OU
/
CWE-ou-minkata
mirror of https://foundry.openuru.org/gitblit/r/CWE-ou-minkata.git
2
1
Fork 0
Code Releases Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
200 Commits
79 Branches
23 Tags
171 MiB
 
 
 
 
 
Tree: 21e2a6de7b
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '21e2a6de7b'
${ noResults }
CWE-ou-minkata/MOULOpenSourceClientPlugin/Plasma20/Sources/Plasma/PubUtilLib/plGImage/plFont.cpp

2176 lines
64 KiB
Raw Blame History

/*==LICENSE==*
CyanWorlds.com Engine - MMOG client, server and tools
Copyright (C) 2011 Cyan Worlds, Inc.
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
Additional permissions under GNU GPL version 3 section 7
If you modify this Program, or any covered work, by linking or
combining it with any of RAD Game Tools Bink SDK, Autodesk 3ds Max SDK,
NVIDIA PhysX SDK, Microsoft DirectX SDK, OpenSSL library, Independent
JPEG Group JPEG library, Microsoft Windows Media SDK, or Apple QuickTime SDK
(or a modified version of those libraries),
containing parts covered by the terms of the Bink SDK EULA, 3ds Max EULA,
PhysX SDK EULA, DirectX SDK EULA, OpenSSL and SSLeay licenses, IJG
JPEG Library README, Windows Media SDK EULA, or QuickTime SDK EULA, the
licensors of this Program grant you additional
permission to convey the resulting work. Corresponding Source for a
non-source form of such a combination shall include the source code for
the parts of OpenSSL and IJG JPEG Library used as well as that of the covered
work.
You can contact Cyan Worlds, Inc. by email legal@cyan.com
or by snail mail at:
Cyan Worlds, Inc.
14617 N Newport Hwy
Mead, WA 99021
*==LICENSE==*/
///////////////////////////////////////////////////////////////////////////////
// //
// plFont Class Header //
// Seems like we've come full circle again. This is our generic Plasma //
// bitmap font class/format. Quick list of things it supports, or needs to: //
// - Antialiasing, either in the font def or at rendertime //
// - Doublebyte character sets //
// - Platform independence, of course //
// - Render to reasonably arbitrary mipmap //
// - Character-level kerning, both before and after, as well as //
// negative kerning (for ligatures) //
// //
// Cyan, Inc. //
// //
//// Version History //////////////////////////////////////////////////////////
// //
// 3.4.2003 mcn - Created. //
// 4.3.2003 mcn - Updated. Casting a char to a UInt16 sign-extends it if //
// the char is > 128, but casting it to an UInt8 first works.//
// Ugly as sin, but hey, so are you. //
// //
///////////////////////////////////////////////////////////////////////////////
#include "hsTypes.h"
#include "hsStlUtils.h"
#include "plFont.h"
#include "plMipmap.h"
#include "hsResMgr.h"
//// plCharacter Stuff ////////////////////////////////////////////////////////
plFont::plCharacter::plCharacter()
{
fBitmapOff = 0;
fHeight = 0;
fBaseline = 0;
fLeftKern = fRightKern = 0.f;
}
void plFont::plCharacter::Read( hsStream *s )
{
// Rocket science here...
s->ReadSwap( &fBitmapOff );
s->ReadSwap( &fHeight );
s->ReadSwap( &fBaseline );
s->ReadSwap( &fLeftKern );
s->ReadSwap( &fRightKern );
}
void plFont::plCharacter::Write( hsStream *s )
{
s->WriteSwap( fBitmapOff );
s->WriteSwap( fHeight );
s->WriteSwap( fBaseline );
s->WriteSwap( fLeftKern );
s->WriteSwap( fRightKern );
}
//// Constructor/Read/Write/Destructor/etc ////////////////////////////////////
plFont::plFont()
{
IClear( true );
}
plFont::~plFont()
{
IClear();
}
void plFont::IClear( bool onConstruct )
{
if( !onConstruct )
delete [] fBMapData;
memset( fFace, 0, sizeof( fFace ) );
fSize = 0;
fFlags = 0;
fWidth = fHeight = 0;
fBPP = 0;
fBMapData = nil;
fFirstChar = 0;
fMaxCharHeight = 0;
fCharacters.Reset();
fRenderInfo.fFlags = 0;
fRenderInfo.fX = fRenderInfo.fY = fRenderInfo.fNumCols = 0;
fRenderInfo.fMaxWidth = fRenderInfo.fMaxHeight = 0;
fRenderInfo.fDestPtr = nil;
fRenderInfo.fDestStride = 0;
fRenderInfo.fColor = 0;
fRenderInfo.fMipmap = nil;
fRenderInfo.fRenderFunc = nil;
fRenderInfo.fVolatileStringPtr = nil;
fRenderInfo.fFirstLineIndent = 0;
fRenderInfo.fLineSpacing = 0;
}
void plFont::SetFace( const char *face )
{
strncpy( fFace, face, sizeof( fFace ) );
}
void plFont::SetSize( UInt8 size )
{
fSize = size;
}
void plFont::Read( hsStream *s, hsResMgr *mgr )
{
hsKeyedObject::Read( s, mgr );
ReadRaw( s );
}
void plFont::Write( hsStream *s, hsResMgr *mgr )
{
hsKeyedObject::Write( s, mgr );
WriteRaw( s );
}
///////////////////////////////////////////////////////////////////////////////
//// Rendering ////////////////////////////////////////////////////////////////
///////////////////////////////////////////////////////////////////////////////
void plFont::SetRenderColor( UInt32 color )
{
fRenderInfo.fColor = color;
}
void plFont::SetRenderClipRect( Int16 x, Int16 y, Int16 width, Int16 height )
{
fRenderInfo.fClipRect.Set( x, y, width, height );
}
void plFont::SetRenderClipping( Int16 x, Int16 y, Int16 width, Int16 height )
{
SetRenderFlag( kRenderWrap, false );
SetRenderFlag( kRenderClip, true );
SetRenderClipRect( x, y, width, height );
}
void plFont::SetRenderWrapping( Int16 x, Int16 y, Int16 width, Int16 height )
{
SetRenderFlag( kRenderWrap, true );
SetRenderFlag( kRenderClip, false );
SetRenderClipRect( x, y, width, height );
}
void plFont::ICalcFontAscent( void )
{
UInt32 i;
// Hack for now, only calc the ascent for characters in the 127 character ASCII range
for( i = 0, fFontAscent = 0, fFontDescent = 0, fMaxCharHeight = 0; i < fCharacters.GetCount(); i++ )
{
if( i + fFirstChar < 128 && fFontAscent < fCharacters[ i ].fBaseline )
fFontAscent = fCharacters[ i ].fBaseline;
Int32 descent = fCharacters[ i ].fHeight - fCharacters[ i ].fBaseline;
if( fFontDescent < descent )
fFontDescent = descent;
if( fMaxCharHeight < fCharacters[ i ].fHeight )
fMaxCharHeight = fCharacters[ i ].fHeight;
}
}
//// IIsWordBreaker //////////////////////////////////////////////////////////
// Returns whether the given character is one that can break a line
static inline bool IIsWordBreaker( const char c )
{
return ( strchr( " \t,.;\n", c ) != nil ) ? true : false;
}
//// IIsDrawableWordBreak /////////////////////////////////////////////////////
// Returns whether the given character is a line breaker that has to be drawn
// (non whitespace)
static inline bool IIsDrawableWordBreak( const char c )
{
return ( strchr( ",.;", c ) != nil ) ? true : false;
}
//// RenderString /////////////////////////////////////////////////////////////
// The base render function. Additional options are specified externally,
// so that their effects can be cached for optimization
void plFont::RenderString( plMipmap *mip, UInt16 x, UInt16 y, const char *string, UInt16 *lastX, UInt16 *lastY )
{
// convert the char string to a wchar_t string
wchar_t *wideString = hsStringToWString(string);
RenderString(mip,x,y,wideString,lastX,lastY);
delete [] wideString;
}
void plFont::RenderString( plMipmap *mip, UInt16 x, UInt16 y, const wchar_t *string, UInt16 *lastX, UInt16 *lastY )
{
if( mip->IsCompressed() )
{
hsAssert( false, "Unable to render string to compressed mipmap" );
return;
}
IRenderString( mip, x, y, string, false );
if( lastX != nil )
*lastX = fRenderInfo.fLastX;
if( lastY != nil )
*lastY = fRenderInfo.fLastY;
}
void plFont::IRenderString( plMipmap *mip, UInt16 x, UInt16 y, const wchar_t *string, hsBool justCalc )
{
fRenderInfo.fMipmap = mip;
fRenderInfo.fX = x;
fRenderInfo.fY = y;
fRenderInfo.fNumCols = (Int16)(( fBPP <= 8 ) ? ( ( fWidth * fBPP ) >> 3 ) : 0);
fRenderInfo.fFloatWidth = (hsScalar)fWidth;
fRenderInfo.fFarthestX = x;
fRenderInfo.fMaxAscent = 0;
fRenderInfo.fVolatileStringPtr = string;
switch( fRenderInfo.fFlags & kRenderJustYMask )
{
case kRenderJustYTop:
fRenderInfo.fY += (Int16)fFontAscent;
break;
case kRenderJustYBottom:
if( fRenderInfo.fFlags & ( kRenderClip | kRenderWrap ) )
fRenderInfo.fY = (Int16)(fRenderInfo.fClipRect.GetBottom() - 1 - fMaxCharHeight + fFontAscent);
else
fRenderInfo.fY = (Int16)(mip->GetHeight() - 1 - fMaxCharHeight + fFontAscent);
break;
case kRenderJustYCenter:
fRenderInfo.fY = (Int16)(( fRenderInfo.fFlags & ( kRenderClip | kRenderWrap ) ) ? fRenderInfo.fClipRect.GetBottom() - 1 : mip->GetHeight() - 1);
fRenderInfo.fY = (Int16)(( fRenderInfo.fY - fMaxCharHeight ) >> 1);
fRenderInfo.fY += (Int16)fFontAscent;
break;
default: // Just the baseline
;
}
if( justCalc )
{
plCharacter &ch = fCharacters[ (UInt16)string[ 0 ] - fFirstChar ];
fRenderInfo.fX = fRenderInfo.fFarthestX = x - (Int16)ch.fLeftKern;
if( fRenderInfo.fX < 0 )
fRenderInfo.fX = 0;
}
else
{
switch( fRenderInfo.fFlags & kRenderJustXMask )
{
case kRenderJustXLeft:
// Default
break;
case kRenderJustXForceLeft:
{
// plCharacter &ch = fCharacters[ (UInt16)(UInt8)string[ 0 ] - fFirstChar ];
// Int32 newX = x - (Int16)ch.fLeftKern;
// if( newX < 0 )
// newX = 0;
// fRenderInfo.fX = fRenderInfo.fFarthestX = newX;
}
break;
case kRenderJustXCenter:
{
UInt16 right = (UInt16)(( fRenderInfo.fFlags & ( kRenderClip | kRenderWrap ) ) ? fRenderInfo.fClipRect.GetRight() : mip->GetWidth());
// UInt16 width = CalcStringWidth( string );
fRenderInfo.fX = fRenderInfo.fFarthestX = ( ( x + right ) >> 1 );// - ( width >> 1 );
}
break;
case kRenderJustXRight:
{
UInt16 width = 0, right = (UInt16)(( fRenderInfo.fFlags & ( kRenderClip | kRenderWrap ) ) ? fRenderInfo.fClipRect.GetRight() : mip->GetWidth());
/* if( fRenderInfo.fFlags & kRenderClip )
{
fRenderInfo.fFlags &= ~kRenderClip;
width = CalcStringWidth( string );
fRenderInfo.fFlags |= kRenderClip;
}
else
*/ width = 0;//CalcStringWidth( string );
fRenderInfo.fX = fRenderInfo.fFarthestX = right - width;
}
break;
}
}
// Choose an optimal rendering function
fRenderInfo.fRenderFunc = nil;
if( justCalc )
fRenderInfo.fRenderFunc = &plFont::IRenderCharNull;
else if( mip->GetPixelSize() == 32 )
{
if( fBPP == 1 )
fRenderInfo.fRenderFunc = ( fRenderInfo.fFlags & kRenderScaleAA ) ? &plFont::IRenderChar1To32AA : &plFont::IRenderChar1To32;
else if( fBPP == 8 )
{
if( fRenderInfo.fFlags & kRenderIntoAlpha )
{
if( fRenderInfo.fFlags & kRenderAlphaPremultiplied )
{
if (fRenderInfo.fFlags & kRenderShadow)
fRenderInfo.fRenderFunc = &plFont::IRenderChar8To32AlphaPremShadow;
else
fRenderInfo.fRenderFunc = &plFont::IRenderChar8To32AlphaPremultiplied;
}
else if( ( fRenderInfo.fColor & 0xff000000 ) != 0xff000000 )
fRenderInfo.fRenderFunc = &plFont::IRenderChar8To32Alpha;
else
fRenderInfo.fRenderFunc = &plFont::IRenderChar8To32FullAlpha;
}
else
fRenderInfo.fRenderFunc = &plFont::IRenderChar8To32;
}
}
if( fRenderInfo.fRenderFunc == nil )
{
hsAssert( false, "Invalid combination of source and destination formats in RenderString()" );
return;
}
// Init our other render values
if( !justCalc )
{
fRenderInfo.fDestStride = mip->GetRowBytes();
fRenderInfo.fDestBPP = mip->GetPixelSize() >> 3;
fRenderInfo.fDestPtr = (UInt8 *)mip->GetImage();
fRenderInfo.fDestPtr += fRenderInfo.fY * fRenderInfo.fDestStride;
fRenderInfo.fDestPtr += fRenderInfo.fX * fRenderInfo.fDestBPP;
}
if( fRenderInfo.fFlags & ( kRenderWrap | kRenderClip ) )
{
fRenderInfo.fMaxHeight = (Int16)( fRenderInfo.fClipRect.GetBottom() - fRenderInfo.fY );
fRenderInfo.fMaxWidth = (Int16)( fRenderInfo.fClipRect.GetRight() - x );
}
else if( justCalc )
{
// Just calculating, no wrapping, so the max is as far as we can go
// Note: 32767 isn't quite right, since we'll be adding the left kern in before we
// calc the first character, so adjust so we make sure we don't underflow
plCharacter &ch = fCharacters[ (UInt16)(UInt8)string[ 0 ] - fFirstChar ];
fRenderInfo.fMaxHeight = (Int16)fMaxCharHeight;
fRenderInfo.fMaxWidth = (Int16)32767 + (Int16)ch.fLeftKern;
}
else
{
fRenderInfo.fMaxHeight = (Int16)( mip->GetHeight() - fRenderInfo.fY );
fRenderInfo.fMaxWidth = (Int16)( mip->GetWidth() - x );
}
fRenderInfo.fMaxDescent = 0;
if( fRenderInfo.fFlags & kRenderWrap )
{
// Hell, gotta word wrap the text
// To avoid backtracking, we step forward in the string one word at a time until we hit a break,
// then render what we have and continue
Int32 lastWord = 0, i;
bool isFirstLine = true;
x = 0;
Int16 firstMaxAscent = 0;
UInt32 lineHt, lineDelta;
if( fRenderInfo.fFlags & kRenderScaleAA )
lineHt = fMaxCharHeight >> 1;
else
lineHt = fMaxCharHeight;
// adjust the line height for spacing
lineHt += fRenderInfo.fLineSpacing;
lineDelta = lineHt * fRenderInfo.fDestStride;
while( *string != 0 && fRenderInfo.fMaxHeight >= fFontDescent )
{
UInt8 *destStartPtr = fRenderInfo.fDestPtr;
UInt32 destStartX = fRenderInfo.fX;
Int16 destMaxWidth = fRenderInfo.fMaxWidth, thisIndent = 0;
if( isFirstLine )
{
// First line, apply indent if applicable
fRenderInfo.fX += fRenderInfo.fFirstLineIndent;
fRenderInfo.fMaxWidth -= fRenderInfo.fFirstLineIndent;
fRenderInfo.fDestPtr += fRenderInfo.fFirstLineIndent * fRenderInfo.fDestBPP;
thisIndent = fRenderInfo.fFirstLineIndent;
isFirstLine = false;
}
std::string ellipsisTracker = ""; // keeps track of ellipsis, since there are three word break chars that can't be split
bool possibleEllipsis = false;
int preEllipsisLastWord = 0; // where the word break was before we started tracking an ellipsis
// Iterate through the string, looking for the next line break
for( lastWord = 0, i = 0; string[ i ] != 0; i++ )
{
// If we're a carriage return, we go ahead and break anyway
if( string[ i ] == L'\n' )
{
lastWord = i;
break;
}
// handle invalid chars discretely
plCharacter* charToDraw = NULL;
if (fCharacters.Count() <= ((UInt16)string[i] - fFirstChar))
charToDraw = &(fCharacters[(UInt16)L' ' - fFirstChar]);
else
charToDraw = &(fCharacters[(UInt16)string[i] - fFirstChar]);
Int16 leftKern = (Int16)charToDraw->fLeftKern;
if( fRenderInfo.fFlags & kRenderScaleAA )
x += leftKern / 2;
else
x += leftKern;
// Update our position and see if we're over
// Note that our wrapping is slightly off, in that it doesn't take into account
// the left kerning of characters. Hopefully that won't matter much...
UInt16 charWidth = (UInt16)(fWidth + (Int16)charToDraw->fRightKern);
if( fRenderInfo.fFlags & kRenderScaleAA )
charWidth >>= 1;
UInt16 nonAdjustedX = (UInt16)(x + fWidth); // just in case the actual bitmap is too big to fit on page and we think the character can (because of right kern)
x += charWidth;
if(( x >= fRenderInfo.fMaxWidth ) || (nonAdjustedX >= fRenderInfo.fMaxWidth))
{
// we're over, but lastWord may not be correct (especially if we're in the middle of an ellipsis)
if (possibleEllipsis)
{
// ellipsisTracker will not be empty since possibleEllipsis is true (so there will be at least one period)
if (ellipsisTracker == ".") // only one period so far
{
if ((string[i] == '.') && (string[i+1] == '.')) // we have an ellipsis, so reset the lastWord back before we found it
lastWord = preEllipsisLastWord;
// otherwise, we don't have an ellipsis, so lastWord is correct (but the grammer might not be ;-)
}
else if (ellipsisTracker == "..") // only two periods so far
{
if (string[i] == '.') // we have an ellipsis, so reset the lastWord back before we found it
lastWord = preEllipsisLastWord;
// otherwise, we don't have an ellipsis, so lastWord is correct (but the grammer might not be ;-)
}
// if neither of the above are true, then the full ellipsis was encountered and the lastWord is correct
ellipsisTracker = "";
possibleEllipsis = false;
}
// Over, so break
break;
}
// Are we a word breaker?
if( IIsWordBreaker( (char)(string[ i ]) ) )
{
if (string[i] == '.') // we might have an ellipsis here
{
if (ellipsisTracker == "...") // we already have a full ellipsis, so break between them
{
preEllipsisLastWord = i;
ellipsisTracker = "";
}
else if (ellipsisTracker == "") // no ellipsis yet, so save the last word
preEllipsisLastWord = lastWord;
ellipsisTracker += '.';
possibleEllipsis = true;
}
else
{
ellipsisTracker = ""; // no chance of an ellipsis, so kill it
possibleEllipsis = false;
}
// Yes, and we didn't go over, so store as the last successfully fit word and move on
lastWord = i;
}
}
if( string[ i ] == 0 )
lastWord = i; // Final catch for end-of-string
else if( lastWord == 0 && string[ i ] != L'\n' && thisIndent == 0 )
lastWord = i; // Catch for a single word that didn't fit (just go up as many chars as we can)
// (but NOT if we have a first line indent, mind you :)
// Got to the end of a line (somewhere), so render up to lastWord, then advance from that point
// to the first non-word-breaker and continue onward
if( lastWord > 0 )
{
bool drawableWordBreak = IIsDrawableWordBreak((char)(string[lastWord]));
if (drawableWordBreak)
lastWord++; // make sure we draw it
if( ( fRenderInfo.fFlags & kRenderJustXMask ) == kRenderJustXRight )
{
UInt16 baseX = fRenderInfo.fX, baseMaxW = fRenderInfo.fMaxWidth;
UInt8 *baseDestPtr = fRenderInfo.fDestPtr;
fRenderInfo.fX = 0;
CharRenderFunc oldFunc = fRenderInfo.fRenderFunc;
fRenderInfo.fRenderFunc = &plFont::IRenderCharNull;
IRenderLoop( string, lastWord );
fRenderInfo.fRenderFunc = oldFunc;
fRenderInfo.fX = baseX - fRenderInfo.fX;
fRenderInfo.fMaxWidth = baseMaxW;
fRenderInfo.fDestPtr = baseDestPtr + ( fRenderInfo.fX - baseX ) * fRenderInfo.fDestBPP;
IRenderLoop( string, lastWord );
fRenderInfo.fLastX = fRenderInfo.fX;
fRenderInfo.fLastY = fRenderInfo.fY;
fRenderInfo.fX = baseX;
}
else if( ( fRenderInfo.fFlags & kRenderJustXMask ) == kRenderJustXCenter )
{
UInt16 baseX = fRenderInfo.fX, baseMaxW = fRenderInfo.fMaxWidth;
UInt8 *baseDestPtr = fRenderInfo.fDestPtr;
fRenderInfo.fX = 0;
CharRenderFunc oldFunc = fRenderInfo.fRenderFunc;
fRenderInfo.fRenderFunc = &plFont::IRenderCharNull;
IRenderLoop( string, lastWord );
fRenderInfo.fRenderFunc = oldFunc;
fRenderInfo.fX = baseX - ( fRenderInfo.fX >> 1 );
fRenderInfo.fMaxWidth = baseMaxW;
fRenderInfo.fDestPtr = baseDestPtr + ( fRenderInfo.fX - baseX ) * fRenderInfo.fDestBPP;
IRenderLoop( string, lastWord );
fRenderInfo.fLastX = fRenderInfo.fX;
fRenderInfo.fLastY = fRenderInfo.fY;
fRenderInfo.fX = baseX;
}
else if( ( fRenderInfo.fFlags & kRenderJustXMask ) == kRenderJustXForceLeft )
{
Int16 baseX = fRenderInfo.fX;
plCharacter &ch = fCharacters[ (UInt16)string[ 0 ] - fFirstChar ];
fRenderInfo.fX -= (Int16)ch.fLeftKern;
fRenderInfo.fDestPtr -= (Int16)ch.fLeftKern * fRenderInfo.fDestBPP;
IRenderLoop( string, lastWord );
fRenderInfo.fLastX = fRenderInfo.fX;
fRenderInfo.fLastY = fRenderInfo.fY;
fRenderInfo.fX = baseX;
}
else
{
IRenderLoop( string, lastWord );
fRenderInfo.fLastX = fRenderInfo.fX;
fRenderInfo.fLastY = fRenderInfo.fY;
}
if (drawableWordBreak)
lastWord--; // reset it
}
// Store farthest X for later reference
if( lastWord > 0 && fRenderInfo.fX > fRenderInfo.fFarthestX )
fRenderInfo.fFarthestX = fRenderInfo.fX;
if( firstMaxAscent == 0 )
firstMaxAscent = fRenderInfo.fMaxAscent;
x = 0;
fRenderInfo.fX = (Int16)destStartX;
fRenderInfo.fDestPtr = destStartPtr;
fRenderInfo.fMaxWidth = destMaxWidth;
fRenderInfo.fMaxAscent = 0;
// Look for the next non-word-breaker. Note that if we have any carriage returns hidden in here,
// we'll want to be advancing down even further
// Advance down
if( string[ i ] != 0 )
{
fRenderInfo.fY += (Int16)lineHt;
fRenderInfo.fMaxHeight -= (Int16)lineHt;
fRenderInfo.fDestPtr += lineDelta;
fRenderInfo.fLastX = fRenderInfo.fX;
fRenderInfo.fLastY = fRenderInfo.fY;
}
for( i = lastWord; string[ i ] != 0 && IIsWordBreaker( (char)(string[ i ]) ) && string[ i ] != L'\n'; i++ )
{
}
// Process any trailing carriage returns as a separate loop b/c we don't want to throw away white space
// after returns
for( ; string[ i ] == L'\n'; i++ )
{
// Don't process if i==lastWord, since we already did that one
if( i > lastWord && string[ i ] == L'\n' )
{
// Advance down
fRenderInfo.fY += (Int16)lineHt;
fRenderInfo.fMaxHeight -= (Int16)lineHt;
fRenderInfo.fDestPtr += lineDelta;
fRenderInfo.fLastY = fRenderInfo.fY;
}
}
// Keep going from here!
string += i;
}
fRenderInfo.fMaxAscent = firstMaxAscent;
// Final y offset from the last line
fRenderInfo.fY += (Int16)fFontDescent;
fRenderInfo.fVolatileStringPtr = string;
}
else
{
if( fRenderInfo.fFlags & kRenderClip )
{
// Advance left past any clipping area
CharRenderFunc oldFunc = fRenderInfo.fRenderFunc;
fRenderInfo.fRenderFunc = &plFont::IRenderCharNull;
Int16 prevX;
do
{
prevX = fRenderInfo.fX;
IRenderLoop( string, 1 );
}
while( fRenderInfo.fX <= fRenderInfo.fClipRect.fX && *++string != 0 );
fRenderInfo.fMaxWidth += fRenderInfo.fX - prevX;
fRenderInfo.fDestPtr -= (fRenderInfo.fX - prevX) * fRenderInfo.fDestBPP;
fRenderInfo.fX = prevX;
fRenderInfo.fRenderFunc = oldFunc;
}
// There used to be an adjustment of the X coordinate by -ch.fLeftKern for the case
// of kRenderJustXForceLeft here, but it was buggy in that it neglected to adjust
// fRenderInfo.fDestPtr and therefore had no visible effect (or almost none - only
// at the end of the line). Fixing the bug however (making kRenderJustXForceLeft
// work as intended) causes the text shadow to be cut off in some places. To ensure
// enough space for the shadow, and considering that all content was developed and
// visually optimized with the bug in place, it seems better to preserve the buggy
// behavior and make kRenderJustXForceLeft work exactly like kRenderJustXLeft.
fRenderInfo.fVolatileStringPtr = string; // Just so we can keep track of when we clip
IRenderLoop( string, -1 );
fRenderInfo.fFarthestX = fRenderInfo.fX;
}
}
void plFont::IRenderLoop( const wchar_t *string, Int32 maxCount )
{
// Render the string straight across, one char at a time
while( *string != 0 && maxCount != 0 ) // Note: if maxCount starts out -1, then it'll just keep
{ // decrementing...well ok, not for forever, but pretty close
UInt16 c = (UInt16)*string;
if( c < fFirstChar )
; // Invalid char
else
{
c -= fFirstChar;
if( c >= fCharacters.GetCount() )
; // Invalid char
else
{
// First pass at supporting left kerning values, but only at the pixel level
Int16 leftKern = (Int16)fCharacters[ c ].fLeftKern;
if( leftKern != 0 )
{
if( fRenderInfo.fFlags & kRenderScaleAA )
leftKern /= 2;
fRenderInfo.fX += leftKern;
fRenderInfo.fMaxWidth -= leftKern;
fRenderInfo.fDestPtr += leftKern * fRenderInfo.fDestBPP;
}
UInt16 thisWidth = (UInt16)(fWidth + fCharacters[ c ].fRightKern);
if( fRenderInfo.fFlags & kRenderScaleAA )
thisWidth >>= 1;
(this->*(fRenderInfo.fRenderFunc))( fCharacters[ c ] );
fRenderInfo.fX += thisWidth;
fRenderInfo.fMaxWidth -= thisWidth;
fRenderInfo.fDestPtr += thisWidth * fRenderInfo.fDestBPP;
Int16 baseline = (Int16)(fCharacters[ c ].fBaseline);
if( fRenderInfo.fFlags & kRenderScaleAA )
baseline >>= 1;
if( baseline > fRenderInfo.fMaxAscent )
fRenderInfo.fMaxAscent = baseline;
Int16 thisHt = (Int16)(fCharacters[ c ].fHeight - baseline);
if( fRenderInfo.fMaxDescent < thisHt )
fRenderInfo.fMaxDescent = thisHt;
}
}
string++;
fRenderInfo.fVolatileStringPtr++;
maxCount--;
}
}
//// The Rendering Functions //////////////////////////////////////////////////
void plFont::IRenderChar1To32( const plFont::plCharacter &c )
{
UInt8 bitMask, *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
UInt16 x, y;
if( (Int32)c.fHeight - (Int32)c.fBaseline >= fRenderInfo.fMaxHeight || c.fBaseline > fRenderInfo.fY )
return;
for( y = 0; y < c.fHeight; y++ )
{
destPtr = destBasePtr;
if( fRenderInfo.fFlags & kRenderItalic )
{
// Faux italic
destPtr += ( c.fHeight - y ) >> 1;
}
for( x = 0; x < fRenderInfo.fNumCols; x++ )
{
for( bitMask = 0x80; bitMask != 0; bitMask >>= 1 )
{
if( *src & bitMask )
*destPtr = fRenderInfo.fColor;
destPtr++;
}
src++;
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
}
}
void plFont::IRenderChar1To32AA( const plFont::plCharacter &c )
{
UInt8 bitMask, *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - ( c.fBaseline >> 1 ) * fRenderInfo.fDestStride );
UInt16 x, y;
if( ( ( (Int32)c.fHeight - (Int32)c.fBaseline ) >> 1 ) >= fRenderInfo.fMaxHeight || ( c.fBaseline >> 1 ) > fRenderInfo.fY )
return;
for( y = 0; y < c.fHeight; y += 2 )
{
destPtr = destBasePtr;
if( fRenderInfo.fFlags & kRenderItalic )
{
// Faux italic
destPtr += ( c.fHeight - y ) >> 2;
}
for( x = 0; x < fRenderInfo.fNumCols; x++ )
{
for( bitMask = 0x80; bitMask != 0; bitMask >>= 2 )
{
// Grab 4 bits and do 4-to-1 AA
UInt8 value = ( *src & bitMask ) ? 1 : 0;
value += ( *src & ( bitMask >> 1 ) ) ? 1 : 0;
value += ( src[ fRenderInfo.fNumCols ] & bitMask ) ? 1 : 0;
value += ( src[ fRenderInfo.fNumCols ] & ( bitMask >> 1 ) ) ? 1 : 0;
switch( value )
{
case 1:
{
UInt32 src = ( fRenderInfo.fColor >> 2 ) & 0x3f3f3f3f;
UInt32 dst = ( (*destPtr) >> 2 ) & 0x3f3f3f3f;
*destPtr = src + dst + dst + dst;
break;
}
case 2:
{
UInt32 src = ( fRenderInfo.fColor >> 1 ) & 0x7f7f7f7f;
UInt32 dst = ( (*destPtr) >> 1 ) & 0x7f7f7f7f;
*destPtr = src + dst;
break;
}
case 3:
{
UInt32 src = ( fRenderInfo.fColor >> 2 ) & 0x3f3f3f3f;
UInt32 dst = ( (*destPtr) >> 2 ) & 0x3f3f3f3f;
*destPtr = src + src + src + dst;
break;
}
case 4:
*destPtr = fRenderInfo.fColor;
}
destPtr++;
}
src++;
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
src += fRenderInfo.fNumCols;
}
}
void plFont::IRenderChar8To32( const plFont::plCharacter &c )
{
UInt8 *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
Int16 x, y, thisHeight, xstart, thisWidth;
UInt32 srcAlpha, oneMinusAlpha, r, g, b, dR, dG, dB, destAlpha;
UInt8 srcR, srcG, srcB;
// Unfortunately for some fonts, their right kern value actually is
// farther left than the right edge of the bitmap (think of overlapping
// script fonts). Ideally, we should store the actual width of each char's
// bitmap and use that here. However, it really shouldn't make too big of a
// difference, especially since the dest pixels that we end up overlapping
// should already be in the cache. If it does, time to upgrade the font
// format (again)
thisWidth = (Int16)fWidth;// + (Int32)c.fRightKern;
if( thisWidth >= fRenderInfo.fMaxWidth )
thisWidth = fRenderInfo.fMaxWidth;
xstart = fRenderInfo.fClipRect.fX - fRenderInfo.fX;
if( xstart < 0 )
xstart = 0;
srcR = (UInt8)(( fRenderInfo.fColor >> 16 ) & 0x000000ff);
srcG = (UInt8)(( fRenderInfo.fColor >> 8 ) & 0x000000ff);
srcB = (UInt8)(( fRenderInfo.fColor ) & 0x000000ff);
y = fRenderInfo.fClipRect.fY - fRenderInfo.fY + (Int16)c.fBaseline;
if( y < 0 )
y = 0;
else
{
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + y*fRenderInfo.fDestStride );
src += y*fRenderInfo.fNumCols;
}
thisHeight = fRenderInfo.fMaxHeight + (Int16)c.fBaseline;
if( thisHeight > (Int16)c.fHeight )
thisHeight = (Int16)c.fHeight;
for( ; y < thisHeight; y++ )
{
destPtr = destBasePtr;
for( x = xstart; x < thisWidth; x++ )
{
if( src[ x ] == 255 )
destPtr[ x ] = fRenderInfo.fColor;
else if( src[ x ] == 0 )
; // Empty
else
{
srcAlpha = ( src[ x ] * ( fRenderInfo.fColor >> 24 ) ) / 255;
oneMinusAlpha = 255 - srcAlpha;
destAlpha = destPtr[ x ] & 0xff000000;
dR = ( destPtr[ x ] >> 16 ) & 0x000000ff;
dG = ( destPtr[ x ] >> 8 ) & 0x000000ff;
dB = ( destPtr[ x ] ) & 0x000000ff;
r = ( srcR * srcAlpha ) >> 8;
g = ( srcG * srcAlpha ) >> 8;
b = ( srcB * srcAlpha ) >> 8;
dR = ( dR * oneMinusAlpha ) >> 8;
dG = ( dG * oneMinusAlpha ) >> 8;
dB = ( dB * oneMinusAlpha ) >> 8;
destPtr[ x ] = ( ( r + dR ) << 16 ) | ( ( g + dG ) << 8 ) | ( b + dB ) | destAlpha;
}
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
src += fWidth;
}
}
void plFont::IRenderChar8To32FullAlpha( const plFont::plCharacter &c )
{
UInt8 *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
Int16 x, y, thisHeight, xstart, thisWidth;
UInt32 destColorOnly;
// Unfortunately for some fonts, their right kern value actually is
// farther left than the right edge of the bitmap (think of overlapping
// script fonts). Ideally, we should store the actual width of each char's
// bitmap and use that here. However, it really shouldn't make too big of a
// difference, especially since the dest pixels that we end up overlapping
// should already be in the cache. If it does, time to upgrade the font
// format (again)
thisWidth = (Int16)fWidth;// + (Int32)c.fRightKern;
if( thisWidth >= fRenderInfo.fMaxWidth )
thisWidth = fRenderInfo.fMaxWidth;
xstart = fRenderInfo.fClipRect.fX - fRenderInfo.fX;
if( xstart < 0 )
xstart = 0;
destColorOnly = fRenderInfo.fColor & 0x00ffffff;
y = fRenderInfo.fClipRect.fY - fRenderInfo.fY + (Int16)c.fBaseline;
if( y < 0 )
y = 0;
else
{
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + y*fRenderInfo.fDestStride );
src += y*fRenderInfo.fNumCols;
}
thisHeight = fRenderInfo.fMaxHeight + (Int16)c.fBaseline;
if( thisHeight > (Int16)c.fHeight )
thisHeight = (Int16)c.fHeight;
for( ; y < thisHeight; y++ )
{
destPtr = destBasePtr;
for( x = xstart; x < thisWidth; x++ )
{
if( src[ x ] != 0 )
destPtr[ x ] = ( src[ x ] << 24 ) | destColorOnly;
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
src += fWidth;
}
}
void plFont::IRenderChar8To32Alpha( const plFont::plCharacter &c )
{
UInt8 val, *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
Int16 x, y, thisHeight, xstart, thisWidth;
UInt32 destColorOnly, alphaMult, fullAlpha;
// Unfortunately for some fonts, their right kern value actually is
// farther left than the right edge of the bitmap (think of overlapping
// script fonts). Ideally, we should store the actual width of each char's
// bitmap and use that here. However, it really shouldn't make too big of a
// difference, especially since the dest pixels that we end up overlapping
// should already be in the cache. If it does, time to upgrade the font
// format (again)
thisWidth = (Int16)fWidth;// + (Int32)c.fRightKern;
if( thisWidth >= fRenderInfo.fMaxWidth )
thisWidth = fRenderInfo.fMaxWidth;
xstart = fRenderInfo.fClipRect.fX - fRenderInfo.fX;
if( xstart < 0 )
xstart = 0;
destColorOnly = fRenderInfo.fColor & 0x00ffffff;
// alphaMult should come out to be a value to satisfy (fontAlpha * alphaMult >> 8) as the right alpha,
// but then we want it so (fontAlpha * alphaMult) will be in the upper 8 bits
fullAlpha = fRenderInfo.fColor & 0xff000000;
alphaMult = fullAlpha / 255;
y = fRenderInfo.fClipRect.fY - fRenderInfo.fY + (Int16)c.fBaseline;
if( y < 0 )
y = 0;
else
{
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + y*fRenderInfo.fDestStride );
src += y*fRenderInfo.fNumCols;
}
thisHeight = fRenderInfo.fMaxHeight + (Int16)c.fBaseline;
if( thisHeight > (Int16)c.fHeight )
thisHeight = (Int16)c.fHeight;
for( ; y < thisHeight; y++ )
{
destPtr = destBasePtr;
for( x = xstart; x < thisWidth; x++ )
{
val = src[ x ];
if( val == 0xff )
destPtr[ x ] = fullAlpha | destColorOnly;
else if( val != 0 )
{
destPtr[ x ] = ( ( alphaMult * val ) & 0xff000000 ) | destColorOnly;
}
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
src += fWidth;
}
}
void plFont::IRenderChar8To32AlphaPremultiplied( const plFont::plCharacter &c )
{
UInt8 *src = fBMapData + c.fBitmapOff;
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
Int16 x, y, thisHeight, xstart, thisWidth;
UInt8 srcA, srcR, srcG, srcB;
// Unfortunately for some fonts, their right kern value actually is
// farther left than the right edge of the bitmap (think of overlapping
// script fonts). Ideally, we should store the actual width of each char's
// bitmap and use that here. However, it really shouldn't make too big of a
// difference, especially since the dest pixels that we end up overlapping
// should already be in the cache. If it does, time to upgrade the font
// format (again)
thisWidth = (Int16)fWidth;// + (Int32)c.fRightKern;
if( thisWidth >= fRenderInfo.fMaxWidth )
thisWidth = fRenderInfo.fMaxWidth;
xstart = fRenderInfo.fClipRect.fX - fRenderInfo.fX;
if( xstart < 0 )
xstart = 0;
srcA = (UInt8)(( fRenderInfo.fColor >> 24 ) & 0x000000ff);
srcR = (UInt8)(( fRenderInfo.fColor >> 16 ) & 0x000000ff);
srcG = (UInt8)(( fRenderInfo.fColor >> 8 ) & 0x000000ff);
srcB = (UInt8)(( fRenderInfo.fColor ) & 0x000000ff);
y = fRenderInfo.fClipRect.fY - fRenderInfo.fY + (Int16)c.fBaseline;
if( y < 0 )
y = 0;
else
{
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + y*fRenderInfo.fDestStride );
src += y*fRenderInfo.fNumCols;
}
thisHeight = fRenderInfo.fMaxHeight + (Int16)c.fBaseline;
if( thisHeight > (Int16)c.fHeight )
thisHeight = (Int16)c.fHeight;
for( ; y < thisHeight; y++ )
{
destPtr = destBasePtr;
for( x = xstart; x < thisWidth; x++ )
{
UInt32 a = src[ x ];
if (a != 0)
{
if (srcA != 0xff)
a = (srcA*a + 127)/255;
destPtr[ x ] = ( a << 24 ) | (((srcR*a + 127)/255) << 16) | (((srcG*a + 127)/255) << 8) | ((srcB*a + 127)/255);
}
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
src += fWidth;
}
}
void plFont::IRenderChar8To32AlphaPremShadow( const plFont::plCharacter &c )
{
UInt32 *destPtr, *destBasePtr = (UInt32 *)( fRenderInfo.fDestPtr - c.fBaseline * fRenderInfo.fDestStride );
Int16 x, y, thisHeight, xstart, thisWidth;
UInt8 srcA, srcR, srcG, srcB;
// Unfortunately for some fonts, their right kern value actually is
// farther left than the right edge of the bitmap (think of overlapping
// script fonts). Ideally, we should store the actual width of each char's
// bitmap and use that here. However, it really shouldn't make too big of a
// difference, especially since the dest pixels that we end up overlapping
// should already be in the cache. If it does, time to upgrade the font
// format (again)
thisWidth = (Int16)fWidth + 2;// + (Int32)c.fRightKern;
if( thisWidth >= fRenderInfo.fMaxWidth )
thisWidth = fRenderInfo.fMaxWidth;
xstart = fRenderInfo.fClipRect.fX - fRenderInfo.fX;
if( xstart < -2 )
xstart = -2;
srcA = (UInt8)(( fRenderInfo.fColor >> 24 ) & 0x000000ff);
if (srcA == 0)
return;
srcR = (UInt8)(( fRenderInfo.fColor >> 16 ) & 0x000000ff);
srcG = (UInt8)(( fRenderInfo.fColor >> 8 ) & 0x000000ff);
srcB = (UInt8)(( fRenderInfo.fColor ) & 0x000000ff);
static const UInt32 kernel[5][5] = {
{1, 2, 2, 2, 1},
{1, 13, 13, 13, 1},
{1, 10, 10, 10, 1},
{1, 7, 7, 7, 1},
{1, 1, 1, 1, 1}
};
UInt32 clamp = 220 - ((2 * srcR + 4 * srcG + srcB) >> 4);
y = fRenderInfo.fClipRect.fY - fRenderInfo.fY + (Int16)c.fBaseline;
if( y < -2 )
y = -2;
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + y*fRenderInfo.fDestStride );
thisHeight = fRenderInfo.fMaxHeight + (Int16)c.fBaseline;
if( thisHeight > (Int16)c.fHeight + 2 )
thisHeight = (Int16)c.fHeight + 2;
for( ; y < thisHeight; y++ )
{
destPtr = destBasePtr;
for( x = xstart; x < thisWidth; x++ )
{
UInt32 sa = 0;
for (Int32 i = -2; i <= 2; i++) {
for (Int32 j = -2; j <= 2; j++) {
UInt32 m = kernel[j+2][i+2];
if (m != 0)
sa += m * IGetCharPixel(c, x+i, y+j);
}
}
sa = (sa * clamp) >> 13;
if (sa > clamp)
sa = clamp;
UInt32 a = IGetCharPixel(c, x, y);
if (srcA != 0xff)
{
a = (srcA * a + 127) / 255;
sa = (srcA * sa + 127) / 255;
}
UInt32 ta = a + sa - ((a*sa + 127) / 255);
if (ta > (destPtr[ x ] >> 24))
destPtr[ x ] = ( ta << 24 ) | (((srcR * a + 127) / 255) << 16) |
(((srcG * a + 127) / 255) << 8) |
(((srcB * a + 127) / 255) << 0);
}
destBasePtr = (UInt32 *)( (UInt8 *)destBasePtr + fRenderInfo.fDestStride );
}
}
void plFont::IRenderCharNull( const plCharacter &c )
{
}
//// CalcString Variations ////////////////////////////////////////////////////
UInt16 plFont::CalcStringWidth( const char *string )
{
UInt16 w, h, a, lX, lY;
UInt32 s;
CalcStringExtents( string, w, h, a, s, lX, lY );
return w;
}
UInt16 plFont::CalcStringWidth( const wchar_t *string )
{
UInt16 w, h, a, lX, lY;
UInt32 s;
CalcStringExtents( string, w, h, a, s, lX, lY );
return w;
}
void plFont::CalcStringExtents( const char *string, UInt16 &width, UInt16 &height, UInt16 &ascent, UInt32 &firstClippedChar, UInt16 &lastX, UInt16 &lastY )
{
// convert the char string to a wchar_t string
wchar_t *wideString = hsStringToWString(string);
CalcStringExtents(wideString,width,height,ascent,firstClippedChar,lastX,lastY);
delete [] wideString;
}
void plFont::CalcStringExtents( const wchar_t *string, UInt16 &width, UInt16 &height, UInt16 &ascent, UInt32 &firstClippedChar, UInt16 &lastX, UInt16 &lastY )
{
IRenderString( nil, 0, 0, string, true );
width = fRenderInfo.fFarthestX;
height = (UInt16)(fRenderInfo.fY + fFontDescent);//fRenderInfo.fMaxDescent;
ascent = fRenderInfo.fMaxAscent;
lastX = fRenderInfo.fLastX;
lastY = fRenderInfo.fLastY;
// firstClippedChar is an index into the given string that points to the start of the part of the string
// that got clipped (i.e. not rendered).
firstClippedChar = (UInt32)fRenderInfo.fVolatileStringPtr - (UInt32)string;
firstClippedChar /= 2; // divide by 2 because a wchar_t is two bytes wide, instead of one (like a char)
}
//// IGetFreeCharData /////////////////////////////////////////////////////////
// Used for constructing fonts one character at a time; finds a pointer to
// the first free space in our allocated font bitmap, based on the heights
// and offsets of all the current characters
UInt8 *plFont::IGetFreeCharData( UInt32 &newOffset )
{
Int32 i;
newOffset = 0;
for( i = fCharacters.GetCount() - 1; i >= 0; i-- )
{
UInt32 thisOff = fCharacters[ i ].fBitmapOff + ( ( fCharacters[ i ].fHeight * fWidth * fBPP ) >> 3 );
if( newOffset < thisOff )
{
newOffset = thisOff;
break;
}
}
if( newOffset >= ( ( fWidth * fHeight * fBPP ) >> 3 ) )
{
hsAssert( false, "Invalid position found in IGetFreeCharData()" );
return nil;
}
// Return pointer to the new area
return fBMapData + newOffset;
}
//// LoadFromP2FFile //////////////////////////////////////////////////////////
// Handy quick wrapper
hsBool plFont::LoadFromP2FFile( const char *path )
{
hsUNIXStream stream;
if( stream.Open( path, "rb" ) )
{
ReadRaw( &stream );
return true;
}
return false;
}
//// LoadFromFNT //////////////////////////////////////////////////////////////
// Load this font from the data found in the given Windows FNT file,
// using the format specified in the Windows 3 Developers Notes.
hsBool plFont::LoadFromFNT( const char *path )
{
hsUNIXStream stream; // Ahh, irony
if( !stream.Open( path, "rb" ) )
return false;
return LoadFromFNTStream( &stream );
}
hsBool plFont::LoadFromFNTStream( hsStream *stream )
{
IClear();
try
{
// Note: hsUNIXStreams just happen to store in the same endian as Windows...
struct FNTInfo
{
UInt16 version;
UInt32 size;
char copyright[ 60 ];
UInt16 type;
UInt16 points;
UInt16 vertRes;
UInt16 horzRes;
UInt16 ascent;
UInt16 internalLeading;
UInt16 externalLeading;
UInt8 italic, underline, strikeout;
UInt16 weight;
UInt8 charSet;
UInt16 pixWidth; // 0 means variable width chars
UInt16 pixHeight;
UInt8 pitchFamily;
UInt16 avgWidth;
UInt16 maxWidth;
UInt8 firstChar, lastChar, defaultChar, breakChar;
UInt16 widthBytes;
UInt32 device, face;
UInt32 bitsPointer, bitsOffset;
UInt8 reserved;
UInt32 flags;
UInt16 aSpace, bSpace, cSpace;
UInt32 colorPointer;
UInt8 reserved1[ 16 ];
void Read( hsStream *s )
{
version = s->ReadSwap16();
size = s->ReadSwap32();
s->Read( sizeof( copyright ), copyright );
s->ReadSwap( &type );
s->ReadSwap( &points );
s->ReadSwap( &vertRes );
s->ReadSwap( &horzRes );
s->ReadSwap( &ascent );
s->ReadSwap( &internalLeading );
s->ReadSwap( &externalLeading );
s->ReadSwap( &italic );
s->ReadSwap( &underline );
s->ReadSwap( &strikeout );
s->ReadSwap( &weight );
s->ReadSwap( &charSet );
s->ReadSwap( &pixWidth );
s->ReadSwap( &pixHeight );
s->ReadSwap( &pitchFamily );
s->ReadSwap( &avgWidth );
s->ReadSwap( &maxWidth );
s->ReadSwap( &firstChar );
s->ReadSwap( &lastChar );
s->ReadSwap( &defaultChar );
s->ReadSwap( &breakChar );
s->ReadSwap( &widthBytes );
s->ReadSwap( &device );
s->ReadSwap( &face );
s->ReadSwap( &bitsPointer );
s->ReadSwap( &bitsOffset );
s->ReadSwap( &reserved );
if( version == 0x0300 )
{
s->ReadSwap( &flags );
s->ReadSwap( &aSpace );
s->ReadSwap( &bSpace );
s->ReadSwap( &cSpace );
s->ReadSwap( &colorPointer );
s->Read( sizeof( reserved1 ), reserved1 );
}
else
{
flags = 0;
aSpace = bSpace = cSpace = 0;
colorPointer = 0;
}
}
} fntInfo;
fntInfo.Read( stream );
struct charEntry
{
UInt16 width;
UInt32 offset;
charEntry() { width = 0; offset = 0; }
} *charEntries;
int i, count = fntInfo.lastChar - fntInfo.firstChar + 2;
charEntries = TRACKED_NEW charEntry[ count ];
for( i = 0; i < count; i++ )
{
charEntries[ i ].width = stream->ReadSwap16();
if( fntInfo.version == 0x0200 )
charEntries[ i ].offset = stream->ReadSwap16();
else
charEntries[ i ].offset = stream->ReadSwap32();
}
char faceName[ 256 ], deviceName[ 256 ];
if( fntInfo.face != 0 )
{
stream->SetPosition( fntInfo.face );
for( i = 0; i < 256; i++ )
{
faceName[ i ] = stream->ReadByte();
if( faceName[ i ] == 0 )
break;
}
strncpy( fFace, faceName, sizeof( fFace ) );
}
if( fntInfo.device != 0 )
{
stream->SetPosition( fntInfo.device );
for( i = 0; i < 256; i++ )
{
deviceName[ i ] = stream->ReadByte();
if( deviceName[ i ] == 0 )
break;
}
}
fSize = (UInt8)(fntInfo.points);
// Figure out what we need to allocate our bitmap as
fWidth = 0;
fHeight = 0;
for( i = 0; i < count; i++ )
{
if( fWidth < charEntries[ i ].width )
fWidth = charEntries[ i ].width;
if( charEntries[ i ].offset > 0 )
fHeight += fntInfo.pixHeight;
}
fBPP = 1;
// Since we're 1 bbp, make sure width is a multiple of 8
fWidth = ( ( fWidth + 7 ) >> 3 ) << 3;
UInt32 widthInBytes = ( fWidth * fBPP ) >> 3;
// Allocate our bitmap now
UInt32 size = widthInBytes * fHeight;
fBMapData = TRACKED_NEW UInt8[ size ];
memset( fBMapData, 0, size );
fFirstChar = fntInfo.firstChar;
fMaxCharHeight = 0;
// Read the bitmap info
UInt32 destOff = 0;
for( i = 0; i < count; i++ )
{
int numCols = ( charEntries[ i ].width + 7 ) >> 3;
int height = fntInfo.pixHeight;
// Write a record for this char
plCharacter outChar;
outChar.fBitmapOff = destOff;
outChar.fHeight = ( charEntries[ i ].offset == 0 ) ? 0 : height;
outChar.fBaseline = fntInfo.ascent;
outChar.fLeftKern = 0.f;
outChar.fRightKern = (hsScalar)(charEntries[ i ].width - fWidth);
fCharacters.Append( outChar );
if( outChar.fHeight > fMaxCharHeight )
fMaxCharHeight = outChar.fHeight;
if( charEntries[ i ].offset == 0 )
continue;
// Seek to this char
stream->SetPosition( charEntries[ i ].offset );
// Write the actual bitmap data. Note: FNTs store the bits one column at a time,
// one col after another, whereas we want it plain row by row
int col, y;
UInt8 *basePtr = fBMapData + destOff;
for( col = 0; col < numCols; col++ )
{
UInt8 *yPtr = basePtr;
for( y = 0; y < height; y++ )
{
*yPtr = stream->ReadByte();
yPtr += widthInBytes;
}
basePtr++;
}
destOff += widthInBytes * outChar.fHeight;
}
delete [] charEntries;
ICalcFontAscent();
return true;
}
catch( ... )
{
// Somehow we crashed converting!
IClear();
return false;
}
}
//// LoadFromBDF //////////////////////////////////////////////////////////////
// Load this font from the data found in the given Adobe Systems BDF file,
// using the format specified in the Glyph Bitmap Distribution Format (BDF)
// Specification Version 2.2 from Adobe Systems.
/*hsBool plFont::LoadFromBDF( const char *path, plBDFConvertCallback *callback )
{
hsUNIXStream stream; // Ahh, irony
if( !stream.Open( path, "rb" ) )
return false;
return LoadFromBDFStream( &stream, callback );
}
*/
// Some parsing helpers
typedef int (*fnCharTester)( int );
static int sQuoteTester( int c )
{
return ( c == '\"' );
}
static int sDashTester( int c )
{
return isspace( c ) || ( c == '-' );
}
class plLineParser
{
protected:
static char fLine[ 512 ];
char *fCursor, fRestore;
void IAdvanceToNextToken( fnCharTester tester = isspace )
{
// Last cursor
*fCursor = fRestore;
// Scan for the start of the next token
while( *fCursor != 0 && (*tester)( *fCursor ) )
fCursor++;
if( *fCursor == 0 )
return;
// This is the start of our token
const char *start = fCursor;
// Put a stopper here
fRestore = *fCursor;
*fCursor = 0;
// And return!
return;
}
const char *IGetNextToken( fnCharTester tester = isspace )
{
// Last cursor
*fCursor = fRestore;
// Scan for the start of the next token
while( *fCursor != 0 && (*tester)( *fCursor ) )
fCursor++;
if( *fCursor == 0 )
return nil;
// This is the start of our token; find the end
const char *start = fCursor;
while( *fCursor != 0 && !(*tester)( *fCursor ) )
fCursor++;
// Put a stopper here
fRestore = *fCursor;
*fCursor = 0;
// And return!
return start;
}
public:
plLineParser( const char *line )
{
strncpy( fLine, line, sizeof( fLine ) );
fCursor = fLine;
fRestore = *fCursor;
}
~plLineParser() { }
const char *GetKeyword( void )
{
return IGetNextToken();
}
const char *GetString( void )
{
IAdvanceToNextToken();
return IGetNextToken( sQuoteTester );
}
const char *GetKeywordNoDashes( void )
{
return IGetNextToken( sDashTester );
}
Int32 GetInt( void )
{
return atoi( IGetNextToken() );
}
hsScalar GetFloat( void )
{
return (hsScalar)atof( IGetNextToken() );
}
};
char plLineParser::fLine[ 512 ];
// Another helper--parser for various sections of the BDF format
class plBDFSectParser
{
protected:
plFont &fFont;
plBDFConvertCallback *fCallback;
public:
plBDFSectParser( plFont &myFont, plBDFConvertCallback *callback ) : fFont( myFont ), fCallback( callback ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
return nil;
}
};
class plBDFLookForEndCharParser : public plBDFSectParser
{
public:
plBDFLookForEndCharParser( plFont &myFont, plBDFConvertCallback *callback ) : plBDFSectParser( myFont, callback ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line );
};
inline UInt8 iHexCharToByte( char c )
{
switch( c )
{
case '0': return 0;
case '1': return 1;
case '2': return 2;
case '3': return 3;
case '4': return 4;
case '5': return 5;
case '6': return 6;
case '7': return 7;
case '8': return 8;
case '9': return 9;
case 'a': case 'A': return 10;
case 'b': case 'B': return 11;
case 'c': case 'C': return 12;
case 'd': case 'D': return 13;
case 'e': case 'E': return 14;
case 'f': case 'F': return 15;
default: return 0;
}
}
inline UInt8 iHexStringToByte( const char *str )
{
return ( iHexCharToByte( str[ 0 ] ) << 4 ) | iHexCharToByte( str[ 1 ] );
}
class plBDFCharsParser : public plBDFSectParser
{
protected:
// Info about the current character we're translating
UInt16 fWhichChar, fRowsLeft;
plFont::plCharacter *fCharacter;
UInt8 *fBitmap;
hsBool fDoingData;
UInt32 fBytesWide, fBMapStride;
inline void IReset( void )
{
fBitmap = nil;
fCharacter = nil;
fWhichChar = 0;
fDoingData = false;
}
public:
static UInt32 fResolution;
plBDFCharsParser( plFont &myFont, plBDFConvertCallback *callback ) : plBDFSectParser( myFont, callback ), fDoingData( false ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
if( strcmp( keyword, "ENDFONT" ) == 0 )
{
// All done!
return nil;
}
else if( strcmp( keyword, "ENDCHAR" ) == 0 )
{
// End of the character, reset
IReset();
if( fCallback != nil )
fCallback->CharDone();
}
else if( fDoingData )
{
// If we're doing data, all lines are hex values until we hit "ENDCHAR"
if( fRowsLeft == 0 )
throw;
int hDigit;
for( hDigit = 0; *keyword != 0 && hDigit < fBytesWide; hDigit++, keyword += 2 )
{
*fBitmap = iHexStringToByte( keyword );
fBitmap++;
}
fBitmap += fBMapStride - hDigit;
fRowsLeft--;
}
else if( strcmp( keyword, "STARTCHAR" ) == 0 )
{
// Start of a new character--ignore the name, we'll just use the encoding
IReset();
}
else if( strcmp( keyword, "ENCODING" ) == 0 )
{
int ch = line.GetInt();
if( ch == -1 )
{
// Nonstandard encoding, skip the character entirely
return TRACKED_NEW plBDFLookForEndCharParser( fFont, fCallback );
}
else
{
fWhichChar = (UInt16)(UInt8)ch;
// Set up our pointer. Note that since BDFs don't tell us the starting character,
// we just make it 0
if( fFont.fCharacters.GetCount() < fWhichChar + 1 )
fFont.fCharacters.SetCount( fWhichChar + 1 );
fCharacter = &fFont.fCharacters[ fWhichChar ];
}
}
// Horizontal mode offsets (writing mode 0)
else if( strcmp( keyword, "SWIDTH" ) == 0 )
{
// In device units, unused
// float fWidth = (float)line.GetInt() * ( (float)fFont.GetSize() / 1000.f ) * ( fResolution / 72.f );
// fCharacter->fRightKern = fWidth - (float)fFont.fWidth;
// fCharacter->fLeftKern = 0.f;
}
else if( strcmp( keyword, "DWIDTH" ) == 0 )
{
// In pixels, basically the offset to the next char. Note that we only
// support the X direction (altho admittedly the idea of a font whose characters cause
// the next chars to go up or down is way nifty)
fCharacter->fRightKern = (float)line.GetInt() - (float)fFont.fWidth;
fCharacter->fLeftKern = 0.f;
}
// Vertical mode offsets (unsupported)
else if( strcmp( keyword, "SWIDTH1" ) == 0 )
{
}
else if( strcmp( keyword, "DWIDTH1" ) == 0 )
{
}
// Bitmap bounding box
else if( strcmp( keyword, "BBX" ) == 0 )
{
int pixW = line.GetInt();
int pixH = line.GetInt();
int xOff = line.GetInt();
int yOff = line.GetInt();
// Got enough info now to allocate us a bitmap for this char
fBitmap = fFont.IGetFreeCharData( fCharacter->fBitmapOff );
if( fBitmap == nil )
throw false;
if( fCharacter->fBitmapOff > ( ( fFont.fWidth * ( fFont.fHeight - pixH ) * fFont.fBPP ) >> 3 ) )
{
hsAssert( false, "Invalid position found in IGetFreeCharData()" );
return nil;
}
// Set these now, since setting them before would've changed the IGetFreeCharData results
fCharacter->fLeftKern = (hsScalar)xOff;
fCharacter->fBaseline = yOff + pixH;
fCharacter->fHeight = fRowsLeft = pixH;
fBytesWide = ( pixW + 7 ) >> 3;
fBMapStride = fFont.fWidth >> 3;
}
// Bitmap data
else if( strcmp( keyword, "BITMAP" ) == 0 )
{
fDoingData = true;
}
// This keyword is outputted from the "getbdf" utility on linux, for some reason...
else if( strcmp( keyword, "ATTRIBUTES" ) == 0 )
{
}
else
throw false; // Invalid keyword
return this;
}
};
UInt32 plBDFCharsParser::fResolution = 72;
plBDFSectParser *plBDFLookForEndCharParser::ParseKeyword( const char *keyword, plLineParser &line )
{
if( strcmp( keyword, "ENDCHAR" ) == 0 )
{
// Horray!
return TRACKED_NEW plBDFCharsParser( fFont, fCallback );
}
// Just gobble lines until we find the endchar section
return this;
}
class plBDFLookForCharParser : public plBDFSectParser
{
public:
plBDFLookForCharParser( plFont &myFont, plBDFConvertCallback *callback ) : plBDFSectParser( myFont, callback ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
if( strcmp( keyword, "CHARS" ) == 0 )
{
// Horray!
return TRACKED_NEW plBDFCharsParser( fFont, fCallback );
}
// Just gobble lines until we find the chars section
return this;
}
};
class plBDFPropertiesParser : public plBDFSectParser
{
public:
plBDFPropertiesParser( plFont &myFont, plBDFConvertCallback *callback ) : plBDFSectParser( myFont, callback ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
// Note: the properties section is entirely optional and arbitrary, but we
// parse it in case we can get more accurate info about the font name and props
if( strcmp( keyword, "FACE_NAME" ) == 0 )
{
fFont.SetFace( line.GetString() );
}
else if( strcmp( keyword, "WEIGHT_NAME" ) == 0 )
{
if( stricmp( line.GetString(), "Bold" ) == 0 )
fFont.SetFlag( plFont::kFlagBold, true );
}
else if( strcmp( keyword, "ENDPROPERTIES" ) == 0 )
{
// Switch to waiting for the chars section
return TRACKED_NEW plBDFLookForCharParser( fFont, fCallback );
}
// All tokens are technically valid, even if we don't recognize them
return this;
}
};
class plBDFHeaderParser : public plBDFSectParser
{
public:
plBDFHeaderParser( plFont &myFont, plBDFConvertCallback *callback ) : plBDFSectParser( myFont, callback ) {}
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
if( strcmp( keyword, "STARTFONT" ) == 0 )
{
// Start of the font; check version #
float version = line.GetFloat();
if( version < 2.1f )
throw false;
// Initial font values
fFont.fFirstChar = 0;
fFont.fMaxCharHeight = 0;
}
else if( strcmp( keyword, "FONT" ) == 0 )
{
// Postscript-style font name: figure out our face name and hopefully attributes too
// Format is usually of the form: vendor-face-weight-otherStuff
const char *vendor = line.GetKeywordNoDashes();
const char *face = line.GetKeywordNoDashes();
fFont.SetFace( face != nil ? face : vendor );
const char *weight = line.GetKeywordNoDashes();
if( weight != nil && stricmp( weight, "Bold" ) == 0 )
fFont.SetFlag( plFont::kFlagBold, true );
}
else if( strcmp( keyword, "COMMENT" ) == 0 )
{
// Comment, ignore
}
else if( strcmp( keyword, "CONTENTVERSION" ) == 0 )
{
// Content version, app specific, not used by us
}
else if( strcmp( keyword, "SIZE" ) == 0 )
{
// Point size and target resolution of this font
fFont.SetSize( (UInt8)(line.GetInt()) );
plBDFCharsParser::fResolution = line.GetInt();
}
else if( strcmp( keyword, "FONTBOUNDINGBOX" ) == 0 )
{
// Max dimensions and base offsets
// Note that we should've already grabbed the width from our
// prescan earlier, to guard against malformed BDFs
int thisWidth = line.GetInt();
if( fFont.fWidth < thisWidth )
fFont.fWidth = thisWidth;
// Since we're 1 bbp, make sure width is a multiple of 8
fFont.fWidth = ( ( fFont.fWidth + 7 ) >> 3 ) << 3;
// Allocate our data now
fFont.fBMapData = TRACKED_NEW UInt8[ ( fFont.fWidth * fFont.fHeight * fFont.fBPP ) >> 3 ];
memset( fFont.fBMapData, 0, ( fFont.fWidth * fFont.fHeight * fFont.fBPP ) >> 3 );
fFont.fMaxCharHeight = line.GetInt();
}
else if( strcmp( keyword, "METRICSSET" ) == 0 )
{
// Metrics set specification; unsupported
}
else if( strcmp( keyword, "STARTPROPERTIES" ) == 0 )
{
// Start of the properties section, switch to that
return TRACKED_NEW plBDFPropertiesParser( fFont, fCallback );
}
else if( strcmp( keyword, "CHARS" ) == 0 )
{
// Allocate our bitmap if we haven't already
// Since we're 1 bbp, make sure width is a multiple of 8
if( fFont.fBMapData == nil )
{
fFont.fWidth = ( ( fFont.fWidth + 7 ) >> 3 ) << 3;
// Allocate our data now
fFont.fBMapData = TRACKED_NEW UInt8[ ( fFont.fWidth * fFont.fHeight * fFont.fBPP ) >> 3 ];
memset( fFont.fBMapData, 0, ( fFont.fWidth * fFont.fHeight * fFont.fBPP ) >> 3 );
}
// Start of the char section
return TRACKED_NEW plBDFCharsParser( fFont, fCallback );
}
else
{
// Unrecognized token, abort
throw false;
}
// Default, keep going with our parser
return this;
}
};
class plBDFCheckDimsParser : public plBDFSectParser
{
hsBool fSkipNext;
public:
UInt32 fMaxWidth, fMaxHeight, fNumChars, fTotalHeight;
UInt16 fMaxChar;
plBDFCheckDimsParser( plFont &myFont ) : plBDFSectParser( myFont, nil ) { fMaxWidth = fMaxHeight = fNumChars = fTotalHeight = 0; fSkipNext = false; fMaxChar = 0; }
virtual plBDFSectParser *ParseKeyword( const char *keyword, plLineParser &line )
{
if( strcmp( keyword, "ENDFONT" ) == 0 )
{
// All done!
return nil;
}
// Encoding
else if( strcmp( keyword, "ENCODING" ) == 0 )
{
int ch = line.GetInt();
if( ch == -1 )
// We skip these
fSkipNext = true;
else
{
fSkipNext = false;
if( fMaxChar < ch )
fMaxChar = ch;
}
}
// Bitmap bounding box
else if( strcmp( keyword, "BBX" ) == 0 )
{
if( fSkipNext )
return this;
int pixW = line.GetInt();
int pixH = line.GetInt();
int xOff = line.GetInt();
int yOff = line.GetInt();
if( fMaxWidth < pixW )
fMaxWidth = pixW;
if( fMaxHeight < pixH )
fMaxHeight = pixH;
fNumChars++;
fTotalHeight += pixH;
}
return this;
}
};
hsBool plFont::LoadFromBDFStream( hsStream *stream, plBDFConvertCallback *callback )
{
return false;
}
hsBool plFont::LoadFromBDF( const char *path, plBDFConvertCallback *callback )
{
FILE *fp = fopen( path, "rt" );
if( fp == nil )
return false;
try
{
IClear();
char line[ 512 ];
// Run through the entire file first with a plBDFCheckDimsParser. This is because
// some BDFs are naughty and don't report the correct fontboundingbox (i.e. too small)
// (See the following loop below for details on the workings of this loop)
plBDFCheckDimsParser checkDims( *this );
while( fgets( line, sizeof( line ), fp ) )
{
plLineParser parser( line );
const char *keyword = parser.GetKeyword();
if( keyword != nil )
{
if( checkDims.ParseKeyword( keyword, parser ) == nil )
break;
}
}
// Set up initial values
fWidth = checkDims.fMaxWidth;
fHeight = checkDims.fTotalHeight;
fBPP = 1;
// Pre-expand our char list
fCharacters.ExpandAndZero( checkDims.fMaxChar + 1 );
fCharacters.SetCount( 0 );
if( callback != nil )
callback->NumChars( (UInt16)(checkDims.fNumChars) );
// Rewind and continue normally
fseek( fp, 0, SEEK_SET );
// Start with the header parser
plBDFSectParser *currParser = TRACKED_NEW plBDFHeaderParser( *this, callback );
// Read from the stream one line at a time (don't want comments, and char #1 should be invalid for a BDF)
while( fgets( line, sizeof( line ), fp ) && currParser != nil )
{
// Parse this one
plLineParser parser( line );
// Get the keyword for this line
const char *keyword = parser.GetKeyword();
if( keyword != nil )
{
// Pass on to the current parser
plBDFSectParser *newParser = currParser->ParseKeyword( keyword, parser );
if( newParser != currParser )
{
delete currParser;
currParser = newParser;
}
}
}
}
catch( ... )
{
IClear();
return false;
}
fclose( fp );
ICalcFontAscent();
return true;
}
hsBool plFont::ReadRaw( hsStream *s )
{
s->Read( sizeof( fFace ), fFace );
fSize = s->ReadByte();
s->ReadSwap( &fFlags );
s->ReadSwap( &fWidth );
s->ReadSwap( &fHeight );
s->ReadSwap( &fMaxCharHeight );
fBPP = s->ReadByte();
UInt32 size = ( fWidth * fHeight * fBPP ) >> 3;
if( size > 0 )
{
fBMapData = TRACKED_NEW UInt8[ size ];
s->Read( size, fBMapData );
}
else
fBMapData = nil;
s->ReadSwap( &fFirstChar );
UInt32 i;
fCharacters.SetCountAndZero( s->ReadSwap32() );
for( i = 0; i < fCharacters.GetCount(); i++ )
fCharacters[ i ].Read( s );
ICalcFontAscent();
return true;
}
hsBool plFont::WriteRaw( hsStream *s )
{
s->Write( sizeof( fFace ), fFace );
s->WriteByte( fSize );
s->WriteSwap( fFlags );
s->WriteSwap( fWidth );
s->WriteSwap( fHeight );
s->WriteSwap( fMaxCharHeight );
s->WriteByte( fBPP );
UInt32 size = ( fWidth * fHeight * fBPP ) >> 3;
if( size > 0 )
s->Write( size, fBMapData );
s->WriteSwap( fFirstChar );
UInt32 i;
s->WriteSwap32( fCharacters.GetCount() );
for( i = 0; i < fCharacters.GetCount(); i++ )
fCharacters[ i ].Write( s );
return true;
}