include/vcl/fontcharmap.hxx | 1
vcl/inc/impfontcharmap.hxx | 1
vcl/source/font/fontcharmap.cxx | 284 ----------------------------------------
3 files changed, 286 deletions(-)
New commits:
commit b59d13a7835b18f4676c3a11e9d70dfa9cb934b5
Author: Khaled Hosny <kha...@aliftype.com>
AuthorDate: Thu Sep 8 02:24:36 2022 +0200
Commit: خالد حسني <kha...@aliftype.com>
CommitDate: Thu Sep 8 16:32:20 2022 +0200
vcl: Drop now unused ParseCMAP()
Change-Id: I2140b7b48ee4cc46a4a6e8e4719f3ce78a600829
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/139624
Tested-by: Jenkins
Reviewed-by: خالد حسني <kha...@aliftype.com>
diff --git a/include/vcl/fontcharmap.hxx b/include/vcl/fontcharmap.hxx
index 4b21142509da..d79ffce7be92 100644
--- a/include/vcl/fontcharmap.hxx
+++ b/include/vcl/fontcharmap.hxx
@@ -164,7 +164,6 @@ public:
const sal_UCS4* mpRangeCodes;
int mnRangeCount;
bool mbSymbolic;
- bool mbRecoded;
};
#endif // INCLUDED_FONTCHARMAP_HXX
diff --git a/vcl/inc/impfontcharmap.hxx b/vcl/inc/impfontcharmap.hxx
index 8a0b0c413d56..36809cdd986b 100644
--- a/vcl/inc/impfontcharmap.hxx
+++ b/vcl/inc/impfontcharmap.hxx
@@ -51,7 +51,6 @@ private:
};
bool VCL_DLLPUBLIC HasSymbolCmap(const char* pRawData, int nRawLength);
-bool VCL_DLLPUBLIC ParseCMAP( const unsigned char* pRawData, int nRawLength,
CmapResult& );
#endif // INCLUDED_VCL_INC_IMPFONTCHARMAP_HXX
diff --git a/vcl/source/font/fontcharmap.cxx b/vcl/source/font/fontcharmap.cxx
index 24f1903f7165..5896e786eca0 100644
--- a/vcl/source/font/fontcharmap.cxx
+++ b/vcl/source/font/fontcharmap.cxx
@@ -18,20 +18,16 @@
#include <utility>
#include <vcl/fontcharmap.hxx>
#include <impfontcharmap.hxx>
-#include <rtl/textcvt.h>
-#include <rtl/textenc.h>
#include <sal/log.hxx>
#include <algorithm>
#include <vector>
-#include <o3tl/sorted_vector.hxx>
CmapResult::CmapResult( bool bSymbolic,
const sal_UCS4* pRangeCodes, int nRangeCount )
: mpRangeCodes( pRangeCodes)
, mnRangeCount( nRangeCount)
, mbSymbolic( bSymbolic)
-, mbRecoded( false)
{}
static ImplFontCharMapRef g_pDefaultImplFontCharMap;
@@ -109,286 +105,6 @@ bool HasSymbolCmap(const char* pCmap, int nLength)
return false;
}
-static unsigned GetUInt( const unsigned char* p ) {
return((p[0]<<24)+(p[1]<<16)+(p[2]<<8)+p[3]);}
-static unsigned GetUShort( const unsigned char* p ){ return((p[0]<<8) | p[1]);}
-
-// TODO: move CMAP parsing directly into the ImplFontCharMap class
-bool ParseCMAP( const unsigned char* pCmap, int nLength, CmapResult& rResult )
-{
- rResult.mpRangeCodes = nullptr;
- rResult.mnRangeCount = 0;
- rResult.mbRecoded = false;
- rResult.mbSymbolic = false;
-
- // parse the table header and check for validity
- if( !pCmap || (nLength < 24) )
- return false;
-
- if( GetUShort( pCmap ) != 0x0000 ) // simple check for CMAP corruption
- return false;
-
- int nSubTables = GetUShort( pCmap + 2 );
- if( (nSubTables <= 0) || (nSubTables > (nLength - 24) / 8) )
- return false;
-
- // find the most interesting subtable in the CMAP
- rtl_TextEncoding eRecodeFrom = RTL_TEXTENCODING_UNICODE;
- int nOffset = 0;
- int nFormat = -1;
- int nBestVal = 0;
- for( const unsigned char* p = pCmap + 4; --nSubTables >= 0; p += 8 )
- {
- int nPlatform = GetUShort( p );
- int nEncoding = GetUShort( p+2 );
- int nPlatformEncoding = (nPlatform << 8) + nEncoding;
-
- int nValue;
- rtl_TextEncoding eTmpEncoding = RTL_TEXTENCODING_UNICODE;
- switch( nPlatformEncoding )
- {
- case 0x000: nValue = 20; break; //
Unicode 1.0
- case 0x001: nValue = 21; break; //
Unicode 1.1
- case 0x002: nValue = 22; break; //
iso10646_1993
- case 0x003: nValue = 23; break; //
UCS-2
- case 0x004: nValue = 24; break; //
UCS-4
- case 0x100: nValue = 22; break; // Mac
Unicode<2.0
- case 0x103: nValue = 23; break; // Mac
Unicode>2.0
- case 0x300: nValue = 5; rResult.mbSymbolic = true; break; // Win
Symbol
- case 0x301: nValue = 28; break; // Win
UCS-2
- case 0x30A: nValue = 29; break; //
Win-UCS-4
- case 0x302: nValue = 11; eTmpEncoding =
RTL_TEXTENCODING_SHIFT_JIS; break;
- case 0x303: nValue = 12; eTmpEncoding = RTL_TEXTENCODING_GB_18030;
break;
- case 0x304: nValue = 11; eTmpEncoding = RTL_TEXTENCODING_BIG5;
break;
- case 0x305: nValue = 11; eTmpEncoding = RTL_TEXTENCODING_MS_949;
break;
- case 0x306: nValue = 11; eTmpEncoding = RTL_TEXTENCODING_MS_1361;
break;
- default: nValue = 0; break;
- }
-
- if( nValue <= 0 ) // ignore unknown encodings
- continue;
-
- int nTmpOffset = GetUInt( p+4 );
-
- if (nTmpOffset > nLength - 2 || nTmpOffset < 0)
- continue;
-
- int nTmpFormat = GetUShort( pCmap + nTmpOffset );
- if( nTmpFormat == 12 ) // 32bit code -> glyph map
format
- nValue += 3;
- else if( nTmpFormat != 4 ) // 16bit code -> glyph map
format
- continue; // ignore other formats
-
- if( nBestVal < nValue )
- {
- nBestVal = nValue;
- nOffset = nTmpOffset;
- nFormat = nTmpFormat;
- eRecodeFrom = eTmpEncoding;
- }
- }
-
- // parse the best CMAP subtable
- int nRangeCount = 0;
- sal_UCS4* pCodePairs = nullptr;
-
- // format 4, the most common 16bit char mapping table
- if( (nFormat == 4) && ((nOffset+16) < nLength) )
- {
- int nSegCountX2 = GetUShort( pCmap + nOffset + 6 );
- nRangeCount = nSegCountX2/2 - 1;
- if (nRangeCount < 0)
- {
- SAL_WARN("vcl.gdi", "negative RangeCount");
- nRangeCount = 0;
- }
-
- const unsigned char* pLimitBase = pCmap + nOffset + 14;
- const unsigned char* pBeginBase = pLimitBase + nSegCountX2 + 2;
- const unsigned char* pDeltaBase = pBeginBase + nSegCountX2;
- const unsigned char* pOffsetBase = pDeltaBase + nSegCountX2;
-
- const int nOffsetBaseStart = pOffsetBase - pCmap;
- const int nRemainingLen = nLength - nOffsetBaseStart;
- const int nMaxPossibleRangeOffsets = nRemainingLen / 2;
- if (nRangeCount > nMaxPossibleRangeOffsets)
- {
- SAL_WARN("vcl.gdi", "more range offsets requested then space
available");
- nRangeCount = std::max(0, nMaxPossibleRangeOffsets);
- }
-
- pCodePairs = new sal_UCS4[ nRangeCount * 2 ];
-
- sal_UCS4* pCP = pCodePairs;
- for( int i = 0; i < nRangeCount; ++i )
- {
- const sal_UCS4 cMinChar = GetUShort( pBeginBase + 2*i );
- const sal_UCS4 cMaxChar = GetUShort( pLimitBase + 2*i );
- if( cMinChar > cMaxChar ) { // no sane font should trigger this
- SAL_WARN("vcl.gdi", "Min char should never be more than the
max char!");
- break;
- }
- if( cMaxChar == 0xFFFF ) {
- SAL_WARN("vcl.gdi", "Format 4 char should not be 0xFFFF");
- break;
- }
- *(pCP++) = cMinChar;
- *(pCP++) = cMaxChar + 1;
- }
- nRangeCount = (pCP - pCodePairs) / 2;
- }
- // format 12, the most common 32bit char mapping table
- else if( (nFormat == 12) && ((nOffset+16) < nLength) )
- {
- nRangeCount = GetUInt( pCmap + nOffset + 12 );
- if (nRangeCount < 0)
- {
- SAL_WARN("vcl.gdi", "negative RangeCount");
- nRangeCount = 0;
- }
-
- const int nGroupOffset = nOffset + 16;
- const int nRemainingLen = nLength - nGroupOffset;
- const int nMaxPossiblePairs = nRemainingLen / 12;
- if (nRangeCount > nMaxPossiblePairs)
- {
- SAL_WARN("vcl.gdi", "more code pairs requested then space
available");
- nRangeCount = std::max(0, nMaxPossiblePairs);
- }
-
- pCodePairs = new sal_UCS4[ nRangeCount * 2 ];
-
- const unsigned char* pGroup = pCmap + nGroupOffset;
- sal_UCS4* pCP = pCodePairs;
- for( int i = 0; i < nRangeCount; ++i )
- {
- sal_UCS4 cMinChar = GetUInt( pGroup + 0 );
- sal_UCS4 cMaxChar = GetUInt( pGroup + 4 );
- pGroup += 12;
-
- if( cMinChar > cMaxChar ) { // no sane font should trigger this
- SAL_WARN("vcl.gdi", "Min char should never be more than the
max char!");
- break;
- }
-
- *(pCP++) = cMinChar;
- *(pCP++) = cMaxChar + 1;
- }
- nRangeCount = (pCP - pCodePairs) / 2;
- }
-
- // check if any subtable resulted in something usable
- if( nRangeCount <= 0 )
- {
- delete[] pCodePairs;
-
- // even when no CMAP is available we know it for symbol fonts
- if( rResult.mbSymbolic )
- {
- pCodePairs = new sal_UCS4[4];
- pCodePairs[0] = 0x0020; // aliased symbols
- pCodePairs[1] = 0x0100;
- pCodePairs[2] = 0xF020; // original symbols
- pCodePairs[3] = 0xF100;
- rResult.mpRangeCodes = pCodePairs;
- rResult.mnRangeCount = 2;
- return true;
- }
-
- return false;
- }
-
- // recode the code ranges to their unicode encoded ranges if needed
- rtl_TextToUnicodeConverter aConverter = nullptr;
- rtl_UnicodeToTextContext aCvtContext = nullptr;
-
- rResult.mbRecoded = ( eRecodeFrom != RTL_TEXTENCODING_UNICODE );
- if( rResult.mbRecoded )
- {
- aConverter = rtl_createTextToUnicodeConverter( eRecodeFrom );
- aCvtContext = rtl_createTextToUnicodeContext( aConverter );
- }
-
- if( aConverter && aCvtContext )
- {
- // determine the set of supported code points from encoded ranges
- o3tl::sorted_vector<sal_UCS4> aSupportedCodePoints;
- aSupportedCodePoints.reserve(256);
-
- static const int NINSIZE = 64;
- static const int NOUTSIZE = 64;
- std::vector<char> cCharsInp;
- cCharsInp.reserve(NINSIZE);
- sal_Unicode cCharsOut[ NOUTSIZE ];
- sal_UCS4* pCP = pCodePairs;
- for( int i = 0; i < nRangeCount; ++i )
- {
- sal_UCS4 cMin = *(pCP++);
- sal_UCS4 cEnd = *(pCP++);
- // ofz#25868 the conversion only makes sense with
- // input codepoints in 0..SAL_MAX_UINT16 range
- while (cMin < cEnd && cMin <= SAL_MAX_UINT16)
- {
- for (int j = 0; (cMin < cEnd) && (j < NINSIZE); ++cMin, ++j)
- {
- if( cMin >= 0x0100 )
- cCharsInp.push_back(static_cast<char>(cMin >> 8));
- if( (cMin >= 0x0100) || (cMin < 0x00A0) )
- cCharsInp.push_back(static_cast<char>(cMin));
- }
-
- sal_uInt32 nCvtInfo;
- sal_Size nSrcCvtBytes;
- int nOutLen = rtl_convertTextToUnicode(
- aConverter, aCvtContext,
- cCharsInp.data(), cCharsInp.size(), cCharsOut, NOUTSIZE,
- RTL_TEXTTOUNICODE_FLAGS_INVALID_IGNORE
- | RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_IGNORE,
- &nCvtInfo, &nSrcCvtBytes );
-
- cCharsInp.clear();
-
- for (int j = 0; j < nOutLen; ++j)
- aSupportedCodePoints.insert( cCharsOut[j] );
- }
- }
-
- rtl_destroyTextToUnicodeConverter( aCvtContext );
- rtl_destroyTextToUnicodeConverter( aConverter );
-
- // convert the set of supported code points to ranges
- std::vector<sal_UCS4> aSupportedRanges;
-
- for (auto const& supportedPoint : aSupportedCodePoints)
- {
- if( aSupportedRanges.empty()
- || (aSupportedRanges.back() != supportedPoint) )
- {
- // add new range beginning with current unicode
- aSupportedRanges.push_back(supportedPoint);
- aSupportedRanges.push_back( 0 );
- }
-
- // extend existing range to include current unicode
- aSupportedRanges.back() = supportedPoint + 1;
- }
-
- // make a pCodePairs array using the vector from above
- delete[] pCodePairs;
- nRangeCount = aSupportedRanges.size() / 2;
- if( nRangeCount <= 0 )
- return false;
- pCodePairs = new sal_UCS4[ nRangeCount * 2 ];
- pCP = pCodePairs;
- for (auto const& supportedRange : aSupportedRanges)
- *(pCP++) = supportedRange;
- }
-
- // update the result struct
- rResult.mpRangeCodes = pCodePairs;
- rResult.mnRangeCount = nRangeCount;
- return true;
-}
-
FontCharMap::FontCharMap()
: mpImplFontCharMap( ImplFontCharMap::getDefaultMap() )
{
