// CLASS HEADER
#include <dali-toolkit/internal/text/multi-language-support-impl.h>
-// INTERNAL INCLUDES
+// EXTERNAL INCLUDES
+#include <memory.h>
+#include <dali/integration-api/debug.h>
#include <dali/public-api/adaptor-framework/singleton-service.h>
#include <dali/public-api/text-abstraction/font-client.h>
+
+// INTERNAL INCLUDES
#include <dali-toolkit/public-api/text/logical-model.h>
#include <dali-toolkit/public-api/text/font-run.h>
#include <dali-toolkit/public-api/text/script.h>
#include <dali-toolkit/public-api/text/script-run.h>
-#include <dali/integration-api/debug.h>
-
-// EXTERNAL INCLUDES
-#include <memory.h>
namespace Dali
{
return script;
}
+/**
+ * @brief Whether the character is valid for all scripts. i.e. the white space.
+ *
+ * @param[in] character The character.
+ *
+ * @return @e true if the character is valid for all scripts.
+ */
+bool IsValidForAllScripts( Character character )
+{
+ return ( IsWhiteSpace( character ) ||
+ IsZeroWidthNonJoiner( character ) ||
+ IsZeroWidthJoiner( character ) ||
+ IsZeroWidthSpace( character ) ||
+ IsLeftToRightMark( character ) ||
+ IsRightToLeftMark( character ) ||
+ IsThinSpace( character ) );
+}
+
bool ValidateFontsPerScript::FindValidFont( FontId fontId ) const
{
for( Vector<FontId>::ConstIterator it = mValidFonts.Begin(),
}
void MultilanguageSupport::SetScripts( const Vector<Character>& text,
+ const Vector<LineBreakInfo>& lineBreakInfo,
Vector<ScriptRun>& scripts )
{
const Length numberOfCharacters = text.Count();
return;
}
- // Traverse all characters and set the scripts.
-
// Stores the current script run.
ScriptRun currentScriptRun;
currentScriptRun.characterRun.characterIndex = 0u;
// Reserve some space to reduce the number of reallocations.
scripts.Reserve( numberOfCharacters << 2u );
- for( Length index = 0u; index < numberOfCharacters; ++index )
- {
- const Character character = *( text.Begin() + index );
+ // Whether the first valid script need to be set.
+ bool firstValidScript = true;
- Script script = GetCharacterScript( character );
+ // Whether the first valid script is a right to left script.
+ bool isParagraphRTL = false;
- if( TextAbstraction::UNKNOWN == script )
+ // Count the number of characters which are valid for all scripts. i.e. white spaces or '\n'.
+ Length numberOfAllScriptCharacters = 0u;
+
+ // Pointers to the text and break info buffers.
+ const Character* textBuffer = text.Begin();
+ const LineBreakInfo* breakInfoBuffer = lineBreakInfo.Begin();
+
+ // Traverse all characters and set the scripts.
+ for( Length index = 0u; index < numberOfCharacters; ++index )
+ {
+ Character character = *( textBuffer + index );
+ LineBreakInfo breakInfo = *( breakInfoBuffer + index );
+
+ // Some characters (like white spaces) are valid for many scripts. The rules to set a script
+ // for them are:
+ // - If they are at the begining of a paragraph they get the script of the first character with
+ // a defined script. If they are at the end, they get the script of the last one.
+ // - If they are between two scripts with the same direction, they get the script of the previous
+ // character with a defined script. If the two scripts have different directions, they get the
+ // script of the first character of the paragraph with a defined script.
+
+ // Skip those characters valid for many scripts like white spaces or '\n'.
+ bool endOfText = index == numberOfCharacters;
+ while( !endOfText &&
+ IsValidForAllScripts( character ) )
{
- if( IsZeroWidthNonJoiner( character ) ||
- IsZeroWidthJoiner( character ) ||
- IsZeroWidthSpace( character ) ||
- IsLeftToRightMark( character ) ||
- IsRightToLeftMark( character ) ||
- IsThinSpace( character ) )
+ // Count all these characters to be added into a script.
+ ++numberOfAllScriptCharacters;
+
+ if( TextAbstraction::LINE_MUST_BREAK == breakInfo )
{
- // Keep previous script if the character is a zero width joiner or a zero width non joiner.
- script = currentScriptRun.script;
+ // The next character is a new paragraph.
+ // Know when there is a new paragraph is needed because if there is a white space
+ // between two scripts with different directions, it is added to the script with
+ // the same direction than the first script of the paragraph.
+ firstValidScript = true;
+ isParagraphRTL = false;
}
- else
+
+ // Get the next character.
+ ++index;
+ endOfText = index == numberOfCharacters;
+ if( !endOfText )
{
- script = TextAbstraction::LATIN;
- DALI_ASSERT_DEBUG( !"MultilanguageSupport::SetScripts. Unkown script!" );
+ character = *( textBuffer + index );
+ breakInfo = *( breakInfoBuffer + index );
}
}
+ if( endOfText )
+ {
+ // Last characters of the text are 'white spaces'.
+ // There is nothing else to do. Just add the remaining characters to the last script after this bucle.
+ break;
+ }
+
+ // Get the script of the character.
+ Script script = GetCharacterScript( character );
+
+ // Check if it is the first character of a paragraph.
+ if( firstValidScript &&
+ ( TextAbstraction::UNKNOWN != script ) )
+ {
+ // Sets the direction of the first valid script.
+ isParagraphRTL = ( TextAbstraction::ARABIC == script );
+ firstValidScript = false;
+ }
+
if( script != currentScriptRun.script )
{
// Current run needs to be stored and a new one initialized.
+ if( isParagraphRTL != ( TextAbstraction::ARABIC == script ) )
+ {
+ // Current script has different direction than the first script of the paragraph.
+ // All the previously skipped characters need to be added to the previous script before it's stored.
+ currentScriptRun.characterRun.numberOfCharacters += numberOfAllScriptCharacters;
+ numberOfAllScriptCharacters = 0u;
+ }
+
if( 0u != currentScriptRun.characterRun.numberOfCharacters )
{
// Store the script run.
// Initialize the new one.
currentScriptRun.characterRun.characterIndex = currentScriptRun.characterRun.characterIndex + currentScriptRun.characterRun.numberOfCharacters;
- currentScriptRun.characterRun.numberOfCharacters = 0u;
+ currentScriptRun.characterRun.numberOfCharacters = numberOfAllScriptCharacters; // Adds the white spaces which are at the begining of the script.
currentScriptRun.script = script;
+ numberOfAllScriptCharacters = 0u;
+ }
+ else
+ {
+ // Adds white spaces between characters.
+ currentScriptRun.characterRun.numberOfCharacters += numberOfAllScriptCharacters;
+ numberOfAllScriptCharacters = 0u;
+ }
+
+ if( TextAbstraction::LINE_MUST_BREAK == breakInfo )
+ {
+ // The next character is a new paragraph.
+ firstValidScript = true;
+ isParagraphRTL = false;
}
// Add one more character to the run.
++currentScriptRun.characterRun.numberOfCharacters;
}
+ // Add remaining characters into the last script.
+ currentScriptRun.characterRun.numberOfCharacters += numberOfAllScriptCharacters;
if( 0u != currentScriptRun.characterRun.numberOfCharacters )
{
+ if( TextAbstraction::UNKNOWN == currentScriptRun.script )
+ {
+ // There are only white spaces in the last script. Set the latin script.
+ currentScriptRun.script = TextAbstraction::LATIN;
+ }
+
// Store the last run.
scripts.PushBack( currentScriptRun );
}
namespace
{
+const unsigned int WHITE_SPACE_THRESHOLD = 0x21; ///< All characters below 0x21 are considered white spaces.
+const unsigned int CHAR_FL = 0x000A; ///< NL Line feed, new line.
+const unsigned int CHAR_VT = 0x000B; ///< Vertical tab.
+const unsigned int CHAR_FF = 0x000C; ///< NP Form feed, new page.
+const unsigned int CHAR_NEL = 0x0085; ///< Next line.
+const unsigned int CHAR_LS = 0x2028; ///< Line separator.
+const unsigned int CHAR_PS = 0x2029; ///< Paragraph separator
+
const unsigned int CHAR_ZWS = 0x200B; ///< Zero width space.
const unsigned int CHAR_ZWNJ = 0x200C; ///< Zero width non joiner.
const unsigned int CHAR_ZWJ = 0x200D; ///< Zero width joiner.
// Burmese script
// 0x1000 - 0x109f Myanmar
-
if( character <= 0x0cff )
{
if( character <= 0x09ff )
return TextAbstraction::UNKNOWN;
}
+bool IsWhiteSpace( Character character )
+{
+ return character < WHITE_SPACE_THRESHOLD;
+}
+
+bool IsNewParagraph( Character character )
+{
+ return ( ( CHAR_FL == character ) ||
+ ( CHAR_VT == character ) ||
+ ( CHAR_FF == character ) ||
+ ( CHAR_NEL == character ) ||
+ ( CHAR_LS == character ) ||
+ ( CHAR_PS == character ) );
+}
+
bool IsZeroWidthNonJoiner( Character character )
{
return CHAR_ZWNJ == character;
#include <dali-toolkit/public-api/text/logical-model.h>
#include <dali-toolkit/public-api/text/multi-language-support.h>
#include <dali-toolkit/public-api/text/script-run.h>
+#include <dali-toolkit/public-api/text/segmentation.h>
#include <dali-toolkit/public-api/text/shaper.h>
#include <dali-toolkit/public-api/text/text-view.h>
#include <dali-toolkit/public-api/text/visual-model.h>
{
Impl()
: mNewText(),
- mOperations( NO_OPERATION )
+ mOperations( NO_OPERATION ),
+ mControlSize()
{
mLogicalModel = LogicalModel::New();
mVisualModel = VisualModel::New();
TextAbstraction::FontClient mFontClient;
OperationsMask mOperations;
+
+ Size mControlSize;
};
ControllerPtr Controller::New()
bool viewUpdated = false;
- if( size != mControlSize )
+ if( size != mImpl->mControlSize )
{
viewUpdated = DoRelayout( size, mImpl->mOperations );
// Do not re-do any operation until something changes.
mImpl->mOperations = NO_OPERATION;
- mControlSize = size;
+ mImpl->mControlSize = size;
}
return viewUpdated;
text.clear();
}
+ Vector<LineBreakInfo> lineBreakInfo;
+ if( GET_LINE_BREAKS & operations )
+ {
+ // Retrieves the line break info. The line break info is used to split the text in 'paragraphs' to
+ // calculate the bidirectional info for each 'paragraph'.
+ // It's also used to layout the text (where it should be a new line) or to shape the text (text in different lines
+ // is not shaped together).
+ lineBreakInfo.Resize( characterCount, TextAbstraction::LINE_NO_BREAK );
+
+ SetLineBreakInfo( utf32Characters,
+ lineBreakInfo );
+
+ mImpl->mLogicalModel->SetLineBreakInfo( lineBreakInfo.Begin(), characterCount );
+ }
+
const bool getScripts = GET_SCRIPTS & operations;
const bool validateFonts = VALIDATE_FONTS & operations;
{
// Retrieves the scripts used in the text.
multilanguageSupport.SetScripts( utf32Characters,
+ lineBreakInfo,
scripts );
// Sets the scripts into the model.
}
}
- Vector<LineBreakInfo> lineBreakInfo;
- if( GET_LINE_BREAKS & operations )
- {
- // Retrieves the line break info. The line break info is used to split the text in 'paragraphs' to
- // calculate the bidirectional info for each 'paragraph'.
- // It's also used to layout the text (where it should be a new line) or to shape the text (text in different lines
- // is not shaped together).
- lineBreakInfo.Resize( characterCount, TextAbstraction::LINE_NO_BREAK );
- mImpl->mLogicalModel->SetLineBreakInfo( lineBreakInfo.Begin(), characterCount );
- }
-
Vector<GlyphInfo> glyphs;
Vector<CharacterIndex> characterIndices;
Vector<Length> charactersPerGlyph;
mImpl->mVisualModel->GetGlyphs( glyphs.Begin(),
0u,
numberOfGlyphs );
-
+
mImpl->mVisualModel->GetGlyphToCharacterMap( characterIndices.Begin(),
0u,
numberOfGlyphs );
}
Controller::Controller()
-: mImpl( NULL ),
- mControlSize()
+: mImpl( NULL )
{
mImpl = new Controller::Impl();
}