summaryrefslogtreecommitdiffstats
path: root/kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp
diff options
context:
space:
mode:
Diffstat (limited to 'kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp')
-rw-r--r--kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp961
1 files changed, 961 insertions, 0 deletions
diff --git a/kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp b/kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp
new file mode 100644
index 00000000..b14220d1
--- /dev/null
+++ b/kviewshell/plugins/djvu/libdjvu/MMRDecoder.cpp
@@ -0,0 +1,961 @@
+//C- -*- C++ -*-
+//C- -------------------------------------------------------------------
+//C- DjVuLibre-3.5
+//C- Copyright (c) 2002 Leon Bottou and Yann Le Cun.
+//C- Copyright (c) 2001 AT&T
+//C-
+//C- This software is subject to, and may be distributed under, the
+//C- GNU General Public License, Version 2. The license should have
+//C- accompanied the software or you may obtain a copy of the license
+//C- from the Free Software Foundation at http://www.fsf.org .
+//C-
+//C- This program is distributed in the hope that it will be useful,
+//C- but WITHOUT ANY WARRANTY; without even the implied warranty of
+//C- MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+//C- GNU General Public License for more details.
+//C-
+//C- DjVuLibre-3.5 is derived from the DjVu(r) Reference Library
+//C- distributed by Lizardtech Software. On July 19th 2002, Lizardtech
+//C- Software authorized us to replace the original DjVu(r) Reference
+//C- Library notice by the following text (see doc/lizard2002.djvu):
+//C-
+//C- ------------------------------------------------------------------
+//C- | DjVu (r) Reference Library (v. 3.5)
+//C- | Copyright (c) 1999-2001 LizardTech, Inc. All Rights Reserved.
+//C- | The DjVu Reference Library is protected by U.S. Pat. No.
+//C- | 6,058,214 and patents pending.
+//C- |
+//C- | This software is subject to, and may be distributed under, the
+//C- | GNU General Public License, Version 2. The license should have
+//C- | accompanied the software or you may obtain a copy of the license
+//C- | from the Free Software Foundation at http://www.fsf.org .
+//C- |
+//C- | The computer code originally released by LizardTech under this
+//C- | license and unmodified by other parties is deemed "the LIZARDTECH
+//C- | ORIGINAL CODE." Subject to any third party intellectual property
+//C- | claims, LizardTech grants recipient a worldwide, royalty-free,
+//C- | non-exclusive license to make, use, sell, or otherwise dispose of
+//C- | the LIZARDTECH ORIGINAL CODE or of programs derived from the
+//C- | LIZARDTECH ORIGINAL CODE in compliance with the terms of the GNU
+//C- | General Public License. This grant only confers the right to
+//C- | infringe patent claims underlying the LIZARDTECH ORIGINAL CODE to
+//C- | the extent such infringement is reasonably necessary to enable
+//C- | recipient to make, have made, practice, sell, or otherwise dispose
+//C- | of the LIZARDTECH ORIGINAL CODE (or portions thereof) and not to
+//C- | any greater extent that may be necessary to utilize further
+//C- | modifications or combinations.
+//C- |
+//C- | The LIZARDTECH ORIGINAL CODE is provided "AS IS" WITHOUT WARRANTY
+//C- | OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED
+//C- | TO ANY WARRANTY OF NON-INFRINGEMENT, OR ANY IMPLIED WARRANTY OF
+//C- | MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+//C- +------------------------------------------------------------------
+//
+// $Id: MMRDecoder.cpp,v 1.8 2003/11/07 22:08:22 leonb Exp $
+// $Name: release_3_5_15 $
+
+#ifdef HAVE_CONFIG_H
+# include "config.h"
+#endif
+#if NEED_GNUG_PRAGMAS
+# pragma implementation
+#endif
+
+#include "MMRDecoder.h"
+#include "JB2Image.h"
+#include "ByteStream.h"
+#include "GBitmap.h"
+
+
+#ifdef HAVE_NAMESPACES
+namespace DJVU {
+# ifdef NOT_DEFINED // Just to fool emacs c++ mode
+}
+#endif
+#endif
+
+
+// ----------------------------------------
+// MMR CODEBOOKS
+
+static const char invalid_mmr_data[]= ERR_MSG("MMRDecoder.bad_data");
+
+struct VLCode
+{
+ unsigned short code;
+ short codelen;
+ short value;
+};
+
+enum MMRMode
+{
+ P=0, H=1, V0=2, VR1=3, VR2=4, VR3=5, VL1=6, VL2=7, VL3=8
+};
+
+static const VLCode mrcodes[] =
+{ // Codes on 7 bits
+ // 7 bit codes
+ { 0x08, 4, P }, // 0001
+ { 0x10, 3, H }, // 001
+ { 0x40, 1, V0 }, // 1
+ { 0x30, 3, VR1 }, // 011
+ { 0x06, 6, VR2 }, // 000011
+ { 0x03, 7, VR3 }, // 0000011
+ { 0x20, 3, VL1 }, // 010
+ { 0x04, 6, VL2 }, // 000010
+ { 0x02, 7, VL3 }, // 0000010
+ { 0x00, 0, -1 } // Illegal entry
+};
+
+
+static const VLCode wcodes[] = {
+ // 13 bit codes
+ { 0x06a0, 8, 0 }, // 00110101
+ { 0x0380, 6, 1 }, // 000111
+ { 0x0e00, 4, 2 }, // 0111
+ { 0x1000, 4, 3 }, // 1000
+ { 0x1600, 4, 4 }, // 1011
+ { 0x1800, 4, 5 }, // 1100
+ { 0x1c00, 4, 6 }, // 1110
+ { 0x1e00, 4, 7 }, // 1111
+ { 0x1300, 5, 8 }, // 10011
+ { 0x1400, 5, 9 }, // 10100
+ { 0x0700, 5, 10 }, // 00111
+ { 0x0800, 5, 11 }, // 01000
+ { 0x0400, 6, 12 }, // 001000
+ { 0x0180, 6, 13 }, // 000011
+ { 0x1a00, 6, 14 }, // 110100
+ { 0x1a80, 6, 15 }, // 110101
+ { 0x1500, 6, 16 }, // 101010
+ { 0x1580, 6, 17 }, // 101011
+ { 0x09c0, 7, 18 }, // 0100111
+ { 0x0300, 7, 19 }, // 0001100
+ { 0x0200, 7, 20 }, // 0001000
+ { 0x05c0, 7, 21 }, // 0010111
+ { 0x00c0, 7, 22 }, // 0000011
+ { 0x0100, 7, 23 }, // 0000100
+ { 0x0a00, 7, 24 }, // 0101000
+ { 0x0ac0, 7, 25 }, // 0101011
+ { 0x04c0, 7, 26 }, // 0010011
+ { 0x0900, 7, 27 }, // 0100100
+ { 0x0600, 7, 28 }, // 0011000
+ { 0x0040, 8, 29 }, // 00000010
+ { 0x0060, 8, 30 }, // 00000011
+ { 0x0340, 8, 31 }, // 00011010
+ { 0x0360, 8, 32 }, // 00011011
+ { 0x0240, 8, 33 }, // 00010010
+ { 0x0260, 8, 34 }, // 00010011
+ { 0x0280, 8, 35 }, // 00010100
+ { 0x02a0, 8, 36 }, // 00010101
+ { 0x02c0, 8, 37 }, // 00010110
+ { 0x02e0, 8, 38 }, // 00010111
+ { 0x0500, 8, 39 }, // 00101000
+ { 0x0520, 8, 40 }, // 00101001
+ { 0x0540, 8, 41 }, // 00101010
+ { 0x0560, 8, 42 }, // 00101011
+ { 0x0580, 8, 43 }, // 00101100
+ { 0x05a0, 8, 44 }, // 00101101
+ { 0x0080, 8, 45 }, // 00000100
+ { 0x00a0, 8, 46 }, // 00000101
+ { 0x0140, 8, 47 }, // 00001010
+ { 0x0160, 8, 48 }, // 00001011
+ { 0x0a40, 8, 49 }, // 01010010
+ { 0x0a60, 8, 50 }, // 01010011
+ { 0x0a80, 8, 51 }, // 01010100
+ { 0x0aa0, 8, 52 }, // 01010101
+ { 0x0480, 8, 53 }, // 00100100
+ { 0x04a0, 8, 54 }, // 00100101
+ { 0x0b00, 8, 55 }, // 01011000
+ { 0x0b20, 8, 56 }, // 01011001
+ { 0x0b40, 8, 57 }, // 01011010
+ { 0x0b60, 8, 58 }, // 01011011
+ { 0x0940, 8, 59 }, // 01001010
+ { 0x0960, 8, 60 }, // 01001011
+ { 0x0640, 8, 61 }, // 00110010
+ { 0x0660, 8, 62 }, // 00110011
+ { 0x0680, 8, 63 }, // 00110100
+ { 0x1b00, 5, 64 }, // 11011
+ { 0x1200, 5, 128 }, // 10010
+ { 0x0b80, 6, 192 }, // 010111
+ { 0x0dc0, 7, 256 }, // 0110111
+ { 0x06c0, 8, 320 }, // 00110110
+ { 0x06e0, 8, 384 }, // 00110111
+ { 0x0c80, 8, 448 }, // 01100100
+ { 0x0ca0, 8, 512 }, // 01100101
+ { 0x0d00, 8, 576 }, // 01101000
+ { 0x0ce0, 8, 640 }, // 01100111
+ { 0x0cc0, 9, 704 }, // 011001100
+ { 0x0cd0, 9, 768 }, // 011001101
+ { 0x0d20, 9, 832 }, // 011010010
+ { 0x0d30, 9, 896 }, // 011010011
+ { 0x0d40, 9, 960 }, // 011010100
+ { 0x0d50, 9, 1024 }, // 011010101
+ { 0x0d60, 9, 1088 }, // 011010110
+ { 0x0d70, 9, 1152 }, // 011010111
+ { 0x0d80, 9, 1216 }, // 011011000
+ { 0x0d90, 9, 1280 }, // 011011001
+ { 0x0da0, 9, 1344 }, // 011011010
+ { 0x0db0, 9, 1408 }, // 011011011
+ { 0x0980, 9, 1472 }, // 010011000
+ { 0x0990, 9, 1536 }, // 010011001
+ { 0x09a0, 9, 1600 }, // 010011010
+ { 0x0c00, 6, 1664 }, // 011000 (what did they think?)
+ { 0x09b0, 9, 1728 }, // 010011011
+ { 0x0020, 11, 1792 }, // 00000001000
+ { 0x0030, 11, 1856 }, // 00000001100
+ { 0x0034, 11, 1920 }, // 00000001101
+ { 0x0024, 12, 1984 }, // 000000010010
+ { 0x0026, 12, 2048 }, // 000000010011
+ { 0x0028, 12, 2112 }, // 000000010100
+ { 0x002a, 12, 2176 }, // 000000010101
+ { 0x002c, 12, 2240 }, // 000000010110
+ { 0x002e, 12, 2304 }, // 000000010111
+ { 0x0038, 12, 2368 }, // 000000011100
+ { 0x003a, 12, 2432 }, // 000000011101
+ { 0x003c, 12, 2496 }, // 000000011110
+ { 0x003e, 12, 2560 }, // 000000011111
+ { 0x0000, 0, -1 } // Illegal entry
+};
+
+
+static const VLCode bcodes[] = {
+ // 13 bit codes
+ { 0x01b8, 10, 0 }, // 0000110111
+ { 0x0800, 3, 1 }, // 010
+ { 0x1800, 2, 2 }, // 11
+ { 0x1000, 2, 3 }, // 10
+ { 0x0c00, 3, 4 }, // 011
+ { 0x0600, 4, 5 }, // 0011
+ { 0x0400, 4, 6 }, // 0010
+ { 0x0300, 5, 7 }, // 00011
+ { 0x0280, 6, 8 }, // 000101
+ { 0x0200, 6, 9 }, // 000100
+ { 0x0100, 7, 10 }, // 0000100
+ { 0x0140, 7, 11 }, // 0000101
+ { 0x01c0, 7, 12 }, // 0000111
+ { 0x0080, 8, 13 }, // 00000100
+ { 0x00e0, 8, 14 }, // 00000111
+ { 0x0180, 9, 15 }, // 000011000
+ { 0x00b8, 10, 16 }, // 0000010111
+ { 0x00c0, 10, 17 }, // 0000011000
+ { 0x0040, 10, 18 }, // 0000001000
+ { 0x019c, 11, 19 }, // 00001100111
+ { 0x01a0, 11, 20 }, // 00001101000
+ { 0x01b0, 11, 21 }, // 00001101100
+ { 0x00dc, 11, 22 }, // 00000110111
+ { 0x00a0, 11, 23 }, // 00000101000
+ { 0x005c, 11, 24 }, // 00000010111
+ { 0x0060, 11, 25 }, // 00000011000
+ { 0x0194, 12, 26 }, // 000011001010
+ { 0x0196, 12, 27 }, // 000011001011
+ { 0x0198, 12, 28 }, // 000011001100
+ { 0x019a, 12, 29 }, // 000011001101
+ { 0x00d0, 12, 30 }, // 000001101000
+ { 0x00d2, 12, 31 }, // 000001101001
+ { 0x00d4, 12, 32 }, // 000001101010
+ { 0x00d6, 12, 33 }, // 000001101011
+ { 0x01a4, 12, 34 }, // 000011010010
+ { 0x01a6, 12, 35 }, // 000011010011
+ { 0x01a8, 12, 36 }, // 000011010100
+ { 0x01aa, 12, 37 }, // 000011010101
+ { 0x01ac, 12, 38 }, // 000011010110
+ { 0x01ae, 12, 39 }, // 000011010111
+ { 0x00d8, 12, 40 }, // 000001101100
+ { 0x00da, 12, 41 }, // 000001101101
+ { 0x01b4, 12, 42 }, // 000011011010
+ { 0x01b6, 12, 43 }, // 000011011011
+ { 0x00a8, 12, 44 }, // 000001010100
+ { 0x00aa, 12, 45 }, // 000001010101
+ { 0x00ac, 12, 46 }, // 000001010110
+ { 0x00ae, 12, 47 }, // 000001010111
+ { 0x00c8, 12, 48 }, // 000001100100
+ { 0x00ca, 12, 49 }, // 000001100101
+ { 0x00a4, 12, 50 }, // 000001010010
+ { 0x00a6, 12, 51 }, // 000001010011
+ { 0x0048, 12, 52 }, // 000000100100
+ { 0x006e, 12, 53 }, // 000000110111
+ { 0x0070, 12, 54 }, // 000000111000
+ { 0x004e, 12, 55 }, // 000000100111
+ { 0x0050, 12, 56 }, // 000000101000
+ { 0x00b0, 12, 57 }, // 000001011000
+ { 0x00b2, 12, 58 }, // 000001011001
+ { 0x0056, 12, 59 }, // 000000101011
+ { 0x0058, 12, 60 }, // 000000101100
+ { 0x00b4, 12, 61 }, // 000001011010
+ { 0x00cc, 12, 62 }, // 000001100110
+ { 0x00ce, 12, 63 }, // 000001100111
+ { 0x0078, 10, 64 }, // 0000001111
+ { 0x0190, 12, 128 }, // 000011001000
+ { 0x0192, 12, 192 }, // 000011001001
+ { 0x00b6, 12, 256 }, // 000001011011
+ { 0x0066, 12, 320 }, // 000000110011
+ { 0x0068, 12, 384 }, // 000000110100
+ { 0x006a, 12, 448 }, // 000000110101
+ { 0x006c, 13, 512 }, // 0000001101100
+ { 0x006d, 13, 576 }, // 0000001101101
+ { 0x004a, 13, 640 }, // 0000001001010
+ { 0x004b, 13, 704 }, // 0000001001011
+ { 0x004c, 13, 768 }, // 0000001001100
+ { 0x004d, 13, 832 }, // 0000001001101
+ { 0x0072, 13, 896 }, // 0000001110010
+ { 0x0073, 13, 960 }, // 0000001110011
+ { 0x0074, 13, 1024 }, // 0000001110100
+ { 0x0075, 13, 1088 }, // 0000001110101
+ { 0x0076, 13, 1152 }, // 0000001110110
+ { 0x0077, 13, 1216 }, // 0000001110111
+ { 0x0052, 13, 1280 }, // 0000001010010
+ { 0x0053, 13, 1344 }, // 0000001010011
+ { 0x0054, 13, 1408 }, // 0000001010100
+ { 0x0055, 13, 1472 }, // 0000001010101
+ { 0x005a, 13, 1536 }, // 0000001011010
+ { 0x005b, 13, 1600 }, // 0000001011011
+ { 0x0064, 13, 1664 }, // 0000001100100
+ { 0x0065, 13, 1728 }, // 0000001100101
+ { 0x0020, 11, 1792 }, // 00000001000
+ { 0x0030, 11, 1856 }, // 00000001100
+ { 0x0034, 11, 1920 }, // 00000001101
+ { 0x0024, 12, 1984 }, // 000000010010
+ { 0x0026, 12, 2048 }, // 000000010011
+ { 0x0028, 12, 2112 }, // 000000010100
+ { 0x002a, 12, 2176 }, // 000000010101
+ { 0x002c, 12, 2240 }, // 000000010110
+ { 0x002e, 12, 2304 }, // 000000010111
+ { 0x0038, 12, 2368 }, // 000000011100
+ { 0x003a, 12, 2432 }, // 000000011101
+ { 0x003c, 12, 2496 }, // 000000011110
+ { 0x003e, 12, 2560 }, // 000000011111
+ { 0x0000, 0, -1 } // Illegal entry
+};
+
+
+
+
+// ----------------------------------------
+// SOURCE OF BITS
+
+#define VLSBUFSIZE 64
+
+class MMRDecoder::VLSource : public GPEnabled
+{
+protected:
+ VLSource(GP<ByteStream> &inp);
+ void init(const bool striped);
+public:
+ // Initializes a bit source on a bytestream
+ static GP<VLSource> create(GP<ByteStream> &inp, const bool striped);
+
+ // Synchronize on the next stripe
+ void nextstripe(void);
+ // Returns a 32 bits integer with at least the
+ // next sixteen code bits in the high order bits.
+ inline unsigned int peek(void);
+ // Ensures that next #peek()# contains at least
+ // the next 24 code bits.
+ void preload(void);
+ // Consumes #n# bits.
+ void shift(const int n);
+private:
+ GP<ByteStream> ginp;
+ ByteStream &inp;
+ unsigned char buffer[ VLSBUFSIZE ];
+ unsigned int codeword;
+ int lowbits;
+ int bufpos;
+ int bufmax;
+ int readmax;
+};
+
+MMRDecoder::VLSource::VLSource(GP<ByteStream> &xinp)
+: ginp(xinp), inp(*ginp), codeword(0),
+ lowbits(0), bufpos(0), bufmax(0),
+ readmax(-1)
+{}
+
+void
+MMRDecoder::VLSource::init(const bool striped)
+{
+ if (striped)
+ readmax = inp.read32();
+ lowbits = 32;
+ preload();
+}
+
+GP<MMRDecoder::VLSource>
+MMRDecoder::VLSource::create(GP<ByteStream> &inp, const bool striped)
+{
+ VLSource *src=new VLSource(inp);
+ GP<VLSource> retval=src;
+ src->init(striped);
+ return retval;
+}
+
+void
+MMRDecoder::VLSource::shift(const int n)
+{
+ codeword<<=n;
+ lowbits+=n;
+ if (lowbits>=16)
+ preload();
+}
+
+inline unsigned int
+MMRDecoder::VLSource::peek(void)
+{
+ return codeword;
+}
+
+
+void
+MMRDecoder::VLSource::nextstripe(void)
+{
+ while (readmax>0)
+ {
+ int size = sizeof(buffer);
+ if (readmax < size)
+ size = readmax;
+ inp.readall(buffer, size);
+ readmax -= size;
+ }
+ bufpos = bufmax = 0;
+ memset(buffer,0,sizeof(buffer));
+ readmax = inp.read32();
+ codeword = 0;
+ lowbits = 32;
+ preload();
+}
+
+void
+MMRDecoder::VLSource::preload(void)
+{
+ while (lowbits>=8)
+ {
+ if (bufpos >= bufmax)
+ {
+ // Refill buffer
+ bufpos = bufmax = 0;
+ int size = sizeof(buffer);
+ if (readmax>=0 && readmax<size)
+ size = readmax;
+ if (size>0)
+ bufmax = inp.read((void*)buffer, size);
+ readmax -= bufmax;
+ if (bufmax <= 0)
+ return;
+ }
+ lowbits -= 8;
+ codeword |= buffer[bufpos++] << lowbits;
+ }
+}
+
+
+
+// ----------------------------------------
+// VARIABLE LENGTH CODES
+
+
+
+class MMRDecoder::VLTable : public GPEnabled
+{
+protected:
+ VLTable(const VLCode *codes);
+ void init(const int nbits);
+public:
+ // Construct a VLTable given a codebook with #nbits# long codes.
+ static GP<VLTable> create(VLCode const * const codes, const int nbits);
+
+ // Reads one symbol from a VLSource
+ int decode(MMRDecoder::VLSource *src);
+
+ const VLCode *code;
+ int codewordshift;
+ unsigned char *index;
+ GPBuffer<unsigned char> gindex;
+};
+
+GP<MMRDecoder::VLTable>
+MMRDecoder::VLTable::create(VLCode const * const codes, const int nbits)
+{
+ VLTable *table=new VLTable(codes);
+ GP<VLTable> retval=table;
+ table->init(nbits);
+ return retval;
+}
+
+inline int
+MMRDecoder::VLTable::decode(MMRDecoder::VLSource *src)
+{
+ const VLCode &c = code[ index[ src->peek() >> codewordshift ] ];
+ src->shift(c.codelen);
+ return c.value;
+}
+
+MMRDecoder::VLTable::VLTable(const VLCode *codes)
+: code(codes), codewordshift(0), gindex(index,0)
+{}
+
+void
+MMRDecoder::VLTable::init(const int nbits)
+{
+ // count entries
+ int ncodes = 0;
+ while (code[ncodes].codelen)
+ ncodes++;
+ // check arguments
+ if (nbits<=1 || nbits>16)
+ G_THROW(invalid_mmr_data);
+ if (ncodes>=256)
+ G_THROW(invalid_mmr_data);
+ codewordshift = 32 - nbits;
+ // allocate table
+ int size = (1<<nbits);
+ gindex.resize(size);
+ gindex.set(ncodes);
+ // process codes
+ for (int i=0; i<ncodes; i++) {
+ const int c = code[i].code;
+ const int b = code[i].codelen;
+ if(b<=0 || b>nbits)
+ {
+ G_THROW(invalid_mmr_data);
+ }
+ // fill table entries whose index high bits are code.
+ int n = c + (1<<(nbits-b));
+ while ( --n >= c ) {
+ if(index[n] != ncodes)
+ G_THROW( ERR_MSG("MMRDecoder.bad_codebook") );
+ index[n] = i;
+ }
+ }
+}
+
+// ----------------------------------------
+// MMR DECODER
+
+
+
+MMRDecoder::~MMRDecoder() {}
+
+MMRDecoder::MMRDecoder( const int xwidth, const int xheight )
+: width(xwidth), height(xheight), lineno(0),
+ striplineno(0), rowsperstrip(0), gline(line,width+8),
+ glineruns(lineruns,width+4), gprevruns(prevruns,width+4)
+{
+ gline.clear();
+ glineruns.clear();
+ gprevruns.clear();
+ lineruns[0] = width;
+ prevruns[0] = width;
+}
+
+void
+MMRDecoder::init(GP<ByteStream> gbs, const bool striped)
+{
+ rowsperstrip = (striped ? gbs->read16() : height);
+ src = VLSource::create(gbs, striped);
+ mrtable = VLTable::create(mrcodes, 7);
+ btable = VLTable::create(bcodes, 13);
+ wtable = VLTable::create(wcodes, 13);
+}
+
+GP<MMRDecoder>
+MMRDecoder::create( GP<ByteStream> gbs, const int width,
+ const int height, const bool striped )
+{
+ MMRDecoder *mmr=new MMRDecoder(width,height);
+ GP<MMRDecoder> retval=mmr;
+ mmr->init(gbs,striped);
+ return retval;
+}
+
+const unsigned short *
+MMRDecoder::scanruns(const unsigned short **endptr)
+{
+ // Check if all lines have been returned
+ if (lineno >= height)
+ return 0;
+ // Check end of stripe
+ if ( striplineno == rowsperstrip )
+ {
+ striplineno=0;
+ lineruns[0] = prevruns[0] = width;
+ src->nextstripe();
+ }
+ // Swap run buffers
+ unsigned short *pr = lineruns;
+ unsigned short *xr = prevruns;
+ prevruns = pr;
+ lineruns = xr;
+ // Loop until scanline is complete
+ bool a0color = false;
+ int a0,rle,b1;
+ for(a0=0,rle=0,b1=*pr++;a0 < width;)
+ {
+ // Process MMR codes
+ const int c=mrtable->decode(src);
+ switch ( c )
+ {
+ /* Pass Mode */
+ case P:
+ {
+ b1 += *pr++;
+ rle += b1 - a0;
+ a0 = b1;
+ b1 += *pr++;
+ break;
+ }
+ /* Horizontal Mode */
+ case H:
+ {
+ // First run
+ VLTable &table1 = *(a0color ? btable : wtable);
+ int inc;
+ do { inc=table1.decode(src); a0+=inc; rle+=inc; } while (inc>=64);
+ *xr = rle; xr++; rle = 0;
+ // Second run
+ VLTable &table2 = *(!a0color ? btable : wtable);
+ do { inc=table2.decode(src); a0+=inc; rle+=inc; } while (inc>=64);
+ *xr = rle; xr++; rle = 0;
+ break;
+ }
+ /* Vertical Modes */
+ case V0:
+ case VR3:
+ case VR2:
+ case VR1:
+ case VL3:
+ case VL2:
+ case VL1:
+ {
+ int inc=b1;
+ switch ( c )
+ {
+ case V0:
+ inc = b1;
+ b1 += *pr++;
+ break;
+ case VR3:
+ inc = b1+3;
+ b1 += *pr++;
+ break;
+ case VR2:
+ inc = b1+2;
+ b1 += *pr++;
+ break;
+ case VR1:
+ inc = b1+1;
+ b1 += *pr++;
+ break;
+ case VL3:
+ inc = b1-3;
+ b1 -= *--pr;
+ break;
+ case VL2:
+ inc = b1-2;
+ b1 -= *--pr;
+ break;
+ case VL1:
+ inc = b1-1;
+ b1 -= *--pr;
+ break;
+ }
+ *xr = inc+rle-a0;
+ xr++;
+ a0 = inc;
+ rle = 0;
+ a0color = !a0color;
+ break;
+ }
+ /* Uncommon modes */
+ default:
+ {
+ src->preload();
+ unsigned int m = src->peek();
+ // -- Could be EOFB ``000000000001000000000001''
+ // TIFF6 says that all remaining lines are white
+ if ((m & 0xffffff00) == 0x00100100)
+ {
+ lineno = height;
+ return 0;
+ }
+ // -- Could be UNCOMPRESSED ``0000001111''
+ // TIFF6 says people should not do this.
+ // RFC1314 says people should do this.
+ else if ((m & 0xffc00000) == 0x03c00000)
+ {
+#ifdef MMRDECODER_REFUSES_UNCOMPRESSED
+ G_THROW( ERR_MSG("MMRDecoder.cant_process") );
+#else
+ // ---THE-FOLLOWING-CODE-IS-POORLY-TESTED---
+ src->shift(10);
+ while ((m = (src->peek() & 0xfc000000)))
+ {
+ if (m == 0x04000000) // 000001
+ {
+ src->shift(6);
+ if (a0color)
+ {
+ *xr = rle;
+ xr++;
+ rle = 0;
+ a0color = !a0color;
+ }
+ rle += 5;
+ a0 += 5;
+ }
+ else // 000010 to 111111
+ {
+ src->shift(1);
+ if (a0color == !(m & 0x80000000))
+ {
+ *xr = rle;
+ xr++;
+ rle = 0;
+ a0color = !a0color;
+ }
+ rle++;
+ a0++;
+ }
+ if (a0 > width)
+ G_THROW(invalid_mmr_data);
+ }
+ // Analyze uncompressed termination code.
+ m = src->peek() & 0xff000000;
+ src->shift(8);
+ if ( (m & 0xfe000000) != 0x02000000 )
+ G_THROW(invalid_mmr_data);
+ if (rle)
+ {
+ *xr = rle;
+ xr++;
+ rle = 0;
+ a0color = !a0color;
+ }
+ if (a0color == !(m & 0x01000000))
+ {
+ *xr = rle;
+ xr++;
+ rle = 0;
+ a0color = !a0color;
+ }
+ // Cross fingers and proceed ...
+ break;
+#endif
+ }
+ // -- Unknown MMR code.
+ G_THROW(invalid_mmr_data);
+ }
+ }
+ // Next reference run
+ for(;b1<=a0 && b1<width;pr+=2)
+ {
+ b1 += pr[0]+pr[1];
+ }
+ }
+ // Final P must be followed by V0 (they say!)
+ if (rle > 0)
+ {
+ if (mrtable->decode(src) != V0)
+ {
+ G_THROW(invalid_mmr_data);
+ }
+ }
+ if (rle > 0)
+ {
+ *xr = rle;
+ xr++;
+ }
+ // At this point we should have A0 equal to WIDTH
+ // But there are buggy files around (Kofax!)
+ // and we are not the CCITT police.
+ if (a0 > width)
+ {
+ while (a0 > width && xr > lineruns)
+ a0 -= *--xr;
+ if (a0 < width)
+ {
+ *xr = width-a0;
+ xr++;
+ }
+ }
+ /* Increment and return */
+ if (endptr)
+ *endptr = xr;
+ xr[0] = 0;
+ xr[1] = 0;
+ lineno ++;
+ striplineno ++;
+ return lineruns;
+}
+
+
+
+const unsigned char *
+MMRDecoder::scanrle(const bool invert, const unsigned char **endptr)
+{
+ // Obtain run lengths
+ const unsigned short *xr = scanruns();
+ if (!xr) return 0;
+ unsigned char *p=line;
+ // Process inversion
+ if (invert)
+ {
+ if (! *xr)
+ {
+ xr++;
+ }else
+ {
+ *p = 0; p++;
+ }
+ }
+ // Encode lenghts using the RLE format
+ for(int a0=0;a0 < width;)
+ {
+ int count = *xr++;
+ a0 += count;
+ GBitmap::append_run(p, count);
+ }
+ if (endptr)
+ *endptr = p;
+ p[0] = 0;
+ p[1] = 0;
+ return line;
+}
+
+
+#if 0
+const unsigned char *
+MMRDecoder::scanline(void)
+{
+ // Obtain run lengths
+ const unsigned short *xr = scanruns();
+ if (!xr) return 0;
+ // Allocate data buffer if needed
+ unsigned char *p = line;
+ // Decode run lengths
+ int a0 = 0;
+ int a0color = 0;
+ while (a0 < width)
+ {
+ int a1 = a0 + *xr++;
+ while (a0<a1 && a0<width)
+ line[a0++] = a0color;
+ a0color = !a0color;
+ }
+ return line;
+}
+#endif
+
+
+
+
+// ----------------------------------------
+// MAIN DECODING ROUTINE
+
+bool
+MMRDecoder::decode_header(
+ ByteStream &inp, int &width, int &height, int &invert)
+{
+ unsigned long int magic = inp.read32();
+ if((magic&0xfffffffc) != 0x4d4d5200)
+ G_THROW( ERR_MSG("MMRDecoder.unrecog_header") );
+ invert = ((magic & 0x1) ? 1 : 0);
+ const bool strip = ((magic & 0x2) ? 1 : 0);
+ width = inp.read16();
+ height = inp.read16();
+ if (width<=0 || height<=0)
+ G_THROW( ERR_MSG("MMRDecoder.bad_header") );
+ return strip;
+}
+
+static inline int MAX(int a, int b) { return a>b ? a : b; }
+static inline int MIN(int a, int b) { return a<b ? a : b; }
+
+GP<JB2Image>
+MMRDecoder::decode(GP<ByteStream> gbs)
+{
+ ByteStream &inp=*gbs;
+ // Read header
+ int width, height, invert;
+ const bool striped=decode_header(inp, width, height, invert);
+ // Prepare image
+ GP<JB2Image> jimg = JB2Image::create();
+ jimg->set_dimension(width, height);
+ // Choose pertinent blocksize
+ int blocksize = MIN(500,MAX(64,MAX(width/17,height/22)));
+ int blocksperline = (width+blocksize-1)/blocksize;
+ // Prepare decoder
+ GP<MMRDecoder> gdcd=MMRDecoder::create(gbs, width, height, striped);
+ MMRDecoder &dcd=*gdcd;
+ // Loop on JB2 bands
+ int line = height-1;
+ while (line >= 0)
+ {
+ int bandline = MIN(blocksize-1,line);
+ GPArray<GBitmap> blocks(0,blocksperline-1);
+ // Loop on scanlines
+ for(; bandline >= 0; bandline--,line--)
+ {
+ // Decode one scanline
+ const unsigned short *s = dcd.scanruns();
+ if (s)
+ {
+ // Loop on blocks
+ int x = 0;
+ int b = 0;
+ int firstx = 0;
+ bool c = !!invert;
+ while (x < width)
+ {
+ int xend = x + *s++;
+ while (b<blocksperline)
+ {
+ int lastx = MIN(firstx+blocksize,width);
+ if (c)
+ {
+ if (!blocks[b])
+ blocks[b] = GBitmap::create(bandline+1, lastx-firstx);
+ unsigned char *bptr = (*blocks[b])[bandline] - firstx;
+ int x1 = MAX(x,firstx);
+ int x2 = MIN(xend,lastx);
+ while (x1 < x2)
+ bptr[x1++] = 1;
+ }
+ if (xend < lastx)
+ break;
+ firstx = lastx;
+ b ++;
+ }
+ x = xend;
+ c = !c;
+ }
+ }
+ }
+ // Insert blocks into JB2Image
+ for (int b=0; b<blocksperline; b++)
+ {
+ JB2Shape shape;
+ shape.bits = blocks[b];
+ if (shape.bits)
+ {
+ shape.parent = -1;
+ shape.bits->compress();
+ JB2Blit blit;
+ blit.left = b*blocksize;
+ blit.bottom = line+1;
+ blit.shapeno = jimg->add_shape(shape);
+ jimg->add_blit(blit);
+ }
+ }
+ }
+ // Return
+ return jimg;
+}
+
+
+
+#ifdef HAVE_NAMESPACES
+}
+# ifndef NOT_USING_DJVU_NAMESPACE
+using namespace DJVU;
+# endif
+#endif