Documentation

Fields

• strings :: [Ptr CChar]

  char *Strings[KBTS_FONT_INFO_STRING_ID_COUNT]

• stringLengths :: [Word16]

  kbts_u16 StringLengths[KBTS_FONT_INFO_STRING_ID_COUNT]

• styleFlags :: FontStyleFlags
   
• weight :: FontWeight
   
• width :: FontWidth
   

Fields

• glyphStorage :: Ptr GlyphStorage
   
• currentGlyph :: Ptr Glyph
   
• lastAdvanceX :: Int
   
• x :: Int
   
• y :: Int
   

Fields

• codepoint :: Ptr Codepoint
   
• context :: ShapeContext
   
• endBlockIndex :: Word32
   
• onePastLastCodepointIndex :: Word32
   
• blockIndex :: Word32
   
• codepointIndex :: Word32
   
• currentBlockCodepointCount :: Word32
   
• flatCodepointIndex :: Word32
   

Fields

• font :: Font
   
• script :: Script
   
• paragraphDirection :: Direction
   
• direction :: Direction
   
• flags :: BreakFlags
   
• glyphs :: GlyphIterator
   

Fields

• allocator :: FunPtr Allocator
   
• allocatorData :: Ptr ()
   
• blockSentinel :: ArenaBlockHeader
   
• freeBlockSentinel :: ArenaBlockHeader
   
• error :: Int
   

Fields

• prev :: Ptr ArenaBlockHeader
   
• next :: Ptr ArenaBlockHeader
   

Fields

• arena :: Arena
   
• glyphSentinel :: Glyph
   
• freeGlyphSentinel :: Glyph
   

Fields

• prev :: Ptr Glyph
   
• next :: Ptr Glyph
   
• codepoint :: Word32
   
• id :: Word16

  Glyph index. This is what you want to use to query outline data.

• uid :: Word16
   
• userIdOrCodepointIndex :: Int

  This field is kept and returned as-is throughout the shaping process.

  When you are using the context API, it contains a codepoint index always! To get the original user ID with the context API, you need to get the corresponding kbts_shape_codepoint with kbts_ShapeGetShapeCodepoint(Context, Glyph->UserIdOrCodepointIndex, ...);

• offsetX :: Int32
   
• offsetY :: Int32
   
• advanceX :: Int32
   
• advanceY :: Int32
   
• attachGlyph :: Ptr Glyph

  Set by GPOS attachments

  Earlier on, we used to assume that, if a glyph had no advance, or had the MARK glyph class, then it could be handled as a mark in layout operations. This is inaccurate. Unicode makes a distinction between attached marks and standalone marks. For our purposes, attached marks are marks that have found a valid base character to attach to. In practice, this means that the font contains a valid display position/configuration for it in the current context. In contrast, standalone marks are marks that aren't attached to anything. Fonts may still have glyphs for them, in which case we want to display those just like regular glyphs that take up horizontal space on the line. When fonts don't have glyphs for them, they simply stay around as zero-width glyphs. Standalone marks have notably different behavior compared to attached marks, and so, once we start applying positioning features, it becomes worthwhile to track exactly which glyph has attached to which.

• config :: Ptr ()
   
• decomposition :: Word64
   
• classes :: Word32
   
• flags :: GlyphFlags
   
• parentInfo :: Word32
   
• ligatureUid :: Word16

  This is set by GSUB and used by GPOS. A 0-index means that we should attach to the last component in the ligature.

  From the Microsoft docs: To correctly access the subtables, the client must keep track of the component associated with the mark.

  For a given mark assigned to a particular class, the appropriate base attachment point is determined by which ligature component the mark is associated with. This is dependent on the original character string and subsequent character- or glyph-sequence processing, not the font data alone. While a text-layout client is performing any character-based preprocessing or any glyph-substitution operations using the GSUB table, the text-layout client must keep track of associations of marks to particular ligature-glyph components.

• ligatureComponentIndexPlusOne :: Word16
   
• ligatureComponentCount :: Word16
   
• joiningFeature :: UnicodeJoiningFeature

  Set in GSUB and used in GPOS, for STCH.

• joiningType :: UnicodeJoiningType
   
• unicodeFlags :: Word8
   
• syllabicClass :: Word8
   
• syllabicPosition :: Word8
   
• useClass :: Word8
   
• combiningClass :: Word8
   
• markOrdering :: Word8

  Only used temporarily in NORMALIZE for Arabic mark reordering.

Fields

• font :: Font

  Only set when (BreakFlags & KBTS_BREAK_FLAG_GRAPHEME) != 0.

• config :: Ptr ()
   
• codepoint :: Int
   
• userId :: Int
   
• breakFlags :: BreakFlags
   
• script :: Script

  Only set when (BreakFlags & KBTS_BREAK_FLAG_SCRIPT) != 0.

• direction :: Direction

  Only set when (BreakFlags & KBTS_BREAK_FLAG_DIRECTION) != 0.

• paragraphDirection :: Direction

  Only set when (BreakFlags & KBTS_BREAK_FLAG_PARAGRAPH_DIRECTION) != 0.

Fields

• tag :: FeatureTag
   
• value :: Int
   

Fields

• decomposition :: Word64
   
• codepoint :: Word32
   

Fields

• font :: Font
   
• baseCodepoint :: Word32
   
• currentBaseError :: Int
   
• error :: Int
   
• baseParents :: Ptr GlyphParent

  kbts_glyph_parent BaseParents[KBTS_MAXIMUM_RECOMPOSITION_PARENTS];

• baseParentCount :: Word32
   

Fields

• position :: Int

  The break code mostly works in relative positions, but we convert to absolute positions for the user. That way, breaks can be trivially stored and compared and such and it just works.

• flags :: BreakFlags
   
• direction :: Direction

  Only valid if KBTS_BREAK_FLAG_DIRECTION is set.

• paragraphDirection :: Direction

  Only valid if KBTS_BREAK_FLAG_PARAGRAPH_DIRECTION is set.

• script :: Script

  Only valid if KBTS_BREAK_FLAG_SCRIPT is set.