private static int LoadSubTable(FontParser parser, int start, FontFace font, Glyph[] glyphs)
{
// Total characters in subtable:
int characterCount = 0;
// Seek to the cmap now:
parser.Position = start;
// Check it's format 4:
int format = parser.ReadUInt16();
#if INFINITEXT_DEBUG
Fonts.OnLogMessage("Cmap subtable format: " + format);
#endif
if (format > 13)
{
// We now have e.g. 14.0 - ulong here ("Length"):
parser.Position += 4;
}
else if (format > 6)
{
// We now have e.g. 12.0 - another short here (reserved):
parser.Position += 2;
// Length and language are both 4 byte ints now:
parser.Position += 8;
}
else
{
// Size of the sub-table (map length, u16):
parser.Position += 2;
// Structure of the sub-table (map language, u16):
parser.Position += 2;
}
switch (format)
{
case 0:
// Byte encoding table:
for (int i = 0; i < 256; i++)
{
int rByte = parser.ReadByte();
Glyph glyph = glyphs[rByte];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(i);
}
}
break;
case 2:
// The offset to the headers:
int subOffset = parser.Position + (256 * 2);
// For each high byte:
for (int i = 0; i < 256; i++)
{
// Read the index to the header and zero out the bottom 3 bits:
int headerPosition = subOffset + (parser.ReadUInt16() & (~7));
// Read the header:
int firstCode = parser.ReadUInt16(ref headerPosition);
int entryCount = parser.ReadUInt16(ref headerPosition);
short idDelta = parser.ReadInt16(ref headerPosition);
// Grab the current position:
int pos = headerPosition;
// Read the idRangeOffset - the last part of the header:
pos += parser.ReadUInt16(ref headerPosition);
int maxCode = firstCode + entryCount;
// Get the high byte value:
int highByte = (i << 8);
// For each low byte:
for (int j = firstCode; j < maxCode; j++)
{
// Get the full charcode (which might not actually exist yet):
int charCode = highByte + j;
// Read the base of the glyphIndex:
int p = parser.ReadUInt16(ref pos);
if (p == 0)
{
continue;
}
p = (p + idDelta) & 0xFFFF;
if (p == 0)
{
continue;
}
Glyph glyph = glyphs[p];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(charCode);
}
}
}
break;
case 4:
// Segment count. It's doubled.
int segCount = (parser.ReadUInt16() >> 1);
// Search range, entry selector and range shift (don't need any):
parser.Position += 6;
int baseIndex = parser.Position;
int endCountIndex = baseIndex;
baseIndex += 2;
int startCountIndex = baseIndex + segCount * 2;
int idDeltaIndex = baseIndex + segCount * 4;
int idRangeOffsetIndex = baseIndex + segCount * 6;
for (int i = 0; i < segCount - 1; i++)
{
int endCount = parser.ReadUInt16(ref endCountIndex);
int startCount = parser.ReadUInt16(ref startCountIndex);
int idDelta = parser.ReadInt16(ref idDeltaIndex);
int idRangeOffset = parser.ReadUInt16(ref idRangeOffsetIndex);
for (int c = startCount; c <= endCount; c++)
{
int glyphIndex;
if (idRangeOffset != 0)
{
// The idRangeOffset is relative to the current position in the idRangeOffset array.
// Take the current offset in the idRangeOffset array.
int glyphIndexOffset = (idRangeOffsetIndex - 2);
// Add the value of the idRangeOffset, which will move us into the glyphIndex array.
glyphIndexOffset += idRangeOffset;
// Then add the character index of the current segment, multiplied by 2 for USHORTs.
glyphIndexOffset += (c - startCount) * 2;
glyphIndex = parser.ReadUInt16(ref glyphIndexOffset);
if (glyphIndex != 0)
{
glyphIndex = (glyphIndex + idDelta) & 0xFFFF;
}
}
else
{
glyphIndex = (c + idDelta) & 0xFFFF;
}
// Add a charcode to the glyph now:
Glyph glyph = glyphs[glyphIndex];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(c);
}
}
}
break;
case 6:
int firstCCode = parser.ReadUInt16();
int entryCCount = parser.ReadUInt16();
for (int i = 0; i < entryCCount; i++)
{
Glyph glyphC = glyphs[parser.ReadUInt16()];
if (glyphC != null)
{
characterCount++;
glyphC.AddCharcode(firstCCode + i);
}
}
break;
case 10:
// Trimmed array. Similar to format 6.
int startCharCode = parser.ReadUInt16();
int numChars = parser.ReadUInt16();
for (int i = 0; i < numChars; i++)
{
Glyph glyphC = glyphs[parser.ReadUInt16()];
if (glyphC != null)
{
characterCount++;
glyphC.AddCharcode(startCharCode + i);
}
}
break;
case 12:
// Segmented coverage.
// Mapping of 1 charcode to 1 glyph. "Segmented" because it can come in blocks called groups.
int groups = (int)parser.ReadUInt32();
// For each group of glyphs..
for (int i = 0; i < groups; i++)
{
// Start/end charcode:
int startCode = (int)parser.ReadUInt32();
int endCode = (int)parser.ReadUInt32();
// Start glyph ID:
int startGlyph = (int)parser.ReadUInt32();
int count = (endCode - startCode);
// For each glyph/charcode pair..
for (int j = 0; j <= count; j++)
{
// Get the glyph:
int glyphIndex = (startGlyph + j);
Glyph glyph = glyphs[glyphIndex];
if (glyph != null)
{
characterCount++;
// Charcode is..
glyph.AddCharcode(startCode + j);
}
}
}
break;
case 13:
// Many to one. Same format as #12 but the meaning is a bit different.
// How many groups?
int glyphCount = (int)parser.ReadUInt32();
for (int i = 0; i < glyphCount; i++)
{
int startCode = (int)parser.ReadUInt32();
int endCode = (int)parser.ReadUInt32();
int glyphID = (int)parser.ReadUInt32();
// Get the glyph:
Glyph glyph = glyphs[glyphID];
if (glyph != null)
{
int count = (endCode - startCode);
// For each charcode..
for (int j = 0; j <= count; j++)
{
characterCount++;
// Hook up glyph to this charcode:
glyph.AddCharcode(startCode + j);
}
}
}
break;
case 14:
Fonts.OnLogMessage("InfiniText partially supports part of the font '" + font.Family.Name + "' - this is harmless. Search for this message for more.");
// This font contains a format 14 CMAP Table.
// Format 14 is "Unicode variation selectors" - essentially different versions of the same character.
// E.g. a text Emoji character and a graphical one.
// In a text system like InfiniText, that just means we must map a bunch of different charcodes
// to the same glyph.
// .. I Think! As usual, the OpenType spec doesn't make too much sense.
// However, it appears to be entirely optional.
// So, approx implementation is below, however getting the utf32 code point from the variation + base character
// is completely undocumented - my best guess unfortunately threw errors.
// See the commented out block below!
break;
/*
*
* case 14:
*
* // How many var selector records?
* int records=(int)parser.ReadUInt32();
*
* for(int i=0;i<records;i++){
*
* // variation selector:
* int varSelector=(int)parser.ReadUInt24();
*
* // Offsets:
* int defaultUVSOffset=(int)parser.ReadUInt32();
* int nonDefaultUVSOffset=(int)parser.ReadUInt32();
*
* // Grab parser position:
* int position=parser.Position;
*
* // Got a ref to a default style table?
* if(defaultUVSOffset!=0){
*
* // Yep! The UVS is simply a list of "base" characters, each with ranges of available charcodes.
* // [BaseCharCode][The extended part. Each of these comes from the range.]
* // The actual glyph is the one that we get by directly looking up each of the base characters.
*
* // Seek to the table:
* parser.Position=start+defaultUVSOffset;
*
* // Read the unicode value ranges count:
* int numUniRangesCount=(int)parser.ReadUInt32();
*
* // For each one..
* for(int m=0;m<numUniRangesCount;m++){
*
* // Read the base charcode:
* int baseCharcode=(int)parser.ReadUInt24();
*
* // Read the size of the range:
* byte rangeSize=parser.ReadByte();
*
* for(int c=0;c<=rangeSize;c++){
*
* // Fetch the base glyph:
* Glyph glyph=font.GetGlyphDirect(baseCharcode);
*
* if(glyph!=null){
*
* // Combine baseCharcode with varSelector next to form the variation (of "glyph").
*
* // Get the full charcode (this is incorrect!):
* // int charcode=char.ConvertToUtf32((char)baseCharcode,(char)varSelector);
*
* // Add:
* //glyph.AddCharcode(charcode);
*
* }
*
* // Move baseCharcode along:
* baseCharcode++;
*
* }
*
* }
*
* // Restore parser:
* parser.Position=position;
*
* }
*
* // Got a ref to a non-default style table?
* if(nonDefaultUVSOffset!=0){
*
* // Yep! The UVS is simply a list of "base" characters, each with ranges of available charcodes.
* // [BaseCharCode][The extended part. Each of these comes from the range.]
* // This time though, the glyph to use is directly specified
* // (that's what gives it the "non-default" property).
*
* // Seek to the table:
* parser.Position=start+nonDefaultUVSOffset;
*
* // Read the number of mappings:
* int numMappings=(int)parser.ReadUInt32();
*
* // For each one..
* for(int m=0;m<numMappings;m++){
*
* // Read the base charcode:
* int baseCharcode=(int)parser.ReadUInt24();
*
* // Read glyph ID:
* int glyphID=(int)parser.ReadUInt16();
*
* // Get the glyph:
* Glyph glyph=glyphs[glyphID];
*
* if(glyph!=null){
*
* // Combine baseCharcode with varSelector next to form the variation (of "glyph").
*
* // Get the full charcode (this is incorrect!):
* // int charcode=char.ConvertToUtf32((char)baseCharcode,(char)varSelector);
*
* // Add:
* //glyph.AddCharcode(charcode);
*
* }
*
* }
*
* // Restore parser:
* parser.Position=position;
*
* }
*
* }
*
* break;
*
*/
default:
Fonts.OnLogMessage("InfiniText does not currently support part of this font. If you need it, please contact us with this: Format: " + format);
break;
}
return(characterCount);
}
public static bool Load(FontParser parser, int start, FontFace font, Glyph[] glyphs)
{
// Seek there:
parser.Position = start;
// Read the version (and check if it's zero):
if (parser.ReadUInt16() != 0)
{
return(false);
}
// Strangely the cmap table can have lots of tables inside it. For now we're looking for the common type 3 table.
// -> Got to grab all tables for a particular "platform".
// -> Prefer platform 3 (Microsoft, most common and best documented), then 0, then anything else.
// So, first let's figure out which platform we'll be using.
// Favourite platform so far:
int selectedPlatform = -1;
// Number of tables:
int tableCount = parser.ReadUInt16();
for (int i = 0; i < tableCount; i += 1)
{
// Grab the platform ID:
int platformId = parser.ReadUInt16();
if (platformId == 0)
{
// Great, got platform 0.
// Just halt there because we know this fonts got 0 so we'll use that.
selectedPlatform = 0;
break;
}
else if (platformId == 3)
{
// Prefer 3 over others:
selectedPlatform = 3;
}
else if (selectedPlatform == -1)
{
// Anything else (last resort):
selectedPlatform = platformId;
}
// Skip encoding ID and offset:
parser.Position += 6;
}
if (selectedPlatform == -1)
{
// Empty table! Return.
return(false);
}
// Round 2. This time, select all subtables of the favourite platform.
// Then collect the offset and load it up.
// Reset parser, skipping count and version:
parser.Position = start + 4;
// For each table..
for (int i = 0; i < tableCount; i += 1)
{
// Grab the platform ID:
int platformId = parser.ReadUInt16();
#if INFINITEXT_DEBUG
// And the encoding ID:
int encodingId = parser.ReadUInt16();
Fonts.OnLogMessage("Font " + font.FamilyName + " cmap subtable. Platform ID: " + platformId + ", encoding ID: " + encodingId);
#else
// And the encoding ID:
parser.ReadUInt16();
#endif
if (platformId == selectedPlatform)
{
// Read offset:
int offset = (int)parser.ReadUInt32();
// Get position:
int position = parser.Position;
// Load the subtable now:
font.CharacterCount += LoadSubTable(parser, start + offset, font, glyphs);
// Reset parser:
parser.Position = position;
}
else
{
// Skip offset:
parser.Position += 4;
}
}
return(true);
}
public static bool Load(FontParser parser, int start, FontFace font, Glyph[] glyphs)
{
// Seek there:
parser.Position = start;
// Read the version (and check if it's zero):
if (parser.ReadUInt16() != 0)
{
return(false);
}
// Strangely the cmap table can have lots of tables inside it. For now we're looking for the common type 3 table.
// Number of tables:
int tableCount = parser.ReadUInt16();
// Total characters in font:
int characterCount = 0;
int offset = -1;
int favour = 0;
for (int i = 0; i < tableCount; i += 1)
{
// Grab the platform ID:
int platformId = parser.ReadUInt16();
// And the encoding ID:
int encodingId = parser.ReadUInt16();
if (platformId == 3 || platformId == 0)
{
if (encodingId == 10)
{
// Top favourite - most broad Unicode encoding.
// Read offset:
offset = (int)parser.ReadUInt32();
break;
}
else if (encodingId == 1 || encodingId == 0)
{
// Read offset:
offset = (int)parser.ReadUInt32();
// Mid-range favourite:
favour = 1;
continue;
}
else if (favour == 0)
{
// Anything else we'll give a try:
// Read offset (but don't break):
offset = (int)parser.ReadUInt32();
continue;
}
}
// Skip offset:
parser.Position += 4;
}
if (offset == -1)
{
// We don't support this font :(
return(false);
}
// Seek to the cmap now:
parser.Position = start + offset;
// Check it's format 4:
int format = parser.ReadUInt16();
if (format > 6)
{
// We now have e.g. 12.0 - another short here:
parser.Position += 2;
// Size/ structure are both 4 byte ints now:
parser.Position += 8;
}
else
{
// Size of the sub-table (map length, u16):
parser.Position += 2;
// Structure of the sub-table (map language, u16):
parser.Position += 2;
}
switch (format)
{
case 0:
// Byte encoding table:
for (int i = 0; i < 256; i++)
{
int rByte = parser.ReadByte();
Glyph glyph = glyphs[rByte];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(i);
}
}
break;
case 2:
// The offset to the headers:
int subOffset = parser.Position + (256 * 2);
// For each high byte:
for (int i = 0; i < 256; i++)
{
// Read the index to the header and zero out the bottom 3 bits:
int headerPosition = subOffset + (parser.ReadUInt16() & (~7));
// Read the header:
int firstCode = parser.ReadUInt16(ref headerPosition);
int entryCount = parser.ReadUInt16(ref headerPosition);
short idDelta = parser.ReadInt16(ref headerPosition);
// Grab the current position:
int pos = headerPosition;
// Read the idRangeOffset - the last part of the header:
pos += parser.ReadUInt16(ref headerPosition);
int maxCode = firstCode + entryCount;
// Get the high byte value:
int highByte = (i << 8);
// For each low byte:
for (int j = firstCode; j < maxCode; j++)
{
// Get the full charcode (which might not actually exist yet):
int charCode = highByte + j;
// Read the base of the glyphIndex:
int p = parser.ReadUInt16(ref pos);
if (p == 0)
{
continue;
}
p = (p + idDelta) & 0xFFFF;
if (p == 0)
{
continue;
}
Glyph glyph = glyphs[p];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(charCode);
}
}
}
break;
case 4:
// Segment count. It's doubled.
int segCount = (parser.ReadUInt16() >> 1);
// Search range, entry selector and range shift (don't need any):
parser.Position += 6;
int baseIndex = parser.Position;
int endCountIndex = baseIndex;
baseIndex += 2;
int startCountIndex = baseIndex + segCount * 2;
int idDeltaIndex = baseIndex + segCount * 4;
int idRangeOffsetIndex = baseIndex + segCount * 6;
for (int i = 0; i < segCount - 1; i++)
{
int endCount = parser.ReadUInt16(ref endCountIndex);
int startCount = parser.ReadUInt16(ref startCountIndex);
int idDelta = parser.ReadInt16(ref idDeltaIndex);
int idRangeOffset = parser.ReadUInt16(ref idRangeOffsetIndex);
for (int c = startCount; c <= endCount; c++)
{
int glyphIndex;
if (idRangeOffset != 0)
{
// The idRangeOffset is relative to the current position in the idRangeOffset array.
// Take the current offset in the idRangeOffset array.
int glyphIndexOffset = (idRangeOffsetIndex - 2);
// Add the value of the idRangeOffset, which will move us into the glyphIndex array.
glyphIndexOffset += idRangeOffset;
// Then add the character index of the current segment, multiplied by 2 for USHORTs.
glyphIndexOffset += (c - startCount) * 2;
glyphIndex = parser.ReadUInt16(ref glyphIndexOffset);
if (glyphIndex != 0)
{
glyphIndex = (glyphIndex + idDelta) & 0xFFFF;
}
}
else
{
glyphIndex = (c + idDelta) & 0xFFFF;
}
// Add a charcode to the glyph now:
Glyph glyph = glyphs[glyphIndex];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(c);
}
}
}
break;
case 6:
int firstCCode = parser.ReadUInt16();
int entryCCount = parser.ReadUInt16();
for (int i = 0; i < entryCCount; i++)
{
Glyph glyphC = glyphs[parser.ReadUInt16()];
if (glyphC != null)
{
characterCount++;
glyphC.AddCharcode(firstCCode + i);
}
}
break;
case 12:
int groups = (int)parser.ReadUInt32();
for (int i = 0; i < groups; i++)
{
int startCode = (int)parser.ReadUInt32();
int endCode = (int)parser.ReadUInt32();
int startGlyph = (int)parser.ReadUInt32();
int count = (endCode - startCode);
for (int j = 0; j <= count; j++)
{
int glyphIndex = (startGlyph + j);
Glyph glyph = glyphs[glyphIndex];
if (glyph != null)
{
characterCount++;
glyph.AddCharcode(startCode + j);
}
}
}
break;
default:
Fonts.OnLogMessage("InfiniText does not currently support this font. If you need it, please contact us with this: Format: " + format);
break;
}
font.CharacterCount = characterCount;
return(true);
}
