 + (addBOM ? unicode::BOM_UTF8_LEN : 0) + 1); + const_iterator iter(*this, 0); + + // Add the byte order mark if the user wants it. + if (addBOM) + { + ret.push_back(unicode::BOM_ENCODE_UTF8[0]); + ret.push_back(unicode::BOM_ENCODE_UTF8[1]); + ret.push_back(unicode::BOM_ENCODE_UTF8[2]); + } + + while (!iter.atEnd()) + { + uchar32_t c = *iter; + if (c > 0xFFFF) + { // 4 bytes + uchar8_t b1 = (0x1E << 3) | ((c >> 18) & 0x7); + uchar8_t b2 = (0x2 << 6) | ((c >> 12) & 0x3F); + uchar8_t b3 = (0x2 << 6) | ((c >> 6) & 0x3F); + uchar8_t b4 = (0x2 << 6) | (c & 0x3F); + ret.push_back(b1); + ret.push_back(b2); + ret.push_back(b3); + ret.push_back(b4); + } + else if (c > 0x7FF) + { // 3 bytes + uchar8_t b1 = (0xE << 4) | ((c >> 12) & 0xF); + uchar8_t b2 = (0x2 << 6) | ((c >> 6) & 0x3F); + uchar8_t b3 = (0x2 << 6) | (c & 0x3F); + ret.push_back(b1); + ret.push_back(b2); + ret.push_back(b3); + } + else if (c > 0x7F) + { // 2 bytes + uchar8_t b1 = (0x6 << 5) | ((c >> 6) & 0x1F); + uchar8_t b2 = (0x2 << 6) | (c & 0x3F); + ret.push_back(b1); + ret.push_back(b2); + } + else + { // 1 byte + ret.push_back(static_cast<uchar8_t>(c)); + } + ++iter; + } + ret.push_back(0); + return ret; + } + + +#ifdef USTRING_CPP0X_NEWLITERALS // C++0x + //! Converts the string to a UTF-16 encoded string. + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return A string containing the UTF-16 encoded string. + core::string<char16_t> toUTF16_s(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + core::string<char16_t> ret; + ret.reserve(used + (addBOM ? unicode::BOM_UTF16_LEN : 0) + 1); + + // Add the BOM if specified. + if (addBOM) + { + if (endian == unicode::EUTFEE_NATIVE) + ret[0] = unicode::BOM; + else if (endian == unicode::EUTFEE_LITTLE) + { + uchar8_t* ptr8 = reinterpret_cast<uchar8_t*>(ret.c_str()); + *ptr8++ = unicode::BOM_ENCODE_UTF16_LE[0]; + *ptr8 = unicode::BOM_ENCODE_UTF16_LE[1]; + } + else + { + uchar8_t* ptr8 = reinterpret_cast<uchar8_t*>(ret.c_str()); + *ptr8++ = unicode::BOM_ENCODE_UTF16_BE[0]; + *ptr8 = unicode::BOM_ENCODE_UTF16_BE[1]; + } + } + + ret.append(array); + if (endian != unicode::EUTFEE_NATIVE && getEndianness() != endian) + { + char16_t* ptr = ret.c_str(); + for (u32 i = 0; i < ret.size(); ++i) + *ptr++ = unicode::swapEndian16(*ptr); + } + return ret; + } +#endif + + + //! Converts the string to a UTF-16 encoded string array. + //! Unfortunately, no toUTF16_s() version exists due to limitations with Irrlicht's string class. + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return An array containing the UTF-16 encoded string. + core::array<uchar16_t> toUTF16(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + core::array<uchar16_t> ret(used + (addBOM ? unicode::BOM_UTF16_LEN : 0) + 1); + uchar16_t* ptr = ret.pointer(); + + // Add the BOM if specified. + if (addBOM) + { + if (endian == unicode::EUTFEE_NATIVE) + *ptr = unicode::BOM; + else if (endian == unicode::EUTFEE_LITTLE) + { + uchar8_t* ptr8 = reinterpret_cast<uchar8_t*>(ptr); + *ptr8++ = unicode::BOM_ENCODE_UTF16_LE[0]; + *ptr8 = unicode::BOM_ENCODE_UTF16_LE[1]; + } + else + { + uchar8_t* ptr8 = reinterpret_cast<uchar8_t*>(ptr); + *ptr8++ = unicode::BOM_ENCODE_UTF16_BE[0]; + *ptr8 = unicode::BOM_ENCODE_UTF16_BE[1]; + } + ++ptr; + } + + memcpy((void*)ptr, (void*)array, used * sizeof(uchar16_t)); + if (endian != unicode::EUTFEE_NATIVE && getEndianness() != endian) + { + for (u32 i = 0; i <= used; ++i) + *ptr++ = unicode::swapEndian16(*ptr); + } + ret.set_used(used + (addBOM ? unicode::BOM_UTF16_LEN : 0)); + ret.push_back(0); + return ret; + } + + +#ifdef USTRING_CPP0X_NEWLITERALS // C++0x + //! Converts the string to a UTF-32 encoded string. + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return A string containing the UTF-32 encoded string. + core::string<char32_t> toUTF32_s(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + core::string<char32_t> ret; + ret.reserve(size() + 1 + (addBOM ? unicode::BOM_UTF32_LEN : 0)); + const_iterator iter(*this, 0); + + // Add the BOM if specified. + if (addBOM) + { + if (endian == unicode::EUTFEE_NATIVE) + ret.append(unicode::BOM); + else + { + union + { + uchar32_t full; + u8 chunk[4]; + } t; + + if (endian == unicode::EUTFEE_LITTLE) + { + t.chunk[0] = unicode::BOM_ENCODE_UTF32_LE[0]; + t.chunk[1] = unicode::BOM_ENCODE_UTF32_LE[1]; + t.chunk[2] = unicode::BOM_ENCODE_UTF32_LE[2]; + t.chunk[3] = unicode::BOM_ENCODE_UTF32_LE[3]; + } + else + { + t.chunk[0] = unicode::BOM_ENCODE_UTF32_BE[0]; + t.chunk[1] = unicode::BOM_ENCODE_UTF32_BE[1]; + t.chunk[2] = unicode::BOM_ENCODE_UTF32_BE[2]; + t.chunk[3] = unicode::BOM_ENCODE_UTF32_BE[3]; + } + ret.append(t.full); + } + } + + while (!iter.atEnd()) + { + uchar32_t c = *iter; + if (endian != unicode::EUTFEE_NATIVE && getEndianness() != endian) + c = unicode::swapEndian32(c); + ret.append(c); + ++iter; + } + return ret; + } +#endif + + + //! Converts the string to a UTF-32 encoded string array. + //! Unfortunately, no toUTF32_s() version exists due to limitations with Irrlicht's string class. + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return An array containing the UTF-32 encoded string. + core::array<uchar32_t> toUTF32(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + core::array<uchar32_t> ret(size() + (addBOM ? unicode::BOM_UTF32_LEN : 0) + 1); + const_iterator iter(*this, 0); + + // Add the BOM if specified. + if (addBOM) + { + if (endian == unicode::EUTFEE_NATIVE) + ret.push_back(unicode::BOM); + else + { + union + { + uchar32_t full; + u8 chunk[4]; + } t; + + if (endian == unicode::EUTFEE_LITTLE) + { + t.chunk[0] = unicode::BOM_ENCODE_UTF32_LE[0]; + t.chunk[1] = unicode::BOM_ENCODE_UTF32_LE[1]; + t.chunk[2] = unicode::BOM_ENCODE_UTF32_LE[2]; + t.chunk[3] = unicode::BOM_ENCODE_UTF32_LE[3]; + } + else + { + t.chunk[0] = unicode::BOM_ENCODE_UTF32_BE[0]; + t.chunk[1] = unicode::BOM_ENCODE_UTF32_BE[1]; + t.chunk[2] = unicode::BOM_ENCODE_UTF32_BE[2]; + t.chunk[3] = unicode::BOM_ENCODE_UTF32_BE[3]; + } + ret.push_back(t.full); + } + } + ret.push_back(0); + + while (!iter.atEnd()) + { + uchar32_t c = *iter; + if (endian != unicode::EUTFEE_NATIVE && getEndianness() != endian) + c = unicode::swapEndian32(c); + ret.push_back(c); + ++iter; + } + return ret; + } + + + //! Converts the string to a wchar_t encoded string. + /** The size of a wchar_t changes depending on the platform. This function will store a + correct UTF-8, -16, or -32 encoded string depending on the size of a wchar_t. **/ + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return A string containing the wchar_t encoded string. + core::string<wchar_t> toWCHAR_s(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + if (sizeof(wchar_t) == 4) + { + core::array<uchar32_t> a(toUTF32(endian, addBOM)); + core::stringw ret(a.pointer()); + return ret; + } + else if (sizeof(wchar_t) == 2) + { + if (endian == unicode::EUTFEE_NATIVE && addBOM == false) + { + core::stringw ret(array); + return ret; + } + else + { + core::array<uchar16_t> a(toUTF16(endian, addBOM)); + core::stringw ret(a.pointer()); + return ret; + } + } + else if (sizeof(wchar_t) == 1) + { + core::array<uchar8_t> a(toUTF8(addBOM)); + core::stringw ret(a.pointer()); + return ret; + } + + // Shouldn't happen. + return core::stringw(); + } + + + //! Converts the string to a wchar_t encoded string array. + /** The size of a wchar_t changes depending on the platform. This function will store a + correct UTF-8, -16, or -32 encoded string depending on the size of a wchar_t. **/ + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return An array containing the wchar_t encoded string. + core::array<wchar_t> toWCHAR(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { + if (sizeof(wchar_t) == 4) + { + core::array<uchar32_t> a(toUTF32(endian, addBOM)); + core::array<wchar_t> ret(a.size()); + ret.set_used(a.size()); + memcpy((void*)ret.pointer(), (void*)a.pointer(), a.size() * sizeof(uchar32_t)); + return ret; + } + if (sizeof(wchar_t) == 2) + { + if (endian == unicode::EUTFEE_NATIVE && addBOM == false) + { + core::array<wchar_t> ret(used); + ret.set_used(used); + memcpy((void*)ret.pointer(), (void*)array, used * sizeof(uchar16_t)); + return ret; + } + else + { + core::array<uchar16_t> a(toUTF16(endian, addBOM)); + core::array<wchar_t> ret(a.size()); + ret.set_used(a.size()); + memcpy((void*)ret.pointer(), (void*)a.pointer(), a.size() * sizeof(uchar16_t)); + return ret; + } + } + if (sizeof(wchar_t) == 1) + { + core::array<uchar8_t> a(toUTF8(addBOM)); + core::array<wchar_t> ret(a.size()); + ret.set_used(a.size()); + memcpy((void*)ret.pointer(), (void*)a.pointer(), a.size() * sizeof(uchar8_t)); + return ret; + } + + // Shouldn't happen. + return core::array<wchar_t>(); + } + + //! Converts the string to a properly encoded io::path string. + //! \param endian The desired endianness of the string. + //! \param addBOM If true, the proper unicode byte-order mark will be prefixed to the string. + //! \return An io::path string containing the properly encoded string. + io::path toPATH_s(const unicode::EUTF_ENDIAN endian = unicode::EUTFEE_NATIVE, const bool addBOM = false) const + { +#if defined(_IRR_WCHAR_FILESYSTEM) + return toWCHAR_s(endian, addBOM); +#else + return toUTF8_s(addBOM); +#endif + } + + //! Loads an unknown stream of data. + //! Will attempt to determine if the stream is unicode data. Useful for loading from files. + //! \param data The data stream to load from. + //! \param data_size The length of the data string. + //! \return A reference to our current string. + ustring16<TAlloc>& loadDataStream(const char* data, size_t data_size) + { + // Clear our string. + *this = ""; + if (!data) + return *this; + + unicode::EUTF_ENCODE e = unicode::determineUnicodeBOM(data); + switch (e) + { + default: + case unicode::EUTFE_UTF8: + append((uchar8_t*)data, data_size); + break; + + case unicode::EUTFE_UTF16: + case unicode::EUTFE_UTF16_BE: + case unicode::EUTFE_UTF16_LE: + append((uchar16_t*)data, data_size / 2); + break; + + case unicode::EUTFE_UTF32: + case unicode::EUTFE_UTF32_BE: + case unicode::EUTFE_UTF32_LE: + append((uchar32_t*)data, data_size / 4); + break; + } + + return *this; + } + + //! Gets the encoding of the Unicode string this class contains. + //! \return An enum describing the current encoding of this string. + const unicode::EUTF_ENCODE getEncoding() const + { + return encoding; + } + + //! Gets the endianness of the Unicode string this class contains. + //! \return An enum describing the endianness of this string. + const unicode::EUTF_ENDIAN getEndianness() const + { + if (encoding == unicode::EUTFE_UTF16_LE || + encoding == unicode::EUTFE_UTF32_LE) + return unicode::EUTFEE_LITTLE; + else return unicode::EUTFEE_BIG; + } + +private: + + //! Reallocate the string, making it bigger or smaller. + //! \param new_size The new size of the string. + void reallocate(u32 new_size) + { + uchar16_t* old_array = array; + + array = allocator.allocate(new_size + 1); //new u16[new_size]; + allocated = new_size + 1; + if (old_array == 0) return; + + u32 amount = used < new_size ? used : new_size; + for (u32 i=0; i<=amount; ++i) + array[i] = old_array[i]; + + if (allocated <= used) + used = allocated - 1; + + array[used] = 0; + + allocator.deallocate(old_array); // delete [] old_array; + } + + //--- member variables + + uchar16_t* array; + unicode::EUTF_ENCODE encoding; + u32 allocated; + u32 used; + TAlloc allocator; + //irrAllocator<uchar16_t> allocator; +}; + +typedef ustring16<irrAllocator<uchar16_t> > ustring; + + +//! Appends two ustring16s. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a null-terminated unicode string. +template <typename TAlloc, class B> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const B* const right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a null-terminated unicode string. +template <class B, typename TAlloc> +inline ustring16<TAlloc> operator+(const B* const left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an Irrlicht string. +template <typename TAlloc, typename B, typename BAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const string<B, BAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an Irrlicht string. +template <typename TAlloc, typename B, typename BAlloc> +inline ustring16<TAlloc> operator+(const string<B, BAlloc>& left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a std::basic_string. +template <typename TAlloc, typename B, typename A, typename BAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const std::basic_string<B, A, BAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a std::basic_string. +template <typename TAlloc, typename B, typename A, typename BAlloc> +inline ustring16<TAlloc> operator+(const std::basic_string<B, A, BAlloc>& left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a char. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const char right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a char. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const char left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +#ifdef USTRING_CPP0X_NEWLITERALS +//! Appends a ustring16 and a uchar32_t. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const uchar32_t right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a uchar32_t. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const uchar32_t left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(left); + ret += right; + return ret; +} +#endif + + +//! Appends a ustring16 and a short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const short right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and a short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const short left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an unsigned short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const unsigned short right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and an unsigned short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned short left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const int right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and an int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const int left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an unsigned int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const unsigned int right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and an unsigned int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned int left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const long right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and a long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const long left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and an unsigned long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const unsigned long right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and an unsigned long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned long left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a float. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const float right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and a float. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const float left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +//! Appends a ustring16 and a double. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const ustring16<TAlloc>& left, const double right) +{ + ustring16<TAlloc> ret(left); + ret += core::stringc(right); + return ret; +} + + +//! Appends a ustring16 and a double. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const double left, const ustring16<TAlloc>& right) +{ + ustring16<TAlloc> ret(core::stringc(left)); + ret += right; + return ret; +} + + +#ifdef USTRING_CPP0X +//! Appends two ustring16s. +template <typename TAlloc> +inline ustring16<TAlloc>&& operator+(const ustring16<TAlloc>& left, ustring16<TAlloc>&& right) +{ + //std::cout << "MOVE operator+(&, &&)" << std::endl; + right.insert(left, 0); + return std::move(right); +} + + +//! Appends two ustring16s. +template <typename TAlloc> +inline ustring16<TAlloc>&& operator+(ustring16<TAlloc>&& left, const ustring16<TAlloc>& right) +{ + //std::cout << "MOVE operator+(&&, &)" << std::endl; + left.append(right); + return std::move(left); +} + + +//! Appends two ustring16s. +template <typename TAlloc> +inline ustring16<TAlloc>&& operator+(ustring16<TAlloc>&& left, ustring16<TAlloc>&& right) +{ + //std::cout << "MOVE operator+(&&, &&)" << std::endl; + if ((right.size_raw() <= left.capacity() - left.size_raw()) || + (right.capacity() - right.size_raw() < left.size_raw())) + { + left.append(right); + return std::move(left); + } + else + { + right.insert(left, 0); + return std::move(right); + } +} + + +//! Appends a ustring16 and a null-terminated unicode string. +template <typename TAlloc, class B> +inline ustring16<TAlloc>&& operator+(ustring16<TAlloc>&& left, const B* const right) +{ + //std::cout << "MOVE operator+(&&, B*)" << std::endl; + left.append(right); + return std::move(left); +} + + +//! Appends a ustring16 and a null-terminated unicode string. +template <class B, typename TAlloc> +inline ustring16<TAlloc>&& operator+(const B* const left, ustring16<TAlloc>&& right) +{ + //std::cout << "MOVE operator+(B*, &&)" << std::endl; + right.insert(left, 0); + return std::move(right); +} + + +//! Appends a ustring16 and an Irrlicht string. +template <typename TAlloc, typename B, typename BAlloc> +inline ustring16<TAlloc>&& operator+(const string<B, BAlloc>& left, ustring16<TAlloc>&& right) +{ + //std::cout << "MOVE operator+(&, &&)" << std::endl; + right.insert(left, 0); + return std::move(right); +} + + +//! Appends a ustring16 and an Irrlicht string. +template <typename TAlloc, typename B, typename BAlloc> +inline ustring16<TAlloc>&& operator+(ustring16<TAlloc>&& left, const string<B, BAlloc>& right) +{ + //std::cout << "MOVE operator+(&&, &)" << std::endl; + left.append(right); + return std::move(left); +} + + +//! Appends a ustring16 and a std::basic_string. +template <typename TAlloc, typename B, typename A, typename BAlloc> +inline ustring16<TAlloc>&& operator+(const std::basic_string<B, A, BAlloc>& left, ustring16<TAlloc>&& right) +{ + //std::cout << "MOVE operator+(&, &&)" << std::endl; + right.insert(core::ustring16<TAlloc>(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and a std::basic_string. +template <typename TAlloc, typename B, typename A, typename BAlloc> +inline ustring16<TAlloc>&& operator+(ustring16<TAlloc>&& left, const std::basic_string<B, A, BAlloc>& right) +{ + //std::cout << "MOVE operator+(&&, &)" << std::endl; + left.append(right); + return std::move(left); +} + + +//! Appends a ustring16 and a char. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const char right) +{ + left.append((uchar32_t)right); + return std::move(left); +} + + +//! Appends a ustring16 and a char. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const char left, ustring16<TAlloc>&& right) +{ + right.insert((uchar32_t)left, 0); + return std::move(right); +} + + +#ifdef USTRING_CPP0X_NEWLITERALS +//! Appends a ustring16 and a uchar32_t. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const uchar32_t right) +{ + left.append(right); + return std::move(left); +} + + +//! Appends a ustring16 and a uchar32_t. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const uchar32_t left, ustring16<TAlloc>&& right) +{ + right.insert(left, 0); + return std::move(right); +} +#endif + + +//! Appends a ustring16 and a short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const short right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and a short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const short left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and an unsigned short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const unsigned short right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and an unsigned short. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned short left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and an int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const int right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and an int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const int left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and an unsigned int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const unsigned int right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and an unsigned int. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned int left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and a long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const long right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and a long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const long left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and an unsigned long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const unsigned long right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and an unsigned long. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const unsigned long left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and a float. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const float right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and a float. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const float left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} + + +//! Appends a ustring16 and a double. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(ustring16<TAlloc>&& left, const double right) +{ + left.append(core::stringc(right)); + return std::move(left); +} + + +//! Appends a ustring16 and a double. +template <typename TAlloc> +inline ustring16<TAlloc> operator+(const double left, ustring16<TAlloc>&& right) +{ + right.insert(core::stringc(left), 0); + return std::move(right); +} +#endif + + +#ifndef USTRING_NO_STL +//! Writes a ustring16 to an ostream. +template <typename TAlloc> +inline std::ostream& operator<<(std::ostream& out, const ustring16<TAlloc>& in) +{ + out << in.toUTF8_s().c_str(); + return out; +} + +//! Writes a ustring16 to a wostream. +template <typename TAlloc> +inline std::wostream& operator<<(std::wostream& out, const ustring16<TAlloc>& in) +{ + out << in.toWCHAR_s().c_str(); + return out; +} +#endif + + +#ifndef USTRING_NO_STL + +namespace unicode +{ + +//! Hashing algorithm for hashing a ustring. Used for things like unordered_maps. +//! Algorithm taken from std::hash<std::string>. +class hash : public std::unary_function<core::ustring, size_t> +{ + public: + size_t operator()(const core::ustring& s) const + { + size_t ret = 2166136261U; + size_t index = 0; + size_t stride = 1 + s.size_raw() / 10; + + core::ustring::const_iterator i = s.begin(); + while (i != s.end()) + { + // TODO: Don't force u32 on an x64 OS. Make it agnostic. + ret = 16777619U * ret ^ (size_t)s[(u32)index]; + index += stride; + i += stride; + } + return (ret); + } +}; + +} // end namespace unicode + +#endif + +} // end namespace core +} // end namespace irr + +#endif diff --git a/src/cguittfont/xCGUITTFont.cpp b/src/cguittfont/xCGUITTFont.cpp new file mode 100644 index 000000000..c51922e4c --- /dev/null +++ b/src/cguittfont/xCGUITTFont.cpp @@ -0,0 +1,5 @@ +// A wrapper source file to avoid hack with gcc and modifying +// the CGUITTFont files. + +#include "xCGUITTFont.h" +#include "CGUITTFont.cpp" diff --git a/src/cguittfont/xCGUITTFont.h b/src/cguittfont/xCGUITTFont.h new file mode 100644 index 000000000..c3efe7f6f --- /dev/null +++ b/src/cguittfont/xCGUITTFont.h @@ -0,0 +1,7 @@ +// A wrapper header to avoid hack with gcc and modifying +// the CGUITTFont files. + +#include <algorithm> +#include <stddef.h> +#include "irrUString.h" +#include "CGUITTFont.h" |