cheng/libtiff
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
3ac0c950c6
libtiff/tools/tiff2ps.c

3133 lines
99 KiB
C
Raw Normal View History

Initial revision
1999-07-27 17:50:26 -04:00
/*
* Copyright (c) 1988-1997 Sam Leffler
* Copyright (c) 1991-1997 Silicon Graphics, Inc.
*
* Permission to use, copy, modify, distribute, and sell this software and
* its documentation for any purpose is hereby granted without fee, provided
* that (i) the above copyright notices and this permission notice appear in
* all copies of the software and related documentation, and (ii) the names of
* Sam Leffler and Silicon Graphics may not be used in any advertising or
* publicity relating to the software without the specific, prior written
* permission of Sam Leffler and Silicon Graphics.
*
* THE SOFTWARE IS PROVIDED "AS-IS" AND WITHOUT WARRANTY OF ANY KIND,
* EXPRESS, IMPLIED OR OTHERWISE, INCLUDING WITHOUT LIMITATION, ANY
* WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
*
* IN NO EVENT SHALL SAM LEFFLER OR SILICON GRAPHICS BE LIABLE FOR
* ANY SPECIAL, INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND,
* OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
* WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE, AND ON ANY THEORY OF
* LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
* OF THIS SOFTWARE.
*/
Avoid zero division in setupPageState() function; properly initialize array in PSDataBW().
2004-08-25 09:14:18 -04:00
#include "tif_config.h"
Initial revision
1999-07-27 17:50:26 -04:00
#include <stdio.h>
#include <stdlib.h> /* for atof */
#include <math.h>
#include <time.h>
#include <strings.h> replaced with #include <string.h>
2003-10-17 17:25:51 -04:00
#include <string.h>
Initial revision
1999-07-27 17:50:26 -04:00
avoid warnings
2004-09-09 14:06:14 -04:00
#ifdef HAVE_UNISTD_H
# include <unistd.h>
#endif
introduce libport.h prototypes to reduce warnings
2009-01-22 15:53:07 -05:00
#ifdef NEED_LIBPORT
# include "libport.h"
#endif
Initial revision
1999-07-27 17:50:26 -04:00
#include "tiffio.h"
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
/*
* Revision history
* tools/tiff2ps.c:Fix bug in auto rotate option code. Once a rotation angle was set by the auto rotate check, it was retained for all pages that followed instead of being retested for each page. Patch by Richard Nolde.
2013-01-25 21:24:05 -05:00
* 2013-Jan-21
* Richard Nolde: Fix bug in auto rotate option code. Once a
* rotation angle was set by the auto rotate check, it was
* retained for all pages that followed instead of being
* retested for each page.
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*
* 2010-Sep-17
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* Richard Nolde: Reinstate code from Feb 2009 that never got
* accepted into CVS with major modifications to handle -H and -W
* options. Replaced original PlaceImage function with several
* new functions that make support for multiple output pages
* from a single image easier to understand. Added additional
* warning messages for incompatible command line options.
* Add new command line options to specify PageOrientation
* Document Structuring Comment for landscape or portrait
* and code to determine the values from ouput width and height
* if not specified on the command line.
* Add new command line option to specify document creator
* as an alterntive to the string "tiff2ps" following model
Fix some typos Most of them were found by codespell. Signed-off-by: Stefan Weil <sw@weilnetz.de>
2018-02-24 15:47:52 -05:00
* of patch submitted by Thomas Jarosch for specifying a
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* document title which is also supported now.
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
*
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* 2009-Feb-11
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
* Richard Nolde: Added support for rotations of 90, 180, 270
* and auto using -r <90|180|270|auto>. Auto picks the best
* fit for the image on the specified paper size (eg portrait
* or landscape) if -h or -w is specified. Rotation is in
* degrees counterclockwise since that is how Postscript does
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* it. The auto opption rotates the image 90 degrees ccw to
Fix some typos Most of them were found by codespell. Signed-off-by: Stefan Weil <sw@weilnetz.de>
2018-02-24 15:47:52 -05:00
* produce landscape if that is a better fit than portrait.
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
*
* Cleaned up code in TIFF2PS and broke into smaller functions
* to simplify rotations.
*
* Identified incompatible options and returned errors, eg
* -i for imagemask operator is only available for Level2 or
* Level3 Postscript in the current implmentation since there
* is a difference in the way the operands are called for Level1
* and there is no function to provide the Level1 version.
* -H was not handled properly if -h and/or -w were specified.
* It should only clip the masked images if the scaled image
* exceeds the maxPageHeight specified with -H.
*
* New design allows for all of the following combinations:
* Conversion of TIFF to Postscript with optional rotations
* of 90, 180, 270, or auto degrees counterclockwise
* Conversion of TIFF to Postscript with entire image scaled
* to maximum of values spedified with -h or -w while
* maintaining aspect ratio. Same rotations apply.
* Conversion of TIFF to Postscript with clipping of output
* viewport to height specified with -H, producing multiple
* pages at this height and original width as needed.
* Same rotations apply.
* Conversion of TIFF to Postscript with image scaled to
* maximum specified by -h and -w and the resulting scaled
* image is presented in an output viewport clipped by -H height.
* The same rotations apply.
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
*
* Added maxPageWidth option using -W flag. MaxPageHeight and
* MaxPageWidth are mutually exclusive since the aspect ratio
* cannot be maintained if you set both.
* Rewrote PlaceImage to allow maxPageHeight and maxPageWidth
* options to work with values smaller or larger than the
* physical paper size and still preserve the aspect ratio.
* This is accomplished by creating multiple pages across
* as well as down if need be.
*
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
* 2001-Mar-21
* I (Bruce A. Mallett) added this revision history comment ;)
*
* Fixed PS_Lvl2page() code which outputs non-ASCII85 raw
* data. Moved test for when to output a line break to
* *after* the output of a character. This just serves
* to fix an eye-nuisance where the first line of raw
* data was one character shorter than subsequent lines.
*
* Added an experimental ASCII85 encoder which can be used
* only when there is a single buffer of bytes to be encoded.
* This version is much faster at encoding a straight-line
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* buffer of data because it can avoid a lot of the loop
* overhead of the byte-by-byte version. To use this version
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
* you need to define EXP_ASCII85ENCODER (experimental ...).
*
* Added bug fix given by Michael Schmidt to PS_Lvl2page()
* in which an end-of-data marker ('>') was not being output
* when producing non-ASCII85 encoded PostScript Level 2
* data.
*
* Fixed PS_Lvl2colorspace() so that it no longer assumes that
* a TIFF having more than 2 planes is a CMYK. This routine
* no longer looks at the samples per pixel but instead looks
* at the "photometric" value. This change allows support of
* CMYK TIFFs.
*
* Modified the PostScript L2 imaging loop so as to test if
* the input stream is still open before attempting to do a
* flushfile on it. This was done because some RIPs close
* the stream after doing the image operation.
*
* Got rid of the realloc() being done inside a loop in the
* PSRawDataBW() routine. The code now walks through the
* byte-size array outside the loop to determine the largest
* size memory block that will be needed.
*
* Added "-m" switch to ask tiff2ps to, where possible, use the
* "imagemask" operator instead of the "image" operator.
*
* Added the "-i #" switch to allow interpolation to be disabled.
*
* Unrolled a loop or two to improve performance.
*/
/*
* Define EXP_ASCII85ENCODER if you want to use an experimental
* version of the ASCII85 encoding routine. The advantage of
* using this routine is that tiff2ps will convert to ASCII85
* encoding at between 3 and 4 times the speed as compared to
* using the old (non-experimental) encoder. The disadvantage
* is that you will be using a new (and unproven) encoding
* routine. So user beware, you have been warned!
*/
#define EXP_ASCII85ENCODER
Initial revision
1999-07-27 17:50:26 -04:00
/*
* NB: this code assumes uint32 works with printf's %l[ud].
*/
#ifndef TRUE
#define TRUE 1
#define FALSE 0
#endif
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
#define DEFAULT_MAX_MALLOC (256 * 1024 * 1024)
/* malloc size limit (in bytes)
* disabled when set to 0 */
static tmsize_t maxMalloc = DEFAULT_MAX_MALLOC;
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
int ascii85 = FALSE; /* use ASCII85 encoding */
Initial revision
1999-07-27 17:50:26 -04:00
int interpolate = TRUE; /* interpolate level2 image */
int level2 = FALSE; /* generate PostScript level 2 */
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
int level3 = FALSE; /* generate PostScript level 3 */
Initial revision
1999-07-27 17:50:26 -04:00
int printAll = FALSE; /* print all images in file */
int generateEPSF = TRUE; /* generate Encapsulated PostScript */
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
int PSduplex = FALSE; /* enable duplex printing */
Initial revision
1999-07-27 17:50:26 -04:00
int PStumble = FALSE; /* enable top edge binding */
int PSavoiddeadzone = TRUE; /* enable avoiding printer deadzone */
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
double maxPageHeight = 0; /* maximum height to select from image and print per page */
double maxPageWidth = 0; /* maximum width to select from image and print per page */
Avoid warnings.
2004-06-04 10:16:14 -04:00
double splitOverlap = 0; /* amount for split pages to overlag */
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
int rotation = 0; /* optional value for rotation angle */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
int auto_rotate = 0; /* rotate image for best fit on the page */
char *filename = NULL; /* input filename */
char *title = NULL; /* optional document title string */
char *creator = NULL; /* optional document creator string */
char pageOrientation[12]; /* set optional PageOrientation DSC to Landscape or Portrait */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
int useImagemask = FALSE; /* Use imagemask instead of image operator */
Calculate margins taking in account resolution unit (Richard Nolde).
2005-02-06 12:58:01 -05:00
uint16 res_unit = 0; /* Resolution units: 2 - inches, 3 - cm */
Initial revision
1999-07-27 17:50:26 -04:00
/*
* ASCII85 Encoding Support.
*/
unsigned char ascii85buf[10];
int ascii85count;
int ascii85breaklen;
Avoid warnings.
2004-06-04 10:16:14 -04:00
int TIFF2PS(FILE*, TIFF*, double, double, double, double, int);
Initial revision
1999-07-27 17:50:26 -04:00
void PSpage(FILE*, TIFF*, uint32, uint32);
void PSColorContigPreamble(FILE*, uint32, uint32, int);
void PSColorSeparatePreamble(FILE*, uint32, uint32, int);
void PSDataColorContig(FILE*, TIFF*, uint32, uint32, int);
void PSDataColorSeparate(FILE*, TIFF*, uint32, uint32, int);
void PSDataPalette(FILE*, TIFF*, uint32, uint32);
void PSDataBW(FILE*, TIFF*, uint32, uint32);
void PSRawDataBW(FILE*, TIFF*, uint32, uint32);
void Ascii85Init(void);
void Ascii85Put(unsigned char code, FILE* fd);
void Ascii85Flush(FILE* fd);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
void PSHead(FILE*, double, double, double, double);
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
void PSTail(FILE*, int);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
int psStart(FILE *, int, int, int *, double *, double, double, double,
double, double, double, double, double, double, double);
int psPageSize(FILE *, int, double, double, double, double, double, double);
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
int psRotateImage(FILE *, int, double, double, double, double);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
int psMaskImage(FILE *, TIFF *, int, int, int *, double, double,
double, double, double, double, double, double, double);
int psScaleImage(FILE *, double, int, int, double, double, double, double,
double, double);
int get_viewport (double, double, double, double, double *, double *, int);
int exportMaskedImage(FILE *, double, double, double, double, int, int,
double, double, double, int, int);
Initial revision
1999-07-27 17:50:26 -04:00
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER)
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t Ascii85EncodeBlock( uint8 * ascii85_p, unsigned f_eod, const uint8 * raw_p, tsize_t raw_l );
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
static void usage(int);
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
/**
* This custom malloc function enforce a maximum allocation size
*/
static void* limitMalloc(tmsize_t s)
{
if (maxMalloc && (s > maxMalloc)) {
fprintf(stderr, "MemoryLimitError: allocation of " TIFF_UINT64_FORMAT " bytes is forbidden. Limit is " TIFF_UINT64_FORMAT ".\n",
(uint64)s, (uint64)maxMalloc);
fprintf(stderr, " use -M option to change limit.\n");
return NULL;
}
return _TIFFmalloc(s);
}
Initial revision
1999-07-27 17:50:26 -04:00
int
main(int argc, char* argv[])
{
Avoid warnings.
2004-06-05 04:13:06 -04:00
int dirnum = -1, c, np = 0;
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
int centered = 0;
Avoid warnings.
2004-06-04 10:16:14 -04:00
double bottommargin = 0;
double leftmargin = 0;
double pageWidth = 0;
double pageHeight = 0;
Initial revision
1999-07-27 17:50:26 -04:00
uint32 diroff = 0;
* libtiff/tif_config.vc.h: Make adjustments to match the new definitions that configure produces, including for WIN64. Still needs to be tested. 'lld' is not assured by the run-time DLLs and so GCC warns. Add TIFF_SIZE_T and TIFF_SIZE_FORMAT to provide a type definition and printf format specifier to deal with printing values of 'size_t' type. In particular, this was necessary for WIN64. Added a configure test for if the system headers provide 'optarg' (normal case) and block out the many explicit 'extern' statements in the utilities. This was found to be necessary under Windows when getopt is in a DLL and the symbols are already imported with dllimport via standard header files.
2015-06-20 21:09:09 -04:00
#if !HAVE_DECL_OPTARG
Initial revision
1999-07-27 17:50:26 -04:00
extern char *optarg;
extern int optind;
* libtiff/tif_config.vc.h: Make adjustments to match the new definitions that configure produces, including for WIN64. Still needs to be tested. 'lld' is not assured by the run-time DLLs and so GCC warns. Add TIFF_SIZE_T and TIFF_SIZE_FORMAT to provide a type definition and printf format specifier to deal with printing values of 'size_t' type. In particular, this was necessary for WIN64. Added a configure test for if the system headers provide 'optarg' (normal case) and block out the many explicit 'extern' statements in the utilities. This was found to be necessary under Windows when getopt is in a DLL and the symbols are already imported with dllimport via standard header files.
2015-06-20 21:09:09 -04:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
FILE* output = stdout;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
pageOrientation[0] = '\0';
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
while ((c = getopt(argc, argv, "b:d:h:H:W:L:M:i:w:l:o:O:P:C:r:t:acemxyzps1238DT")) != -1)
Initial revision
1999-07-27 17:50:26 -04:00
switch (c) {
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
case 'M':
maxMalloc = (tmsize_t)strtoul(optarg, NULL, 0) << 20;
break;
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
case 'b':
bottommargin = atof(optarg);
break;
case 'c':
centered = 1;
break;
Sync tool usage printouts and man pages with reality.
2013-03-05 22:35:09 -05:00
case 'C':
creator = optarg;
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 'd': /* without -a, this only processes one image at this IFD */
Initial revision
1999-07-27 17:50:26 -04:00
dirnum = atoi(optarg);
break;
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
case 'D':
Initial revision
1999-07-27 17:50:26 -04:00
PSduplex = TRUE;
break;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
case 'i':
interpolate = atoi(optarg) ? TRUE:FALSE;
break;
Initial revision
1999-07-27 17:50:26 -04:00
case 'T':
PStumble = TRUE;
break;
case 'e':
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
PSavoiddeadzone = FALSE;
Initial revision
1999-07-27 17:50:26 -04:00
generateEPSF = TRUE;
break;
case 'h':
pageHeight = atof(optarg);
break;
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
case 'H':
maxPageHeight = atof(optarg);
break;
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
case 'W':
maxPageWidth = atof(optarg);
break;
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
case 'L':
splitOverlap = atof(optarg);
break;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
case 'm':
useImagemask = TRUE;
break;
Initial revision
1999-07-27 17:50:26 -04:00
case 'o':
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
switch (optarg[0])
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9': diroff = (uint32) strtoul(optarg, NULL, 0);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
break;
default: TIFFError ("-o", "Offset must be a numeric value.");
exit (1);
}
Initial revision
1999-07-27 17:50:26 -04:00
break;
case 'O': /* XXX too bad -o is already taken */
output = fopen(optarg, "w");
if (output == NULL) {
fprintf(stderr,
"%s: %s: Cannot open output file.\n",
argv[0], optarg);
exit(-2);
}
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 'P':
switch (optarg[0])
{
case 'l':
case 'L': strcpy (pageOrientation, "Landscape");
break;
case 'p':
case 'P': strcpy (pageOrientation, "Portrait");
break;
default: TIFFError ("-P", "Page orientation must be Landscape or Portrait");
exit (-1);
}
break;
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
case 'l':
leftmargin = atof(optarg);
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 'a': /* removed fall through to generate warning below, R Nolde 09-01-2010 */
Initial revision
1999-07-27 17:50:26 -04:00
printAll = TRUE;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
break;
Initial revision
1999-07-27 17:50:26 -04:00
case 'p':
generateEPSF = FALSE;
break;
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
case 'r':
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
if (strcmp (optarg, "auto") == 0)
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
{
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
rotation = 0;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
auto_rotate = TRUE;
}
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
else
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
{
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
rotation = atoi(optarg);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
auto_rotate = FALSE;
}
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
switch (rotation)
{
case 0:
case 90:
case 180:
case 270:
break;
default:
fprintf (stderr, "Rotation angle must be 90, 180, 270 (degrees ccw) or auto\n");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
exit (-1);
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
}
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
break;
Initial revision
1999-07-27 17:50:26 -04:00
case 's':
printAll = FALSE;
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 't':
title = optarg;
break;
Initial revision
1999-07-27 17:50:26 -04:00
case 'w':
pageWidth = atof(optarg);
break;
case 'z':
PSavoiddeadzone = FALSE;
break;
case '1':
level2 = FALSE;
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
level3 = FALSE;
Initial revision
1999-07-27 17:50:26 -04:00
ascii85 = FALSE;
break;
case '2':
level2 = TRUE;
ascii85 = TRUE; /* default to yes */
break;
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
case '3':
level3 = TRUE;
ascii85 = TRUE; /* default to yes */
break;
Initial revision
1999-07-27 17:50:26 -04:00
case '8':
ascii85 = FALSE;
break;
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
case 'x':
New commandline switches to override resolution units obtained from the input file. Closes http://bugzilla.remotesensing.org/show_bug.cgi?id=131
2002-05-10 06:45:43 -04:00
res_unit = RESUNIT_CENTIMETER;
break;
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
case 'y':
New commandline switches to override resolution units obtained from the input file. Closes http://bugzilla.remotesensing.org/show_bug.cgi?id=131
2002-05-10 06:45:43 -04:00
res_unit = RESUNIT_INCH;
break;
Initial revision
1999-07-27 17:50:26 -04:00
case '?':
usage(-1);
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (useImagemask == TRUE)
{
if ((level2 == FALSE) && (level3 == FALSE))
{
TIFFError ("-m "," imagemask operator requres Postscript Level2 or Level3");
exit (1);
}
}
if (pageWidth && (maxPageWidth > pageWidth))
{
TIFFError ("-W", "Max viewport width cannot exceed page width");
exit (1);
}
/* auto rotate requires a specified page width and height */
if (auto_rotate == TRUE)
{
* tools/tiff2ps.c:Fix bug in auto rotate option code. Once a rotation angle was set by the auto rotate check, it was retained for all pages that followed instead of being retested for each page. Patch by Richard Nolde.
2013-01-25 21:24:05 -05:00
/*
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if ((pageWidth == 0) || (pageHeight == 0))
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
TIFFWarning ("-r auto", " requires page height and width specified with -h and -w");
* tools/tiff2ps.c:Fix bug in auto rotate option code. Once a rotation angle was set by the auto rotate check, it was retained for all pages that followed instead of being retested for each page. Patch by Richard Nolde.
2013-01-25 21:24:05 -05:00
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if ((maxPageWidth > 0) || (maxPageHeight > 0))
{
TIFFError ("-r auto", " is incompatible with maximum page width/height specified by -H or -W");
exit (1);
}
}
if ((maxPageWidth > 0) && (maxPageHeight > 0))
{
TIFFError ("-H and -W", " Use only one of -H or -W to define a viewport");
exit (1);
}
if ((generateEPSF == TRUE) && (printAll == TRUE))
{
Fix spelling typo in tiff2ps.
2011-04-09 12:53:08 -04:00
TIFFError(" -e and -a", "Warning: Cannot generate Encapsulated Postscript for multiple images");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
generateEPSF = FALSE;
}
if ((generateEPSF == TRUE) && (PSduplex == TRUE))
{
Fix spelling typo in tiff2ps.
2011-04-09 12:53:08 -04:00
TIFFError(" -e and -D", "Warning: Encapsulated Postscript does not support Duplex option");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
PSduplex = FALSE;
}
if ((generateEPSF == TRUE) && (PStumble == TRUE))
{
Fix spelling typo in tiff2ps.
2011-04-09 12:53:08 -04:00
TIFFError(" -e and -T", "Warning: Encapsulated Postscript does not support Top Edge Binding option");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
PStumble = FALSE;
}
if ((generateEPSF == TRUE) && (PSavoiddeadzone == TRUE))
PSavoiddeadzone = FALSE;
Initial revision
1999-07-27 17:50:26 -04:00
for (; argc - optind > 0; optind++) {
TIFF* tif = TIFFOpen(filename = argv[optind], "r");
if (tif != NULL) {
Avoid warnings.
2004-06-05 04:13:06 -04:00
if (dirnum != -1
&& !TIFFSetDirectory(tif, (tdir_t)dirnum))
* libtiff/tif_dirread.c: fix memory leak in non DEFER_STRILE_LOAD mode (ie default) when there is both a StripOffsets and TileOffsets tag, or a StripByteCounts and TileByteCounts Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2689 * tools/tiff2ps.c: call TIFFClose() in error code paths.
2017-04-27 11:46:22 -04:00
{
TIFFClose(tif);
Initial revision
1999-07-27 17:50:26 -04:00
return (-1);
* libtiff/tif_dirread.c: fix memory leak in non DEFER_STRILE_LOAD mode (ie default) when there is both a StripOffsets and TileOffsets tag, or a StripByteCounts and TileByteCounts Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2689 * tools/tiff2ps.c: call TIFFClose() in error code paths.
2017-04-27 11:46:22 -04:00
}
Initial revision
1999-07-27 17:50:26 -04:00
else if (diroff != 0 &&
!TIFFSetSubDirectory(tif, diroff))
* libtiff/tif_dirread.c: fix memory leak in non DEFER_STRILE_LOAD mode (ie default) when there is both a StripOffsets and TileOffsets tag, or a StripByteCounts and TileByteCounts Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2689 * tools/tiff2ps.c: call TIFFClose() in error code paths.
2017-04-27 11:46:22 -04:00
{
TIFFClose(tif);
Initial revision
1999-07-27 17:50:26 -04:00
return (-1);
* libtiff/tif_dirread.c: fix memory leak in non DEFER_STRILE_LOAD mode (ie default) when there is both a StripOffsets and TileOffsets tag, or a StripByteCounts and TileByteCounts Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2689 * tools/tiff2ps.c: call TIFFClose() in error code paths.
2017-04-27 11:46:22 -04:00
}
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
np = TIFF2PS(output, tif, pageWidth, pageHeight,
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
leftmargin, bottommargin, centered);
if (np < 0)
{
TIFFError("Error", "Unable to process %s", filename);
}
Initial revision
1999-07-27 17:50:26 -04:00
TIFFClose(tif);
}
}
if (np)
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
PSTail(output, np);
Initial revision
1999-07-27 17:50:26 -04:00
else
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
usage(-1);
Initial revision
1999-07-27 17:50:26 -04:00
if (output != stdout)
fclose(output);
return (0);
}
static uint16 samplesperpixel;
static uint16 bitspersample;
static uint16 planarconfiguration;
static uint16 photometric;
static uint16 compression;
static uint16 extrasamples;
static int alpha;
static int
checkImage(TIFF* tif)
{
switch (photometric) {
case PHOTOMETRIC_YCBCR:
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
if ((compression == COMPRESSION_JPEG || compression == COMPRESSION_OJPEG)
&& planarconfiguration == PLANARCONFIG_CONTIG) {
Initial revision
1999-07-27 17:50:26 -04:00
/* can rely on libjpeg to convert to RGB */
TIFFSetField(tif, TIFFTAG_JPEGCOLORMODE,
JPEGCOLORMODE_RGB);
photometric = PHOTOMETRIC_RGB;
} else {
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
if (level2 || level3)
Initial revision
1999-07-27 17:50:26 -04:00
break;
TIFFError(filename, "Can not handle image with %s",
"PhotometricInterpretation=YCbCr");
return (0);
}
Fix some typos Most of them were found by codespell. Signed-off-by: Stefan Weil <sw@weilnetz.de>
2018-02-24 15:47:52 -05:00
/* fall through... */
Initial revision
1999-07-27 17:50:26 -04:00
case PHOTOMETRIC_RGB:
if (alpha && bitspersample != 8) {
TIFFError(filename,
"Can not handle %d-bit/sample RGB image with alpha",
bitspersample);
return (0);
}
Fix some typos Most of them were found by codespell. Signed-off-by: Stefan Weil <sw@weilnetz.de>
2018-02-24 15:47:52 -05:00
/* fall through... */
Initial revision
1999-07-27 17:50:26 -04:00
case PHOTOMETRIC_SEPARATED:
case PHOTOMETRIC_PALETTE:
case PHOTOMETRIC_MINISBLACK:
case PHOTOMETRIC_MINISWHITE:
break;
case PHOTOMETRIC_LOGL:
case PHOTOMETRIC_LOGLUV:
if (compression != COMPRESSION_SGILOG &&
compression != COMPRESSION_SGILOG24) {
TIFFError(filename,
"Can not handle %s data with compression other than SGILog",
(photometric == PHOTOMETRIC_LOGL) ?
"LogL" : "LogLuv"
);
return (0);
}
/* rely on library to convert to RGB/greyscale */
TIFFSetField(tif, TIFFTAG_SGILOGDATAFMT, SGILOGDATAFMT_8BIT);
photometric = (photometric == PHOTOMETRIC_LOGL) ?
PHOTOMETRIC_MINISBLACK : PHOTOMETRIC_RGB;
bitspersample = 8;
break;
case PHOTOMETRIC_CIELAB:
Fix some typos Most of them were found by codespell. Signed-off-by: Stefan Weil <sw@weilnetz.de>
2018-02-24 15:47:52 -05:00
/* fall through... */
Initial revision
1999-07-27 17:50:26 -04:00
default:
TIFFError(filename,
"Can not handle image with PhotometricInterpretation=%d",
photometric);
return (0);
}
switch (bitspersample) {
case 1: case 2:
case 4: case 8:
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
case 16:
Initial revision
1999-07-27 17:50:26 -04:00
break;
default:
TIFFError(filename, "Can not handle %d-bit/sample image",
bitspersample);
return (0);
}
if (planarconfiguration == PLANARCONFIG_SEPARATE && extrasamples > 0)
TIFFWarning(filename, "Ignoring extra samples");
return (1);
}
Fixed wrong variable data type when read Resolution tags (Peter Fales).
2004-10-28 13:56:46 -04:00
#define PS_UNIT_SIZE 72.0F
Initial revision
1999-07-27 17:50:26 -04:00
#define PSUNITS(npix,res) ((npix) * (PS_UNIT_SIZE / (res)))
static char RGBcolorimage[] = "\
/bwproc {\n\
rgbproc\n\
dup length 3 idiv string 0 3 0\n\
5 -1 roll {\n\
add 2 1 roll 1 sub dup 0 eq {\n\
pop 3 idiv\n\
3 -1 roll\n\
dup 4 -1 roll\n\
dup 3 1 roll\n\
5 -1 roll put\n\
1 add 3 0\n\
} { 2 1 roll } ifelse\n\
} forall\n\
pop pop pop\n\
} def\n\
/colorimage where {pop} {\n\
/colorimage {pop pop /rgbproc exch def {bwproc} image} bind def\n\
} ifelse\n\
";
/*
* Adobe Photoshop requires a comment line of the form:
*
* %ImageData: <cols> <rows> <depth> <main channels> <pad channels>
* <block size> <1 for binary|2 for hex> "data start"
*
* It is claimed to be part of some future revision of the EPS spec.
*/
static void
PhotoshopBanner(FILE* fd, uint32 w, uint32 h, int bs, int nc, char* startline)
{
fprintf(fd, "%%ImageData: %ld %ld %d %d 0 %d 2 \"",
(long) w, (long) h, bitspersample, nc, bs);
fprintf(fd, startline, nc);
fprintf(fd, "\"\n");
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
/* Convert pixel width and height pw, ph, to points pprw, pprh
* using image resolution and resolution units from TIFF tags.
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
* pw : image width in pixels
* ph : image height in pixels
* pprw : image width in PS units (72 dpi)
* pprh : image height in PS units (72 dpi)
*/
Initial revision
1999-07-27 17:50:26 -04:00
static void
Avoid warnings.
2004-06-04 10:16:14 -04:00
setupPageState(TIFF* tif, uint32* pw, uint32* ph, double* pprw, double* pprh)
Initial revision
1999-07-27 17:50:26 -04:00
{
Fixed wrong variable data type when read Resolution tags (Peter Fales).
2004-10-28 13:56:46 -04:00
float xres = 0.0F, yres = 0.0F;
Initial revision
1999-07-27 17:50:26 -04:00
TIFFGetField(tif, TIFFTAG_IMAGEWIDTH, pw);
TIFFGetField(tif, TIFFTAG_IMAGELENGTH, ph);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (res_unit == 0) /* Not specified as command line option */
if (!TIFFGetFieldDefaulted(tif, TIFFTAG_RESOLUTIONUNIT, &res_unit))
res_unit = RESUNIT_INCH;
Initial revision
1999-07-27 17:50:26 -04:00
/*
* Calculate printable area.
*/
Avoid zero division in setupPageState() function; properly initialize array in PSDataBW().
2004-08-25 09:14:18 -04:00
if (!TIFFGetField(tif, TIFFTAG_XRESOLUTION, &xres)
|| fabs(xres) < 0.0000001)
Initial revision
1999-07-27 17:50:26 -04:00
xres = PS_UNIT_SIZE;
Avoid zero division in setupPageState() function; properly initialize array in PSDataBW().
2004-08-25 09:14:18 -04:00
if (!TIFFGetField(tif, TIFFTAG_YRESOLUTION, &yres)
|| fabs(yres) < 0.0000001)
Initial revision
1999-07-27 17:50:26 -04:00
yres = PS_UNIT_SIZE;
switch (res_unit) {
case RESUNIT_CENTIMETER:
Fixed wrong variable data type when read Resolution tags (Peter Fales).
2004-10-28 13:56:46 -04:00
xres *= 2.54F, yres *= 2.54F;
Initial revision
1999-07-27 17:50:26 -04:00
break;
*** empty log message ***
2002-06-02 07:41:23 -04:00
case RESUNIT_INCH:
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case RESUNIT_NONE: /* Subsequent code assumes we have converted to inches! */
res_unit = RESUNIT_INCH;
break;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
default: /* Last ditch guess for unspecified RESUNIT case
* check that the resolution is not inches before scaling it.
* Moved to end of function with additional check, RJN, 08-31-2010
* if (xres != PS_UNIT_SIZE || yres != PS_UNIT_SIZE)
* xres *= PS_UNIT_SIZE, yres *= PS_UNIT_SIZE;
*/
Initial revision
1999-07-27 17:50:26 -04:00
break;
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
/* This is a hack to deal with images that have no meaningful Resolution Size
* but may have x and/or y resolutions of 1 pixel per undefined unit.
*/
if ((xres > 1.0) && (xres != PS_UNIT_SIZE))
*pprw = PSUNITS(*pw, xres);
else
*pprw = PSUNITS(*pw, PS_UNIT_SIZE);
if ((yres > 1.0) && (yres != PS_UNIT_SIZE))
*pprh = PSUNITS(*ph, yres);
else
*pprh = PSUNITS(*ph, PS_UNIT_SIZE);
Initial revision
1999-07-27 17:50:26 -04:00
}
static int
isCCITTCompression(TIFF* tif)
{
uint16 compress;
TIFFGetField(tif, TIFFTAG_COMPRESSION, &compress);
return (compress == COMPRESSION_CCITTFAX3 ||
compress == COMPRESSION_CCITTFAX4 ||
compress == COMPRESSION_CCITTRLE ||
compress == COMPRESSION_CCITTRLEW);
}
static tsize_t tf_bytesperrow;
static tsize_t ps_bytesperrow;
tiff2ps: fix use of wrong data type that caused issues (/Height being written as 0) on 64-bit big endian platforms
2019-08-18 04:52:45 -04:00
static uint32 tf_rowsperstrip;
static uint32 tf_numberstrips;
Initial revision
1999-07-27 17:50:26 -04:00
static char *hex = "0123456789abcdef";
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
/*
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
* Pagewidth and pageheight are the output size in points,
* may refer to values specified with -h and -w, or to
* values read from the image if neither -h nor -w are used.
* Imagewidth and imageheight are image size in points.
* Ximages and Yimages are number of pages across and down.
* Only one of maxPageHeight or maxPageWidth can be used.
* These are global variables unfortunately.
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
int get_subimage_count(double pagewidth, double pageheight,
double imagewidth, double imageheight,
int *ximages, int *yimages,
int rotation, double scale)
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
{
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
int pages = 1;
double splitheight = 0; /* Requested Max Height in points */
double splitwidth = 0; /* Requested Max Width in points */
double overlap = 0; /* Repeated edge width in points */
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
splitheight = maxPageHeight * PS_UNIT_SIZE;
splitwidth = maxPageWidth * PS_UNIT_SIZE;
overlap = splitOverlap * PS_UNIT_SIZE;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
pagewidth *= PS_UNIT_SIZE;
pageheight *= PS_UNIT_SIZE;
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if ((imagewidth < 1.0) || (imageheight < 1.0))
{
TIFFError("get_subimage_count", "Invalid image width or height");
return (0);
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
switch (rotation)
{
case 0:
case 180: if (splitheight > 0) /* -H maxPageHeight */
{
if (imageheight > splitheight) /* More than one vertical image segment */
{
if (pagewidth)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imagewidth) / (pagewidth - overlap));
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*ximages = 1;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imageheight) / (splitheight - overlap)); /* Max vert pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else
{
if (pagewidth)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imagewidth) / (pagewidth - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*ximages = 1;
*yimages = 1; /* Max vert pages needed */
}
}
else
{
if (splitwidth > 0) /* -W maxPageWidth */
{
if (imagewidth >splitwidth)
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imagewidth) / (splitwidth - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (pageheight)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imageheight) / (pageheight - overlap)); /* Max vert pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*yimages = 1;
}
else
{
*ximages = 1; /* Max vert pages needed */
if (pageheight)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imageheight) / (pageheight - overlap)); /* Max vert pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*yimages = 1;
}
}
else
{
*ximages = 1;
*yimages = 1;
}
}
break;
case 90:
case 270: if (splitheight > 0) /* -H maxPageHeight */
{
if (imagewidth > splitheight) /* More than one vertical image segment */
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imagewidth) / (splitheight - overlap)); /* Max vert pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (pagewidth)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imageheight) / (pagewidth - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*ximages = 1;
}
else
{
*yimages = 1; /* Max vert pages needed */
if (pagewidth)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imageheight) / (pagewidth - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*ximages = 1;
}
}
else
{
if (splitwidth > 0) /* -W maxPageWidth */
{
if (imageheight > splitwidth)
{
if (pageheight)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imagewidth) / (pageheight - overlap)); /* Max vert pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*yimages = 1;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*ximages = (int)ceil((scale * imageheight) / (splitwidth - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else
{
if (pageheight)
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
*yimages = (int)ceil((scale * imagewidth) / (pageheight - overlap)); /* Max horz pages needed */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
*yimages = 1;
*ximages = 1; /* Max vert pages needed */
}
}
else
{
*ximages = 1;
*yimages = 1;
}
}
break;
default: *ximages = 1;
*yimages = 1;
}
pages = (*ximages) * (*yimages);
return (pages);
}
/* New version of PlaceImage that handles only the translation and rotation
* for a single output page.
*/
int exportMaskedImage(FILE *fp, double pagewidth, double pageheight,
double imagewidth, double imageheight,
int row, int column,
double left_offset, double bott_offset,
double scale, int center, int rotation)
{
double xtran = 0.0;
double ytran = 0.0;
double xscale = 1.0;
double yscale = 1.0;
double splitheight = 0; /* Requested Max Height in points */
double splitwidth = 0; /* Requested Max Width in points */
double overlap = 0; /* Repeated edge width in points */
double subimage_height = 0.0;
splitheight = maxPageHeight * PS_UNIT_SIZE;
splitwidth = maxPageWidth * PS_UNIT_SIZE;
overlap = splitOverlap * PS_UNIT_SIZE;
xscale = scale * imagewidth;
yscale = scale * imageheight;
if ((xscale < 0.0) || (yscale < 0.0))
{
TIFFError("exportMaskedImage", "Invalid parameters.");
return (-1);
}
Initial revision
1999-07-27 17:50:26 -04:00
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
/* If images are cropped to a vewport with -H or -W, the output pages are shifted to
* the top of each output page rather than the Postscript default lower edge.
*/
switch (rotation)
{
case 0:
case 180: if (splitheight > 0) /* -H maxPageHeight */
{
if (splitheight < imageheight) /* More than one vertical image segments */
{
From Eduardo Robles Elvira: 1. libtiffcrop-fix.patch fixes a small problem in tiffcrop, it seems it was incorrectly using TIFFSetField instead of CopyField. And in libtiff-correct-fax-scaling.patch we have some other changes: 2. I had to remove a check in main() that didn't allow maxPageWidth to be bigger than pageWidth. 3. [ Omitted due to question on universality ] 4. the pagewidth variable was being set as the maxpagewidth instead, which made all the calculations bad. This made sense when the check in point 2 was in place, but not anymore. I've modified it so that pagewidth is set with the specified pagewidth when maxpagewidth is bigger. 5. The remaining lines of the patch - in exportMaskedImage() - basically fix the scaling.
2015-06-13 01:49:18 -04:00
/* Intra2net: Keep correct apspect ratio */
xscale = (imagewidth + overlap) * (pageheight / splitheight) * scale;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
xtran = -1.0 * column * (pagewidth - overlap);
subimage_height = imageheight - ((splitheight - overlap) * row);
ytran = pageheight - subimage_height * (pageheight / splitheight);
}
else /* Only one page in vertical direction */
{
xtran = -1.0 * column * (pagewidth - overlap);
ytran = splitheight - imageheight;
}
}
else
{
if (splitwidth > 0) /* maxPageWidth */
{
if (splitwidth < imagewidth)
{
xtran = -1.0 * column * splitwidth;
ytran = -1.0 * row * (pageheight - overlap);
}
else /* Only one page in horizontal direction */
{
ytran = -1.0 * row * (pageheight - overlap);
xtran = 0;
}
}
else /* Simple case, no splitting */
{
ytran = pageheight - imageheight;
xtran = 0;
}
}
From Eduardo Robles Elvira: 1. libtiffcrop-fix.patch fixes a small problem in tiffcrop, it seems it was incorrectly using TIFFSetField instead of CopyField. And in libtiff-correct-fax-scaling.patch we have some other changes: 2. I had to remove a check in main() that didn't allow maxPageWidth to be bigger than pageWidth. 3. [ Omitted due to question on universality ] 4. the pagewidth variable was being set as the maxpagewidth instead, which made all the calculations bad. This made sense when the check in point 2 was in place, but not anymore. I've modified it so that pagewidth is set with the specified pagewidth when maxpagewidth is bigger. 5. The remaining lines of the patch - in exportMaskedImage() - basically fix the scaling.
2015-06-13 01:49:18 -04:00
if (imagewidth <= pagewidth) {
/* Intra2net: Crop page at the bottom instead of the top (-> output starts at the top).
Only do this in non-page-split mode */
if (imageheight <= splitheight) {
ytran = pageheight - imageheight; /* Note: Will be negative for images longer than page size */
}
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
bott_offset += ytran / (center ? 2 : 1);
left_offset += xtran / (center ? 2 : 1);
break;
case 90:
case 270: if (splitheight > 0) /* -H maxPageHeight */
{
if (splitheight < imagewidth) /* More than one vertical image segments */
{
xtran = -1.0 * column * (pageheight - overlap);
/* Commented code places image at bottom of page instead of top.
ytran = -1.0 * row * splitheight;
*/
if (row == 0)
ytran = -1.0 * (imagewidth - splitheight);
else
ytran = -1.0 * (imagewidth - (splitheight - overlap) * (row + 1));
}
else /* Only one page in vertical direction */
{
xtran = -1.0 * column * (pageheight - overlap);
ytran = splitheight - imagewidth;
}
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
else
{
if (splitwidth > 0) /* maxPageWidth */
{
if (splitwidth < imageheight)
{
xtran = -1.0 * column * splitwidth;
ytran = -1.0 * row * (pagewidth - overlap);
}
else /* Only one page in horizontal direction */
{
ytran = -1.0 * row * (pagewidth - overlap);
xtran = 0;
}
}
else /* Simple case, no splitting */
{
ytran = pageheight - imageheight;
xtran = 0; /* pagewidth - imagewidth; */
}
}
bott_offset += ytran / (center ? 2 : 1);
left_offset += xtran / (center ? 2 : 1);
break;
default: xtran = 0;
ytran = 0;
}
switch (rotation)
{
case 0: fprintf(fp, "%f %f translate\n", left_offset, bott_offset);
fprintf(fp, "%f %f scale\n", xscale, yscale);
break;
case 180: fprintf(fp, "%f %f translate\n", left_offset, bott_offset);
fprintf(fp, "%f %f scale\n1 1 translate 180 rotate\n", xscale, yscale);
break;
case 90: fprintf(fp, "%f %f translate\n", left_offset, bott_offset);
fprintf(fp, "%f %f scale\n1 0 translate 90 rotate\n", yscale, xscale);
break;
case 270: fprintf(fp, "%f %f translate\n", left_offset, bott_offset);
fprintf(fp, "%f %f scale\n0 1 translate 270 rotate\n", yscale, xscale);
break;
default: TIFFError ("exportMaskedImage", "Unsupported rotation angle %d. No rotation", rotation);
fprintf( fp, "%f %f scale\n", xscale, yscale);
break;
}
return (0);
}
/* Rotate an image without scaling or clipping */
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
int psRotateImage (FILE * fd, int rotation, double pswidth, double psheight,
double left_offset, double bottom_offset)
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
if ((left_offset != 0.0) || (bottom_offset != 0))
fprintf (fd, "%f %f translate\n", left_offset, bottom_offset);
/* Exchange width and height for 90/270 rotations */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
switch (rotation)
{
case 0: fprintf (fd, "%f %f scale\n", pswidth, psheight);
break;
case 90: fprintf (fd, "%f %f scale\n1 0 translate 90 rotate\n", psheight, pswidth);
break;
case 180: fprintf (fd, "%f %f scale\n1 1 translate 180 rotate\n", pswidth, psheight);
break;
case 270: fprintf (fd, "%f %f scale\n0 1 translate 270 rotate\n", psheight, pswidth);
break;
default: TIFFError ("psRotateImage", "Unsupported rotation %d.", rotation);
fprintf( fd, "%f %f scale\n", pswidth, psheight);
return (1);
}
return (0);
}
/* Scale and rotate an image to a single output page. */
int psScaleImage(FILE * fd, double scale, int rotation, int center,
double reqwidth, double reqheight, double pswidth, double psheight,
double left_offset, double bottom_offset)
{
double hcenter = 0.0, vcenter = 0.0;
/* Adjust offsets for centering */
if (center)
{
switch (rotation)
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
case 90: vcenter = (reqheight - pswidth * scale) / 2;
hcenter = (reqwidth - psheight * scale) / 2;
fprintf (fd, "%f %f translate\n", hcenter, vcenter);
fprintf (fd, "%f %f scale\n1 0 translate 90 rotate\n", psheight * scale, pswidth * scale);
break;
case 180: hcenter = (reqwidth - pswidth * scale) / 2;
vcenter = (reqheight - psheight * scale) / 2;
fprintf (fd, "%f %f translate\n", hcenter, vcenter);
fprintf (fd, "%f %f scale\n1 1 translate 180 rotate\n", pswidth * scale, psheight * scale);
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 270: vcenter = (reqheight - pswidth * scale) / 2;
hcenter = (reqwidth - psheight * scale) / 2;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
fprintf (fd, "%f %f translate\n", hcenter, vcenter);
fprintf (fd, "%f %f scale\n0 1 translate 270 rotate\n", psheight * scale, pswidth * scale);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
break;
case 0:
default: hcenter = (reqwidth - pswidth * scale) / 2;
vcenter = (reqheight - psheight * scale) / 2;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
fprintf (fd, "%f %f translate\n", hcenter, vcenter);
fprintf (fd, "%f %f scale\n", pswidth * scale, psheight * scale);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
break;
}
}
else /* Not centered */
{
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
switch (rotation)
{
case 0: fprintf (fd, "%f %f translate\n", left_offset ? left_offset : 0.0,
bottom_offset ? bottom_offset : reqheight - (psheight * scale));
fprintf (fd, "%f %f scale\n", pswidth * scale, psheight * scale);
break;
case 90: fprintf (fd, "%f %f translate\n", left_offset ? left_offset : 0.0,
bottom_offset ? bottom_offset : reqheight - (pswidth * scale));
fprintf (fd, "%f %f scale\n1 0 translate 90 rotate\n", psheight * scale, pswidth * scale);
break;
case 180: fprintf (fd, "%f %f translate\n", left_offset ? left_offset : 0.0,
bottom_offset ? bottom_offset : reqheight - (psheight * scale));
fprintf (fd, "%f %f scale\n1 1 translate 180 rotate\n", pswidth * scale, psheight * scale);
break;
case 270: fprintf (fd, "%f %f translate\n", left_offset ? left_offset : 0.0,
bottom_offset ? bottom_offset : reqheight - (pswidth * scale));
fprintf (fd, "%f %f scale\n0 1 translate 270 rotate\n", psheight * scale, pswidth * scale);
break;
default: TIFFError ("psScaleImage", "Unsupported rotation %d", rotation);
fprintf (fd, "%f %f scale\n", pswidth * scale, psheight * scale);
return (1);
}
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
return (0);
}
/* This controls the visible portion of the page which is displayed.
* N.B. Setting maxPageHeight no longer sets pageheight if not set explicitly
*/
int psPageSize (FILE * fd, int rotation, double pgwidth, double pgheight,
double reqwidth, double reqheight, double pswidth, double psheight)
{
double xscale = 1.0, yscale = 1.0, scale = 1.0;
double splitheight;
double splitwidth;
double new_width;
double new_height;
splitheight = maxPageHeight * PS_UNIT_SIZE;
splitwidth = maxPageWidth * PS_UNIT_SIZE;
switch (rotation)
{
case 0:
case 180: if ((splitheight > 0) || (splitwidth > 0))
{
if (pgwidth != 0 || pgheight != 0)
{
xscale = reqwidth / (splitwidth ? splitwidth : pswidth);
yscale = reqheight / (splitheight ? splitheight : psheight);
scale = (xscale < yscale) ? xscale : yscale;
}
new_width = splitwidth ? splitwidth : scale * pswidth;
new_height = splitheight ? splitheight : scale * psheight;
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
else
fprintf (fd, "%%%%PageOrientation: %s\n", (new_width > new_height) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)new_width, (long)new_height);
fprintf (fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
new_width, new_height);
}
else /* No viewport defined with -H or -W */
{
if ((pgwidth == 0) && (pgheight == 0)) /* Image not scaled */
{
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
else
fprintf (fd, "%%%%PageOrientation: %s\n", (pswidth > psheight) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)pswidth, (long)psheight);
fprintf(fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
pswidth, psheight);
}
else /* Image scaled */
{
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
else
fprintf (fd, "%%%%PageOrientation: %s\n", (reqwidth > reqheight) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)reqwidth, (long)reqheight);
fprintf(fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
reqwidth, reqheight);
}
}
break;
case 90:
case 270: if ((splitheight > 0) || (splitwidth > 0))
{
if (pgwidth != 0 || pgheight != 0)
{
xscale = reqwidth / (splitwidth ? splitwidth : pswidth);
yscale = reqheight / (splitheight ? splitheight : psheight);
scale = (xscale < yscale) ? xscale : yscale;
}
new_width = splitwidth ? splitwidth : scale * psheight;
new_height = splitheight ? splitheight : scale * pswidth;
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
else
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
fprintf (fd, "%%%%PageOrientation: %s\n", (new_width > new_height) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)new_width, (long)new_height);
fprintf (fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
new_width, new_height);
}
else
{
if ((pgwidth == 0) && (pgheight == 0)) /* Image not scaled */
{
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
else
fprintf (fd, "%%%%PageOrientation: %s\n", (psheight > pswidth) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)psheight, (long)pswidth);
fprintf(fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
psheight, pswidth);
}
else /* Image scaled */
{
if (strlen(pageOrientation))
fprintf (fd, "%%%%PageOrientation: %s\n", pageOrientation);
else
fprintf (fd, "%%%%PageOrientation: %s\n", (reqwidth > reqheight) ? "Landscape" : "Portrait");
fprintf (fd, "%%%%PageBoundingBox: 0 0 %ld %ld\n", (long)reqwidth, (long)reqheight);
fprintf(fd, "1 dict begin /PageSize [ %f %f ] def currentdict end setpagedevice\n",
reqwidth, reqheight);
}
}
break;
default: TIFFError ("psPageSize", "Invalid rotation %d", rotation);
return (1);
}
fputs("<<\n /Policies <<\n /PageSize 3\n >>\n>> setpagedevice\n", fd);
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
return (0);
} /* end psPageSize */
/* Mask an image as a series of pages, each only showing a section defined
* by the maxPageHeight or maxPageWidth options.
*/
int psMaskImage(FILE *fd, TIFF *tif, int rotation, int center,
int *npages, double pixwidth, double pixheight,
double left_margin, double bottom_margin,
double pgwidth, double pgheight,
double pswidth, double psheight, double scale)
{
int i, j;
int ximages = 1, yimages = 1;
int pages = *npages;
double view_width = 0;
double view_height = 0;
if (get_viewport (pgwidth, pgheight, pswidth, psheight, &view_width, &view_height, rotation))
{
TIFFError ("get_viewport", "Unable to set image viewport");
return (-1);
}
if (get_subimage_count(pgwidth, pgheight, pswidth, psheight,
&ximages, &yimages, rotation, scale) < 1)
{
TIFFError("get_subimage_count", "Invalid image count: %d columns, %d rows", ximages, yimages);
return (-1);
}
for (i = 0; i < yimages; i++)
{
for (j = 0; j < ximages; j++)
{
pages++;
*npages = pages;
fprintf(fd, "%%%%Page: %d %d\n", pages, pages);
/* Write out the PageSize info for non EPS files */
if (!generateEPSF && ( level2 || level3 ))
{
if (psPageSize(fd, rotation, pgwidth, pgheight,
view_width, view_height, pswidth, psheight))
return (-1);
}
fprintf(fd, "gsave\n");
fprintf(fd, "100 dict begin\n");
if (exportMaskedImage(fd, view_width, view_height, pswidth, psheight,
i, j, left_margin, bottom_margin,
scale, center, rotation))
{
TIFFError("exportMaskedImage", "Invalid image parameters.");
return (-1);
}
PSpage(fd, tif, pixwidth, pixheight);
fprintf(fd, "end\n");
fprintf(fd, "grestore\n");
fprintf(fd, "showpage\n");
}
}
return (pages);
}
/* Compute scale factor and write out file header */
int psStart(FILE *fd, int npages, int auto_rotate, int *rotation, double *scale,
double ox, double oy, double pgwidth, double pgheight,
double reqwidth, double reqheight, double pswidth, double psheight,
double left_offset, double bottom_offset)
{
double maxsource = 0.0; /* Used for auto rotations */
double maxtarget = 0.0;
double xscale = 1.0, yscale = 1.0;
double splitheight;
double splitwidth;
double view_width = 0.0, view_height = 0.0;
double page_width = 0.0, page_height = 0.0;
/* Splitheight and splitwidth are in inches */
splitheight = maxPageHeight * PS_UNIT_SIZE;
splitwidth = maxPageWidth * PS_UNIT_SIZE;
page_width = pgwidth * PS_UNIT_SIZE;
page_height = pgheight * PS_UNIT_SIZE;
/* If user has specified a page width and height and requested the
* image to be auto-rotated to fit on that media, we match the
* longest dimension of the image to the longest dimension of the
* target media but we have to ignore auto rotate if user specified
* maxPageHeight since this makes life way too complicated. */
if (auto_rotate)
{
if ((splitheight != 0) || (splitwidth != 0))
{
TIFFError ("psStart", "Auto-rotate is incompatible with page splitting ");
return (1);
}
/* Find longest edges in image and output media */
maxsource = (pswidth >= psheight) ? pswidth : psheight;
maxtarget = (reqwidth >= reqheight) ? reqwidth : reqheight;
if (((maxsource == pswidth) && (maxtarget != reqwidth)) ||
((maxsource == psheight) && (maxtarget != reqheight)))
{ /* optimal orientaion does not match input orientation */
*rotation = 90;
xscale = (reqwidth - left_offset)/psheight;
yscale = (reqheight - bottom_offset)/pswidth;
}
else /* optimal orientaion matches input orientation */
{
xscale = (reqwidth - left_offset)/pswidth;
yscale = (reqheight - bottom_offset)/psheight;
}
*scale = (xscale < yscale) ? xscale : yscale;
/* Do not scale image beyound original size */
if (*scale > 1.0)
*scale = 1.0;
/* Set the size of the displayed image to requested page size
* and optimal orientation.
*/
if (!npages)
PSHead(fd, reqwidth, reqheight, ox, oy);
return (0);
}
/* N.B. If pgwidth or pgheight are set from maxPageHeight/Width,
* we have a problem with the tests below under splitheight.
*/
switch (*rotation) /* Auto rotate has NOT been specified */
{
case 0:
case 180: if ((splitheight != 0) || (splitwidth != 0))
{ /* Viewport clipped to maxPageHeight or maxPageWidth */
if ((page_width != 0) || (page_height != 0)) /* Image scaled */
{
xscale = (reqwidth - left_offset) / (page_width ? page_width : pswidth);
yscale = (reqheight - bottom_offset) / (page_height ? page_height : psheight);
*scale = (xscale < yscale) ? xscale : yscale;
/*
if (*scale > 1.0)
*scale = 1.0;
*/
}
else /* Image clipped but not scaled */
*scale = 1.0;
view_width = splitwidth ? splitwidth : *scale * pswidth;
view_height = splitheight ? splitheight: *scale * psheight;
}
else /* Viewport not clipped to maxPageHeight or maxPageWidth */
{
if ((page_width != 0) || (page_height != 0))
{ /* Image scaled */
xscale = (reqwidth - left_offset) / pswidth;
yscale = (reqheight - bottom_offset) / psheight;
view_width = reqwidth;
view_height = reqheight;
}
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
else
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
{ /* Image not scaled */
xscale = (pswidth - left_offset)/pswidth;
yscale = (psheight - bottom_offset)/psheight;
view_width = pswidth;
view_height = psheight;
}
}
break;
case 90:
case 270: if ((splitheight != 0) || (splitwidth != 0))
{ /* Viewport clipped to maxPageHeight or maxPageWidth */
if ((page_width != 0) || (page_height != 0)) /* Image scaled */
{
xscale = (reqwidth - left_offset)/ psheight;
yscale = (reqheight - bottom_offset)/ pswidth;
*scale = (xscale < yscale) ? xscale : yscale;
/*
if (*scale > 1.0)
*scale = 1.0;
*/
}
else /* Image clipped but not scaled */
*scale = 1.0;
view_width = splitwidth ? splitwidth : *scale * psheight;
view_height = splitheight ? splitheight : *scale * pswidth;
}
else /* Viewport not clipped to maxPageHeight or maxPageWidth */
{
if ((page_width != 0) || (page_height != 0)) /* Image scaled */
{
xscale = (reqwidth - left_offset) / psheight;
yscale = (reqheight - bottom_offset) / pswidth;
view_width = reqwidth;
view_height = reqheight;
}
else
{
xscale = (pswidth - left_offset)/ psheight;
yscale = (psheight - bottom_offset)/ pswidth;
view_width = psheight;
view_height = pswidth;
}
}
break;
default: TIFFError ("psPageSize", "Invalid rotation %d", *rotation);
return (1);
}
if (!npages)
PSHead(fd, (page_width ? page_width : view_width), (page_height ? page_height : view_height), ox, oy);
*scale = (xscale < yscale) ? xscale : yscale;
if (*scale > 1.0)
*scale = 1.0;
return (0);
}
int get_viewport (double pgwidth, double pgheight, double pswidth, double psheight,
double *view_width, double *view_height, int rotation)
{
/* Only one of maxPageHeight or maxPageWidth can be specified */
if (maxPageHeight != 0) /* Clip the viewport to maxPageHeight on each page */
{
From Eduardo Robles Elvira: 1. libtiffcrop-fix.patch fixes a small problem in tiffcrop, it seems it was incorrectly using TIFFSetField instead of CopyField. And in libtiff-correct-fax-scaling.patch we have some other changes: 2. I had to remove a check in main() that didn't allow maxPageWidth to be bigger than pageWidth. 3. [ Omitted due to question on universality ] 4. the pagewidth variable was being set as the maxpagewidth instead, which made all the calculations bad. This made sense when the check in point 2 was in place, but not anymore. I've modified it so that pagewidth is set with the specified pagewidth when maxpagewidth is bigger. 5. The remaining lines of the patch - in exportMaskedImage() - basically fix the scaling.
2015-06-13 01:49:18 -04:00
if (pgheight != 0 && pgheight < maxPageHeight)
*view_height = pgheight * PS_UNIT_SIZE;
else
*view_height = maxPageHeight * PS_UNIT_SIZE;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
/*
* if (res_unit == RESUNIT_CENTIMETER)
* *view_height /= 2.54F;
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else
{
if (pgheight != 0) /* User has set PageHeight with -h flag */
{
*view_height = pgheight * PS_UNIT_SIZE; /* Postscript size for Page Height in inches */
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
/* if (res_unit == RESUNIT_CENTIMETER)
* *view_height /= 2.54F;
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else /* If no width or height are specified, use the original size from image */
switch (rotation)
{
default:
case 0:
case 180: *view_height = psheight;
break;
case 90:
case 270: *view_height = pswidth;
break;
}
}
if (maxPageWidth != 0) /* Clip the viewport to maxPageWidth on each page */
{
From Eduardo Robles Elvira: 1. libtiffcrop-fix.patch fixes a small problem in tiffcrop, it seems it was incorrectly using TIFFSetField instead of CopyField. And in libtiff-correct-fax-scaling.patch we have some other changes: 2. I had to remove a check in main() that didn't allow maxPageWidth to be bigger than pageWidth. 3. [ Omitted due to question on universality ] 4. the pagewidth variable was being set as the maxpagewidth instead, which made all the calculations bad. This made sense when the check in point 2 was in place, but not anymore. I've modified it so that pagewidth is set with the specified pagewidth when maxpagewidth is bigger. 5. The remaining lines of the patch - in exportMaskedImage() - basically fix the scaling.
2015-06-13 01:49:18 -04:00
if (pgwidth != 0 && pgwidth < maxPageWidth)
*view_width = pgwidth * PS_UNIT_SIZE;
else
*view_width = maxPageWidth * PS_UNIT_SIZE;
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
/* if (res_unit == RESUNIT_CENTIMETER)
* *view_width /= 2.54F;
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else
{
if (pgwidth != 0) /* User has set PageWidth with -w flag */
{
*view_width = pgwidth * PS_UNIT_SIZE; /* Postscript size for Page Width in inches */
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
/* if (res_unit == RESUNIT_CENTIMETER)
* *view_width /= 2.54F;
*/
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
}
else /* If no width or height are specified, use the original size from image */
switch (rotation)
{
default:
case 0:
case 180: *view_width = pswidth;
break;
case 90:
case 270: *view_width = psheight; /* (*view_height / psheight) * psheight; */
break;
}
}
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
return (0);
}
/* pgwidth and pgheight specify page width and height in inches from -h and -w flags
* lm and bm are the LeftMargin and BottomMargin in inches
* center causes the image to be centered on the page if the paper size is
* larger than the image size
* returns the sequence number of the page processed or -1 on error
*/
int TIFF2PS(FILE* fd, TIFF* tif, double pgwidth, double pgheight, double lm, double bm, int center)
{
uint32 pixwidth = 0, pixheight = 0; /* Image width and height in pixels */
double ox = 0.0, oy = 0.0; /* Offset from current Postscript origin */
double pswidth, psheight; /* Original raw image width and height in points */
double view_width, view_height; /* Viewport width and height in points */
double scale = 1.0;
double left_offset = lm * PS_UNIT_SIZE;
double bottom_offset = bm * PS_UNIT_SIZE;
uint32 subfiletype;
uint16* sampleinfo;
static int npages = 0;
if (!TIFFGetField(tif, TIFFTAG_XPOSITION, &ox))
ox = 0;
if (!TIFFGetField(tif, TIFFTAG_YPOSITION, &oy))
oy = 0;
/* Consolidated all the tag information into one code segment, Richard Nolde */
do {
tf_numberstrips = TIFFNumberOfStrips(tif);
TIFFGetFieldDefaulted(tif, TIFFTAG_ROWSPERSTRIP, &tf_rowsperstrip);
TIFFGetFieldDefaulted(tif, TIFFTAG_BITSPERSAMPLE, &bitspersample);
TIFFGetFieldDefaulted(tif, TIFFTAG_SAMPLESPERPIXEL, &samplesperpixel);
TIFFGetFieldDefaulted(tif, TIFFTAG_PLANARCONFIG, &planarconfiguration);
TIFFGetField(tif, TIFFTAG_COMPRESSION, &compression);
TIFFGetFieldDefaulted(tif, TIFFTAG_EXTRASAMPLES, &extrasamples, &sampleinfo);
alpha = (extrasamples == 1 && sampleinfo[0] == EXTRASAMPLE_ASSOCALPHA);
if (!TIFFGetField(tif, TIFFTAG_PHOTOMETRIC, &photometric))
{
switch (samplesperpixel - extrasamples)
{
case 1: if (isCCITTCompression(tif))
photometric = PHOTOMETRIC_MINISWHITE;
else
photometric = PHOTOMETRIC_MINISBLACK;
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 3: photometric = PHOTOMETRIC_RGB;
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
break;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
case 4: photometric = PHOTOMETRIC_SEPARATED;
break;
}
}
/* Read image tags for width and height in pixels pixwidth, pixheight,
* and convert to points pswidth, psheight
*/
setupPageState(tif, &pixwidth, &pixheight, &pswidth, &psheight);
view_width = pswidth;
view_height = psheight;
if (get_viewport (pgwidth, pgheight, pswidth, psheight, &view_width, &view_height, rotation))
{
TIFFError("get_viewport", "Unable to set image viewport");
return (1);
}
/* Write the Postscript file header with Bounding Box and Page Size definitions */
if (psStart(fd, npages, auto_rotate, &rotation, &scale, ox, oy,
pgwidth, pgheight, view_width, view_height, pswidth, psheight,
left_offset, bottom_offset))
return (-1);
if (checkImage(tif)) /* Aborts if unsupported image parameters */
{
tf_bytesperrow = TIFFScanlineSize(tif);
/* Set viewport clipping and scaling options */
if ((maxPageHeight) || (maxPageWidth) || (pgwidth != 0) || (pgheight != 0))
{
if ((maxPageHeight) || (maxPageWidth)) /* used -H or -W option */
{
if (psMaskImage(fd, tif, rotation, center, &npages, pixwidth, pixheight,
left_offset, bottom_offset, pgwidth, pgheight,
pswidth, psheight, scale) < 0)
return (-1);
}
else /* N.B. Setting maxPageHeight no longer sets pgheight */
{
if (pgwidth != 0 || pgheight != 0)
{
/* User did not specify a maxium page height or width using -H or -W flag
* but did use -h or -w flag to scale to a specific size page.
*/
npages++;
fprintf(fd, "%%%%Page: %d %d\n", npages, npages);
if (!generateEPSF && ( level2 || level3 ))
{
/* Write out the PageSize info for non EPS files */
if (psPageSize(fd, rotation, pgwidth, pgheight,
view_width, view_height, pswidth, psheight))
return (-1);
}
fprintf(fd, "gsave\n");
fprintf(fd, "100 dict begin\n");
if (psScaleImage(fd, scale, rotation, center, view_width, view_height,
pswidth, psheight, left_offset, bottom_offset))
return (-1);
PSpage(fd, tif, pixwidth, pixheight);
fprintf(fd, "end\n");
fprintf(fd, "grestore\n");
fprintf(fd, "showpage\n");
}
}
}
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
else /* Simple rotation: user did not use -H, -W, -h or -w */
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
{
npages++;
fprintf(fd, "%%%%Page: %d %d\n", npages, npages);
if (!generateEPSF && ( level2 || level3 ))
{
/* Write out the PageSize info for non EPS files */
if (psPageSize(fd, rotation, pgwidth, pgheight,
view_width, view_height, pswidth, psheight))
return (-1);
}
fprintf(fd, "gsave\n");
fprintf(fd, "100 dict begin\n");
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
if (psRotateImage(fd, rotation, pswidth, psheight, left_offset, bottom_offset))
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
return (-1);
PSpage(fd, tif, pixwidth, pixheight);
fprintf(fd, "end\n");
fprintf(fd, "grestore\n");
fprintf(fd, "showpage\n");
}
}
if (generateEPSF)
break;
* tools/tiff2ps.c:Fix bug in auto rotate option code. Once a rotation angle was set by the auto rotate check, it was retained for all pages that followed instead of being retested for each page. Patch by Richard Nolde.
2013-01-25 21:24:05 -05:00
if (auto_rotate)
rotation = 0.0;
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
TIFFGetFieldDefaulted(tif, TIFFTAG_SUBFILETYPE, &subfiletype);
} while (((subfiletype & FILETYPE_PAGE) || printAll) && TIFFReadDirectory(tif));
return(npages);
Initial revision
1999-07-27 17:50:26 -04:00
}
static char DuplexPreamble[] = "\
%%BeginFeature: *Duplex True\n\
systemdict begin\n\
/languagelevel where { pop languagelevel } { 1 } ifelse\n\
2 ge { 1 dict dup /Duplex true put setpagedevice }\n\
{ statusdict /setduplex known { statusdict begin setduplex true end } if\n\
} ifelse\n\
end\n\
%%EndFeature\n\
";
static char TumblePreamble[] = "\
%%BeginFeature: *Tumble True\n\
systemdict begin\n\
/languagelevel where { pop languagelevel } { 1 } ifelse\n\
2 ge { 1 dict dup /Tumble true put setpagedevice }\n\
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
{ statusdict /settumble known { statusdict begin true settumble end } if\n\
Initial revision
1999-07-27 17:50:26 -04:00
} ifelse\n\
end\n\
%%EndFeature\n\
";
static char AvoidDeadZonePreamble[] = "\
gsave newpath clippath pathbbox grestore\n\
4 2 roll 2 copy translate\n\
exch 3 1 roll sub 3 1 roll sub exch\n\
currentpagedevice /PageSize get aload pop\n\
exch 3 1 roll div 3 1 roll div abs exch abs\n\
2 copy gt { exch } if pop\n\
dup 1 lt { dup scale } { pop } ifelse\n\
";
void
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
PSHead(FILE *fd, double pagewidth, double pageheight, double xoff, double yoff)
Initial revision
1999-07-27 17:50:26 -04:00
{
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
time_t t;
Initial revision
1999-07-27 17:50:26 -04:00
t = time(0);
fprintf(fd, "%%!PS-Adobe-3.0%s\n", generateEPSF ? " EPSF-3.0" : "");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
fprintf(fd, "%%%%Creator: %s\n", creator ? creator : "tiff2ps");
fprintf(fd, "%%%%Title: %s\n", title ? title : filename);
Initial revision
1999-07-27 17:50:26 -04:00
fprintf(fd, "%%%%CreationDate: %s", ctime(&t));
fprintf(fd, "%%%%DocumentData: Clean7Bit\n");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
/* NB: should use PageBoundingBox for each page instead of BoundingBox *
* PageBoundingBox DSC added in PSPageSize function, R Nolde 09-01-2010
*/
fprintf(fd, "%%%%Origin: %ld %ld\n", (long) xoff, (long) yoff);
fprintf(fd, "%%%%BoundingBox: 0 0 %ld %ld\n",
(long) ceil(pagewidth), (long) ceil(pageheight));
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
fprintf(fd, "%%%%LanguageLevel: %d\n", (level3 ? 3 : (level2 ? 2 : 1)));
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (generateEPSF == TRUE)
fprintf(fd, "%%%%Pages: 1 1\n");
else
fprintf(fd, "%%%%Pages: (atend)\n");
Initial revision
1999-07-27 17:50:26 -04:00
fprintf(fd, "%%%%EndComments\n");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (generateEPSF == FALSE)
{
fprintf(fd, "%%%%BeginSetup\n");
if (PSduplex)
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
fprintf(fd, "%s", DuplexPreamble);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (PStumble)
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
fprintf(fd, "%s", TumblePreamble);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (PSavoiddeadzone && (level2 || level3))
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
fprintf(fd, "%s", AvoidDeadZonePreamble);
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
fprintf(fd, "%%%%EndSetup\n");
}
Initial revision
1999-07-27 17:50:26 -04:00
}
void
PSTail(FILE *fd, int npages)
{
fprintf(fd, "%%%%Trailer\n");
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
if (generateEPSF == FALSE)
fprintf(fd, "%%%%Pages: %d\n", npages);
Initial revision
1999-07-27 17:50:26 -04:00
fprintf(fd, "%%%%EOF\n");
}
static int
checkcmap(TIFF* tif, int n, uint16* r, uint16* g, uint16* b)
{
(void) tif;
while (n-- > 0)
if (*r++ >= 256 || *g++ >= 256 || *b++ >= 256)
return (16);
TIFFWarning(filename, "Assuming 8-bit colormap");
return (8);
}
static void
PS_Lvl2colorspace(FILE* fd, TIFF* tif)
{
uint16 *rmap, *gmap, *bmap;
int i, num_colors;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
const char * colorspace_p;
switch ( photometric )
{
case PHOTOMETRIC_SEPARATED:
colorspace_p = "CMYK";
break;
case PHOTOMETRIC_RGB:
colorspace_p = "RGB";
break;
default:
colorspace_p = "Gray";
}
Initial revision
1999-07-27 17:50:26 -04:00
/*
* Set up PostScript Level 2 colorspace according to
* section 4.8 in the PostScript refenence manual.
*/
fputs("% PostScript Level 2 only.\n", fd);
if (photometric != PHOTOMETRIC_PALETTE) {
if (photometric == PHOTOMETRIC_YCBCR) {
/* MORE CODE HERE */
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
fprintf(fd, "/Device%s setcolorspace\n", colorspace_p );
Initial revision
1999-07-27 17:50:26 -04:00
return;
}
/*
* Set up an indexed/palette colorspace
*/
num_colors = (1 << bitspersample);
if (!TIFFGetField(tif, TIFFTAG_COLORMAP, &rmap, &gmap, &bmap)) {
TIFFError(filename,
"Palette image w/o \"Colormap\" tag");
return;
}
if (checkcmap(tif, num_colors, rmap, gmap, bmap) == 16) {
/*
* Convert colormap to 8-bits values.
*/
#define CVT(x) (((x) * 255) / ((1L<<16)-1))
for (i = 0; i < num_colors; i++) {
rmap[i] = CVT(rmap[i]);
gmap[i] = CVT(gmap[i]);
bmap[i] = CVT(bmap[i]);
}
#undef CVT
}
fprintf(fd, "[ /Indexed /DeviceRGB %d", num_colors - 1);
if (ascii85) {
Ascii85Init();
fputs("\n<~", fd);
ascii85breaklen -= 2;
} else
fputs(" <", fd);
for (i = 0; i < num_colors; i++) {
if (ascii85) {
Avoid warnings.
2004-06-04 10:16:14 -04:00
Ascii85Put((unsigned char)rmap[i], fd);
Ascii85Put((unsigned char)gmap[i], fd);
Ascii85Put((unsigned char)bmap[i], fd);
Initial revision
1999-07-27 17:50:26 -04:00
} else {
fputs((i % 8) ? " " : "\n ", fd);
fprintf(fd, "%02x%02x%02x",
rmap[i], gmap[i], bmap[i]);
}
}
if (ascii85)
Ascii85Flush(fd);
else
fputs(">\n", fd);
fputs("] setcolorspace\n", fd);
}
static int
PS_Lvl2ImageDict(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
int use_rawdata;
uint32 tile_width, tile_height;
uint16 predictor, minsamplevalue, maxsamplevalue;
tiff2ps: fix use of wrong data type that caused issues (/Height being written as 0) on 64-bit big endian platforms
2019-08-18 04:52:45 -04:00
uint32 repeat_count;
Initial revision
1999-07-27 17:50:26 -04:00
char im_h[64], im_x[64], im_y[64];
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
char * imageOp = "image";
if ( useImagemask && (bitspersample == 1) )
imageOp = "imagemask";
Initial revision
1999-07-27 17:50:26 -04:00
(void)strcpy(im_x, "0");
Replace sprintf with snprintf, to fix CVE-2013-1961.
2013-05-02 10:44:29 -04:00
(void)snprintf(im_y, sizeof(im_y), "%lu", (long) h);
(void)snprintf(im_h, sizeof(im_h), "%lu", (long) h);
Initial revision
1999-07-27 17:50:26 -04:00
tile_width = w;
tile_height = h;
if (TIFFIsTiled(tif)) {
repeat_count = TIFFNumberOfTiles(tif);
TIFFGetField(tif, TIFFTAG_TILEWIDTH, &tile_width);
TIFFGetField(tif, TIFFTAG_TILELENGTH, &tile_height);
if (tile_width > w || tile_height > h ||
(w % tile_width) != 0 || (h % tile_height != 0)) {
/*
* The tiles does not fit image width and height.
* Set up a clip rectangle for the image unit square.
*/
fputs("0 0 1 1 rectclip\n", fd);
}
if (tile_width < w) {
fputs("/im_x 0 def\n", fd);
(void)strcpy(im_x, "im_x neg");
}
if (tile_height < h) {
fputs("/im_y 0 def\n", fd);
Replace sprintf with snprintf, to fix CVE-2013-1961.
2013-05-02 10:44:29 -04:00
(void)snprintf(im_y, sizeof(im_y), "%lu im_y sub", (unsigned long) h);
Initial revision
1999-07-27 17:50:26 -04:00
}
} else {
repeat_count = tf_numberstrips;
tile_height = tf_rowsperstrip;
if (tile_height > h)
tile_height = h;
if (repeat_count > 1) {
fputs("/im_y 0 def\n", fd);
fprintf(fd, "/im_h %lu def\n",
(unsigned long) tile_height);
(void)strcpy(im_h, "im_h");
Replace sprintf with snprintf, to fix CVE-2013-1961.
2013-05-02 10:44:29 -04:00
(void)snprintf(im_y, sizeof(im_y), "%lu im_y sub", (unsigned long) h);
Initial revision
1999-07-27 17:50:26 -04:00
}
}
/*
* Output start of exec block
*/
fputs("{ % exec\n", fd);
if (repeat_count > 1)
tiff2ps: fix use of wrong data type that caused issues (/Height being written as 0) on 64-bit big endian platforms
2019-08-18 04:52:45 -04:00
fprintf(fd, "%u { %% repeat\n", repeat_count);
Initial revision
1999-07-27 17:50:26 -04:00
/*
* Output filter options and image dictionary.
*/
if (ascii85)
fputs(" /im_stream currentfile /ASCII85Decode filter def\n",
fd);
fputs(" <<\n", fd);
fputs(" /ImageType 1\n", fd);
fprintf(fd, " /Width %lu\n", (unsigned long) tile_width);
Added workaround for some software that may crash when last strip of image contains fewer number of scanlines than specified by the `/Height' variable. See http://bugzilla.remotesensing.org/show_bug.cgi?id=164 for explanation.
2002-06-22 06:32:56 -04:00
/*
* Workaround for some software that may crash when last strip
* of image contains fewer number of scanlines than specified
* by the `/Height' variable. So for stripped images with multiple
* strips we will set `/Height' as `im_h', because one is
* recalculated for each strip - including the (smaller) final strip.
* For tiled images and images with only one strip `/Height' will
* contain number of scanlines in tile (or image height in case of
* one-stripped image).
*/
if (TIFFIsTiled(tif) || tf_numberstrips == 1)
fprintf(fd, " /Height %lu\n", (unsigned long) tile_height);
else
fprintf(fd, " /Height im_h\n");
Initial revision
1999-07-27 17:50:26 -04:00
if (planarconfiguration == PLANARCONFIG_SEPARATE && samplesperpixel > 1)
fputs(" /MultipleDataSources true\n", fd);
fprintf(fd, " /ImageMatrix [ %lu 0 0 %ld %s %s ]\n",
(unsigned long) w, - (long)h, im_x, im_y);
fprintf(fd, " /BitsPerComponent %d\n", bitspersample);
fprintf(fd, " /Interpolate %s\n", interpolate ? "true" : "false");
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
switch (samplesperpixel - extrasamples) {
Initial revision
1999-07-27 17:50:26 -04:00
case 1:
switch (photometric) {
case PHOTOMETRIC_MINISBLACK:
fputs(" /Decode [0 1]\n", fd);
break;
case PHOTOMETRIC_MINISWHITE:
switch (compression) {
case COMPRESSION_CCITTRLE:
case COMPRESSION_CCITTRLEW:
case COMPRESSION_CCITTFAX3:
case COMPRESSION_CCITTFAX4:
/*
* Manage inverting with /Blackis1 flag
* since there migth be uncompressed parts
*/
fputs(" /Decode [0 1]\n", fd);
break;
default:
/*
* ERROR...
*/
fputs(" /Decode [1 0]\n", fd);
break;
}
break;
case PHOTOMETRIC_PALETTE:
TIFFGetFieldDefaulted(tif, TIFFTAG_MINSAMPLEVALUE,
&minsamplevalue);
TIFFGetFieldDefaulted(tif, TIFFTAG_MAXSAMPLEVALUE,
&maxsamplevalue);
fprintf(fd, " /Decode [%u %u]\n",
minsamplevalue, maxsamplevalue);
break;
default:
/*
* ERROR ?
*/
fputs(" /Decode [0 1]\n", fd);
break;
}
break;
case 3:
switch (photometric) {
case PHOTOMETRIC_RGB:
fputs(" /Decode [0 1 0 1 0 1]\n", fd);
break;
case PHOTOMETRIC_MINISWHITE:
case PHOTOMETRIC_MINISBLACK:
default:
/*
* ERROR??
*/
fputs(" /Decode [0 1 0 1 0 1]\n", fd);
break;
}
break;
case 4:
/*
* ERROR??
*/
fputs(" /Decode [0 1 0 1 0 1 0 1]\n", fd);
break;
}
fputs(" /DataSource", fd);
if (planarconfiguration == PLANARCONFIG_SEPARATE &&
samplesperpixel > 1)
fputs(" [", fd);
if (ascii85)
fputs(" im_stream", fd);
else
fputs(" currentfile /ASCIIHexDecode filter", fd);
use_rawdata = TRUE;
switch (compression) {
case COMPRESSION_NONE: /* 1: uncompressed */
break;
case COMPRESSION_CCITTRLE: /* 2: CCITT modified Huffman RLE */
case COMPRESSION_CCITTRLEW: /* 32771: #1 w/ word alignment */
case COMPRESSION_CCITTFAX3: /* 3: CCITT Group 3 fax encoding */
case COMPRESSION_CCITTFAX4: /* 4: CCITT Group 4 fax encoding */
fputs("\n\t<<\n", fd);
if (compression == COMPRESSION_CCITTFAX3) {
uint32 g3_options;
fputs("\t /EndOfLine true\n", fd);
fputs("\t /EndOfBlock false\n", fd);
if (!TIFFGetField(tif, TIFFTAG_GROUP3OPTIONS,
&g3_options))
g3_options = 0;
if (g3_options & GROUP3OPT_2DENCODING)
fprintf(fd, "\t /K %s\n", im_h);
if (g3_options & GROUP3OPT_UNCOMPRESSED)
fputs("\t /Uncompressed true\n", fd);
if (g3_options & GROUP3OPT_FILLBITS)
fputs("\t /EncodedByteAlign true\n", fd);
}
if (compression == COMPRESSION_CCITTFAX4) {
uint32 g4_options;
fputs("\t /K -1\n", fd);
TIFFGetFieldDefaulted(tif, TIFFTAG_GROUP4OPTIONS,
&g4_options);
if (g4_options & GROUP4OPT_UNCOMPRESSED)
fputs("\t /Uncompressed true\n", fd);
}
if (!(tile_width == w && w == 1728U))
fprintf(fd, "\t /Columns %lu\n",
(unsigned long) tile_width);
fprintf(fd, "\t /Rows %s\n", im_h);
if (compression == COMPRESSION_CCITTRLE ||
compression == COMPRESSION_CCITTRLEW) {
fputs("\t /EncodedByteAlign true\n", fd);
fputs("\t /EndOfBlock false\n", fd);
}
if (photometric == PHOTOMETRIC_MINISBLACK)
fputs("\t /BlackIs1 true\n", fd);
fprintf(fd, "\t>> /CCITTFaxDecode filter");
break;
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
case COMPRESSION_LZW: /* 5: Lempel-Ziv & Welch */
Initial revision
1999-07-27 17:50:26 -04:00
TIFFGetFieldDefaulted(tif, TIFFTAG_PREDICTOR, &predictor);
if (predictor == 2) {
fputs("\n\t<<\n", fd);
fprintf(fd, "\t /Predictor %u\n", predictor);
fprintf(fd, "\t /Columns %lu\n",
(unsigned long) tile_width);
fprintf(fd, "\t /Colors %u\n", samplesperpixel);
fprintf(fd, "\t /BitsPerComponent %u\n",
bitspersample);
fputs("\t>>", fd);
}
fputs(" /LZWDecode filter", fd);
break;
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
case COMPRESSION_DEFLATE: /* 5: ZIP */
case COMPRESSION_ADOBE_DEFLATE:
if ( level3 ) {
TIFFGetFieldDefaulted(tif, TIFFTAG_PREDICTOR, &predictor);
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
if (predictor > 1) {
Fixed PS comment emitted when FlateDecode is being used. Reported by Tom Kacvinsky.
2003-08-08 15:46:20 -04:00
fprintf(fd, "\t %% PostScript Level 3 only.");
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
fputs("\n\t<<\n", fd);
fprintf(fd, "\t /Predictor %u\n", predictor);
fprintf(fd, "\t /Columns %lu\n",
(unsigned long) tile_width);
fprintf(fd, "\t /Colors %u\n", samplesperpixel);
fprintf(fd, "\t /BitsPerComponent %u\n",
bitspersample);
fputs("\t>>", fd);
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
}
fputs(" /FlateDecode filter", fd);
} else {
use_rawdata = FALSE ;
}
break;
Initial revision
1999-07-27 17:50:26 -04:00
case COMPRESSION_PACKBITS: /* 32773: Macintosh RLE */
fputs(" /RunLengthDecode filter", fd);
use_rawdata = TRUE;
break;
case COMPRESSION_OJPEG: /* 6: !6.0 JPEG */
case COMPRESSION_JPEG: /* 7: %JPEG DCT compression */
#ifdef notdef
/*
* Code not tested yet
*/
fputs(" /DCTDecode filter", fd);
use_rawdata = TRUE;
#else
use_rawdata = FALSE;
#endif
break;
case COMPRESSION_NEXT: /* 32766: NeXT 2-bit RLE */
case COMPRESSION_THUNDERSCAN: /* 32809: ThunderScan RLE */
case COMPRESSION_PIXARFILM: /* 32908: Pixar companded 10bit LZW */
case COMPRESSION_JBIG: /* 34661: ISO JBIG */
use_rawdata = FALSE;
break;
case COMPRESSION_SGILOG: /* 34676: SGI LogL or LogLuv */
case COMPRESSION_SGILOG24: /* 34677: SGI 24-bit LogLuv */
use_rawdata = FALSE;
break;
default:
/*
* ERROR...
*/
use_rawdata = FALSE;
break;
}
if (planarconfiguration == PLANARCONFIG_SEPARATE &&
samplesperpixel > 1) {
uint16 i;
/*
* NOTE: This code does not work yet...
*/
for (i = 1; i < samplesperpixel; i++)
fputs(" dup", fd);
fputs(" ]", fd);
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
fprintf( fd, "\n >> %s\n", imageOp );
Initial revision
1999-07-27 17:50:26 -04:00
if (ascii85)
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
fputs(" im_stream status { im_stream flushfile } if\n", fd);
Initial revision
1999-07-27 17:50:26 -04:00
if (repeat_count > 1) {
if (tile_width < w) {
fprintf(fd, " /im_x im_x %lu add def\n",
(unsigned long) tile_width);
if (tile_height < h) {
fprintf(fd, " im_x %lu ge {\n",
(unsigned long) w);
fputs(" /im_x 0 def\n", fd);
fprintf(fd, " /im_y im_y %lu add def\n",
(unsigned long) tile_height);
fputs(" } if\n", fd);
}
}
if (tile_height < h) {
if (tile_width >= w) {
fprintf(fd, " /im_y im_y %lu add def\n",
(unsigned long) tile_height);
if (!TIFFIsTiled(tif)) {
fprintf(fd, " /im_h %lu im_y sub",
(unsigned long) h);
fprintf(fd, " dup %lu gt { pop",
(unsigned long) tile_height);
fprintf(fd, " %lu } if def\n",
(unsigned long) tile_height);
}
}
}
fputs("} repeat\n", fd);
}
/*
* End of exec function
*/
fputs("}\n", fd);
return(use_rawdata);
}
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
/* Flip the byte order of buffers with 16 bit samples */
static void
fix 64bit type issues
2008-01-07 10:47:54 -05:00
PS_FlipBytes(unsigned char* buf, tsize_t count)
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
{
int i;
unsigned char temp;
if (count <= 0 || bitspersample <= 8) {
return;
}
count--;
for (i = 0; i < count; i += 2) {
temp = buf[i];
buf[i] = buf[i + 1];
buf[i + 1] = temp;
}
}
Readability fixes.
2004-10-10 07:58:16 -04:00
#define MAXLINE 36
Initial revision
1999-07-27 17:50:26 -04:00
int
PS_Lvl2page(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
uint16 fillorder;
Readability fixes.
2004-10-10 07:58:16 -04:00
int use_rawdata, tiled_image, breaklen = MAXLINE;
* tools/{tiff2pdf.c, tiff2ps.c, tiffinfo.c}: Offset and length tags now require 64-bit parameter rather than 32-bit. * libtiff/tif_dirread.c: Fixed issues with unaligned access to 64-bit values.
2008-12-30 22:06:27 -05:00
uint32 chunk_no, num_chunks;
uint64 *bc;
Initial revision
1999-07-27 17:50:26 -04:00
unsigned char *buf_data, *cp;
tsize_t chunk_size, byte_count;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t ascii85_l; /* Length, in bytes, of ascii85_p[] data */
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
uint8 * ascii85_p = 0; /* Holds ASCII85 encoded data */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
PS_Lvl2colorspace(fd, tif);
use_rawdata = PS_Lvl2ImageDict(fd, tif, w, h);
bug80 - remove BeginData/EndData from PostScript output
2002-09-08 13:49:22 -04:00
/* See http://bugzilla.remotesensing.org/show_bug.cgi?id=80 */
#ifdef ENABLE_BROKEN_BEGINENDDATA
Initial revision
1999-07-27 17:50:26 -04:00
fputs("%%BeginData:\n", fd);
bug80 - remove BeginData/EndData from PostScript output
2002-09-08 13:49:22 -04:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
fputs("exec\n", fd);
tiled_image = TIFFIsTiled(tif);
if (tiled_image) {
num_chunks = TIFFNumberOfTiles(tif);
TIFFGetField(tif, TIFFTAG_TILEBYTECOUNTS, &bc);
} else {
num_chunks = TIFFNumberOfStrips(tif);
TIFFGetField(tif, TIFFTAG_STRIPBYTECOUNTS, &bc);
}
if (use_rawdata) {
Avoid warnings.
2004-06-04 10:16:14 -04:00
chunk_size = (tsize_t) bc[0];
Initial revision
1999-07-27 17:50:26 -04:00
for (chunk_no = 1; chunk_no < num_chunks; chunk_no++)
Avoid warnings.
2004-06-04 10:16:14 -04:00
if ((tsize_t) bc[chunk_no] > chunk_size)
chunk_size = (tsize_t) bc[chunk_no];
Initial revision
1999-07-27 17:50:26 -04:00
} else {
if (tiled_image)
chunk_size = TIFFTileSize(tif);
else
chunk_size = TIFFStripSize(tif);
}
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
buf_data = (unsigned char *)limitMalloc(chunk_size);
Initial revision
1999-07-27 17:50:26 -04:00
if (!buf_data) {
* libtiff/tiffio.h: GCC will now validate format specifications for TIFFError(), TIFFErrorExt(), TIFFWarning(), and TIFFWarningExt() in order to reveal bugs. * Many fixes throughout to work better as a 64-bit build.
2008-12-31 18:48:01 -05:00
TIFFError(filename, "Can't alloc %lu bytes for %s.",
(unsigned long) chunk_size, tiled_image ? "tiles" : "strips");
Initial revision
1999-07-27 17:50:26 -04:00
return(FALSE);
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
if ( ascii85 ) {
/*
* Allocate a buffer to hold the ASCII85 encoded data. Note
* that it is allocated with sufficient room to hold the
* encoded data (5*chunk_size/4) plus the EOD marker (+8)
* and formatting line breaks. The line breaks are more
* than taken care of by using 6*chunk_size/4 rather than
* 5*chunk_size/4.
*/
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
ascii85_p = limitMalloc( (chunk_size+(chunk_size/2)) + 8 );
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( !ascii85_p ) {
_TIFFfree( buf_data );
TIFFError( filename, "Cannot allocate ASCII85 encoding buffer." );
return ( FALSE );
}
}
#endif
Initial revision
1999-07-27 17:50:26 -04:00
TIFFGetFieldDefaulted(tif, TIFFTAG_FILLORDER, &fillorder);
for (chunk_no = 0; chunk_no < num_chunks; chunk_no++) {
if (ascii85)
Ascii85Init();
else
Readability fixes.
2004-10-10 07:58:16 -04:00
breaklen = MAXLINE;
Initial revision
1999-07-27 17:50:26 -04:00
if (use_rawdata) {
if (tiled_image)
byte_count = TIFFReadRawTile(tif, chunk_no,
buf_data, chunk_size);
else
byte_count = TIFFReadRawStrip(tif, chunk_no,
buf_data, chunk_size);
if (fillorder == FILLORDER_LSB2MSB)
TIFFReverseBits(buf_data, byte_count);
} else {
if (tiled_image)
byte_count = TIFFReadEncodedTile(tif,
chunk_no, buf_data,
chunk_size);
else
byte_count = TIFFReadEncodedStrip(tif,
chunk_no, buf_data,
chunk_size);
}
if (byte_count < 0) {
TIFFError(filename, "Can't read %s %d.",
tiled_image ? "tile" : "strip", chunk_no);
if (ascii85)
Ascii85Put('\0', fd);
}
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
/*
* for 16 bits, the two bytes must be most significant
* byte first
*/
if (bitspersample == 16 && !TIFFIsBigEndian(tif)) {
PS_FlipBytes(buf_data, byte_count);
}
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
/*
* For images with alpha, matte against a white background;
* i.e. Cback * (1 - Aimage) where Cback = 1. We will fill the
* lower part of the buffer with the modified values.
*
* XXX: needs better solution
*/
if (alpha) {
int adjust, i, j = 0;
int ncomps = samplesperpixel - extrasamples;
tiff2ps: fix heap-buffer-overflow http://bugzilla.maptools.org/show_bug.cgi?id=2834 usually the test (i < byte_count) is OK because the byte_count is divisible by samplesperpixel. But if that is not the case, (i + ncomps) < byte_count should be used, or maybe (i + samplesperpixel) <= byte_count
2019-01-29 04:42:19 -05:00
for (i = 0; (i + ncomps) < byte_count; i+=samplesperpixel) {
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
adjust = 255 - buf_data[i + ncomps];
switch (ncomps) {
case 1:
buf_data[j++] = buf_data[i] + adjust;
break;
case 2:
buf_data[j++] = buf_data[i] + adjust;
buf_data[j++] = buf_data[i+1] + adjust;
break;
case 3:
buf_data[j++] = buf_data[i] + adjust;
buf_data[j++] = buf_data[i+1] + adjust;
buf_data[j++] = buf_data[i+2] + adjust;
break;
}
}
byte_count -= j;
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if (ascii85) {
#if defined( EXP_ASCII85ENCODER )
ascii85_l = Ascii85EncodeBlock(ascii85_p, 1, buf_data, byte_count );
if ( ascii85_l > 0 )
fwrite( ascii85_p, ascii85_l, 1, fd );
#else
for (cp = buf_data; byte_count > 0; byte_count--)
Initial revision
1999-07-27 17:50:26 -04:00
Ascii85Put(*cp++, fd);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif
}
else
{
for (cp = buf_data; byte_count > 0; byte_count--) {
putc(hex[((*cp)>>4)&0xf], fd);
putc(hex[(*cp)&0xf], fd);
cp++;
Initial revision
1999-07-27 17:50:26 -04:00
if (--breaklen <= 0) {
putc('\n', fd);
Readability fixes.
2004-10-10 07:58:16 -04:00
breaklen = MAXLINE;
Initial revision
1999-07-27 17:50:26 -04:00
}
}
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( !ascii85 ) {
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
if ( level2 || level3 )
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
putc( '>', fd );
Initial revision
1999-07-27 17:50:26 -04:00
putc('\n', fd);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
}
#if !defined( EXP_ASCII85ENCODER )
else
Ascii85Flush(fd);
#endif
Initial revision
1999-07-27 17:50:26 -04:00
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
if ( ascii85_p )
_TIFFfree( ascii85_p );
#endif
Initial revision
1999-07-27 17:50:26 -04:00
_TIFFfree(buf_data);
bug80 - remove BeginData/EndData from PostScript output
2002-09-08 13:49:22 -04:00
#ifdef ENABLE_BROKEN_BEGINENDDATA
Initial revision
1999-07-27 17:50:26 -04:00
fputs("%%EndData\n", fd);
bug80 - remove BeginData/EndData from PostScript output
2002-09-08 13:49:22 -04:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
return(TRUE);
}
void
PSpage(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
char * imageOp = "image";
if ( useImagemask && (bitspersample == 1) )
imageOp = "imagemask";
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
if ((level2 || level3) && PS_Lvl2page(fd, tif, w, h))
Initial revision
1999-07-27 17:50:26 -04:00
return;
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
ps_bytesperrow = tf_bytesperrow - (extrasamples * bitspersample / 8)*w;
Initial revision
1999-07-27 17:50:26 -04:00
switch (photometric) {
case PHOTOMETRIC_RGB:
if (planarconfiguration == PLANARCONFIG_CONTIG) {
fprintf(fd, "%s", RGBcolorimage);
PSColorContigPreamble(fd, w, h, 3);
PSDataColorContig(fd, tif, w, h, 3);
} else {
PSColorSeparatePreamble(fd, w, h, 3);
PSDataColorSeparate(fd, tif, w, h, 3);
}
break;
case PHOTOMETRIC_SEPARATED:
/* XXX should emit CMYKcolorimage */
if (planarconfiguration == PLANARCONFIG_CONTIG) {
PSColorContigPreamble(fd, w, h, 4);
PSDataColorContig(fd, tif, w, h, 4);
} else {
PSColorSeparatePreamble(fd, w, h, 4);
PSDataColorSeparate(fd, tif, w, h, 4);
}
break;
case PHOTOMETRIC_PALETTE:
fprintf(fd, "%s", RGBcolorimage);
PhotoshopBanner(fd, w, h, 1, 3, "false 3 colorimage");
fprintf(fd, "/scanLine %ld string def\n",
The change below fixes a problem where "tiff2ps -1e" did not make the scanline buffer long enough when writing rgb triplets. The scanline needs to be 3 X the number of dots or else it will contain an incomplete triplet and programs that try to separate the eps by redefining the colorimage operator will get messed up. William http://williambader.com
2002-10-14 22:39:09 -04:00
(long) ps_bytesperrow * 3L);
Initial revision
1999-07-27 17:50:26 -04:00
fprintf(fd, "%lu %lu 8\n",
(unsigned long) w, (unsigned long) h);
fprintf(fd, "[%lu 0 0 -%lu 0 %lu]\n",
(unsigned long) w, (unsigned long) h, (unsigned long) h);
fprintf(fd, "{currentfile scanLine readhexstring pop} bind\n");
fprintf(fd, "false 3 colorimage\n");
PSDataPalette(fd, tif, w, h);
break;
case PHOTOMETRIC_MINISBLACK:
case PHOTOMETRIC_MINISWHITE:
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
PhotoshopBanner(fd, w, h, 1, 1, imageOp);
Initial revision
1999-07-27 17:50:26 -04:00
fprintf(fd, "/scanLine %ld string def\n",
(long) ps_bytesperrow);
fprintf(fd, "%lu %lu %d\n",
(unsigned long) w, (unsigned long) h, bitspersample);
fprintf(fd, "[%lu 0 0 -%lu 0 %lu]\n",
(unsigned long) w, (unsigned long) h, (unsigned long) h);
fprintf(fd,
"{currentfile scanLine readhexstring pop} bind\n");
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
fprintf(fd, "%s\n", imageOp);
Initial revision
1999-07-27 17:50:26 -04:00
PSDataBW(fd, tif, w, h);
break;
}
putc('\n', fd);
}
void
PSColorContigPreamble(FILE* fd, uint32 w, uint32 h, int nc)
{
ps_bytesperrow = nc * (tf_bytesperrow / samplesperpixel);
PhotoshopBanner(fd, w, h, 1, nc, "false %d colorimage");
fprintf(fd, "/line %ld string def\n", (long) ps_bytesperrow);
fprintf(fd, "%lu %lu %d\n",
(unsigned long) w, (unsigned long) h, bitspersample);
fprintf(fd, "[%lu 0 0 -%lu 0 %lu]\n",
(unsigned long) w, (unsigned long) h, (unsigned long) h);
fprintf(fd, "{currentfile line readhexstring pop} bind\n");
fprintf(fd, "false %d colorimage\n", nc);
}
void
PSColorSeparatePreamble(FILE* fd, uint32 w, uint32 h, int nc)
{
int i;
PhotoshopBanner(fd, w, h, ps_bytesperrow, nc, "true %d colorimage");
for (i = 0; i < nc; i++)
fprintf(fd, "/line%d %ld string def\n",
i, (long) ps_bytesperrow);
fprintf(fd, "%lu %lu %d\n",
(unsigned long) w, (unsigned long) h, bitspersample);
fprintf(fd, "[%lu 0 0 -%lu 0 %lu] \n",
(unsigned long) w, (unsigned long) h, (unsigned long) h);
for (i = 0; i < nc; i++)
fprintf(fd, "{currentfile line%d readhexstring pop}bind\n", i);
fprintf(fd, "true %d colorimage\n", nc);
}
#define DOBREAK(len, howmany, fd) \
if (((len) -= (howmany)) <= 0) { \
putc('\n', fd); \
(len) = MAXLINE-(howmany); \
}
#define PUTHEX(c,fd) putc(hex[((c)>>4)&0xf],fd); putc(hex[(c)&0xf],fd)
void
PSDataColorContig(FILE* fd, TIFF* tif, uint32 w, uint32 h, int nc)
{
uint32 row;
fix 64bit type issues
2008-01-07 10:47:54 -05:00
int breaklen = MAXLINE, es = samplesperpixel - nc;
tsize_t cc;
Initial revision
1999-07-27 17:50:26 -04:00
unsigned char *tf_buf;
unsigned char *cp, c;
(void) w;
Fix tiff2ps error regarding "Inconsistent value of es" by allowing es to be zero. Problem was reported to the tiff mailing list by Julian H. Stacey on January 5, 2019.
2019-01-05 14:56:09 -05:00
if( es < 0 )
* tools/tiff2ps.c: fix 2 heap-based buffer overflows (in PSDataBW and PSDataColorContig). Reported by Agostino Sarubbo. Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2633 and http://bugzilla.maptools.org/show_bug.cgi?id=2634.
2016-12-17 14:45:28 -05:00
{
Fix tiff2ps error regarding "Inconsistent value of es" by allowing es to be zero. Problem was reported to the tiff mailing list by Julian H. Stacey on January 5, 2019.
2019-01-05 14:56:09 -05:00
TIFFError(filename, "Inconsistent value of es: %d (samplesperpixel=%u, nc=%d)", es, samplesperpixel, nc);
* tools/tiff2ps.c: fix 2 heap-based buffer overflows (in PSDataBW and PSDataColorContig). Reported by Agostino Sarubbo. Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2633 and http://bugzilla.maptools.org/show_bug.cgi?id=2634.
2016-12-17 14:45:28 -05:00
return;
}
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
tf_buf = (unsigned char *) limitMalloc(tf_bytesperrow);
Initial revision
1999-07-27 17:50:26 -04:00
if (tf_buf == NULL) {
TIFFError(filename, "No space for scanline buffer");
return;
}
for (row = 0; row < h; row++) {
if (TIFFReadScanline(tif, tf_buf, row, 0) < 0)
break;
cp = tf_buf;
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
/*
* for 16 bits, the two bytes must be most significant
* byte first
*/
if (bitspersample == 16 && !HOST_BIGENDIAN) {
PS_FlipBytes(cp, tf_bytesperrow);
}
Initial revision
1999-07-27 17:50:26 -04:00
if (alpha) {
int adjust;
tiff2ps: fix heap buffer read overflow in PSDataColorContig() fixes #161 / http://bugzilla.maptools.org/show_bug.cgi?id=2855 in 05029fb7f1ecf771abaf90b5705b6cab9eb522a7 I missed that 1 extra byte is read in this loop.
2020-02-08 05:17:08 -05:00
/*
* the code inside this loop reads nc bytes + 1 extra byte (for adjust)
*/
for (cc = 0; (cc + nc) < tf_bytesperrow; cc += samplesperpixel) {
Initial revision
1999-07-27 17:50:26 -04:00
DOBREAK(breaklen, nc, fd);
/*
* For images with alpha, matte against
* a white background; i.e.
* Cback * (1 - Aimage)
* where Cback = 1.
*/
adjust = 255 - cp[nc];
switch (nc) {
case 4: c = *cp++ + adjust; PUTHEX(c,fd);
case 3: c = *cp++ + adjust; PUTHEX(c,fd);
case 2: c = *cp++ + adjust; PUTHEX(c,fd);
case 1: c = *cp++ + adjust; PUTHEX(c,fd);
}
cp += es;
}
} else {
tiff2ps: fix heap buffer read overflow in PSDataColorContig() fixes #161 / http://bugzilla.maptools.org/show_bug.cgi?id=2855 in 05029fb7f1ecf771abaf90b5705b6cab9eb522a7 I missed that 1 extra byte is read in this loop.
2020-02-08 05:17:08 -05:00
/*
* the code inside this loop reads nc bytes per iteration
*/
for (cc = 0; (cc + nc) <= tf_bytesperrow; cc += samplesperpixel) {
Initial revision
1999-07-27 17:50:26 -04:00
DOBREAK(breaklen, nc, fd);
switch (nc) {
case 4: c = *cp++; PUTHEX(c,fd);
case 3: c = *cp++; PUTHEX(c,fd);
case 2: c = *cp++; PUTHEX(c,fd);
case 1: c = *cp++; PUTHEX(c,fd);
}
cp += es;
}
}
}
_TIFFfree((char *) tf_buf);
}
void
PSDataColorSeparate(FILE* fd, TIFF* tif, uint32 w, uint32 h, int nc)
{
uint32 row;
fix 64bit type issues
2008-01-07 10:47:54 -05:00
int breaklen = MAXLINE;
tsize_t cc;
Avoid warnings.
2004-06-04 10:16:14 -04:00
tsample_t s, maxs;
Initial revision
1999-07-27 17:50:26 -04:00
unsigned char *tf_buf;
unsigned char *cp, c;
(void) w;
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
tf_buf = (unsigned char *) limitMalloc(tf_bytesperrow);
Initial revision
1999-07-27 17:50:26 -04:00
if (tf_buf == NULL) {
TIFFError(filename, "No space for scanline buffer");
return;
}
maxs = (samplesperpixel > nc ? nc : samplesperpixel);
for (row = 0; row < h; row++) {
for (s = 0; s < maxs; s++) {
if (TIFFReadScanline(tif, tf_buf, row, s) < 0)
break;
for (cp = tf_buf, cc = 0; cc < tf_bytesperrow; cc++) {
DOBREAK(breaklen, 1, fd);
c = *cp++;
PUTHEX(c,fd);
}
}
}
_TIFFfree((char *) tf_buf);
}
#define PUTRGBHEX(c,fd) \
PUTHEX(rmap[c],fd); PUTHEX(gmap[c],fd); PUTHEX(bmap[c],fd)
void
PSDataPalette(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
uint16 *rmap, *gmap, *bmap;
uint32 row;
fix 64bit type issues
2008-01-07 10:47:54 -05:00
int breaklen = MAXLINE, nc;
tsize_t cc;
Initial revision
1999-07-27 17:50:26 -04:00
unsigned char *tf_buf;
unsigned char *cp, c;
(void) w;
if (!TIFFGetField(tif, TIFFTAG_COLORMAP, &rmap, &gmap, &bmap)) {
TIFFError(filename, "Palette image w/o \"Colormap\" tag");
return;
}
switch (bitspersample) {
case 8: case 4: case 2: case 1:
break;
default:
TIFFError(filename, "Depth %d not supported", bitspersample);
return;
}
nc = 3 * (8 / bitspersample);
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
tf_buf = (unsigned char *) limitMalloc(tf_bytesperrow);
Initial revision
1999-07-27 17:50:26 -04:00
if (tf_buf == NULL) {
TIFFError(filename, "No space for scanline buffer");
return;
}
if (checkcmap(tif, 1<<bitspersample, rmap, gmap, bmap) == 16) {
int i;
Avoid warnings.
2004-06-04 10:16:14 -04:00
#define CVT(x) ((unsigned short) (((x) * 255) / ((1U<<16)-1)))
Initial revision
1999-07-27 17:50:26 -04:00
for (i = (1<<bitspersample)-1; i >= 0; i--) {
rmap[i] = CVT(rmap[i]);
gmap[i] = CVT(gmap[i]);
bmap[i] = CVT(bmap[i]);
}
#undef CVT
}
for (row = 0; row < h; row++) {
if (TIFFReadScanline(tif, tf_buf, row, 0) < 0)
break;
for (cp = tf_buf, cc = 0; cc < tf_bytesperrow; cc++) {
DOBREAK(breaklen, nc, fd);
switch (bitspersample) {
case 8:
c = *cp++; PUTRGBHEX(c, fd);
break;
case 4:
c = *cp++; PUTRGBHEX(c&0xf, fd);
c >>= 4; PUTRGBHEX(c, fd);
break;
case 2:
c = *cp++; PUTRGBHEX(c&0x3, fd);
c >>= 2; PUTRGBHEX(c&0x3, fd);
c >>= 2; PUTRGBHEX(c&0x3, fd);
c >>= 2; PUTRGBHEX(c, fd);
break;
case 1:
c = *cp++; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c&0x1, fd);
c >>= 1; PUTRGBHEX(c, fd);
break;
}
}
}
_TIFFfree((char *) tf_buf);
}
void
PSDataBW(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
int breaklen = MAXLINE;
unsigned char* tf_buf;
unsigned char* cp;
tsize_t stripsize = TIFFStripSize(tif);
tstrip_t s;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t ascii85_l; /* Length, in bytes, of ascii85_p[] data */
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
uint8 *ascii85_p = 0; /* Holds ASCII85 encoded data */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif
Initial revision
1999-07-27 17:50:26 -04:00
(void) w; (void) h;
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
tf_buf = (unsigned char *) limitMalloc(stripsize);
Initial revision
1999-07-27 17:50:26 -04:00
if (tf_buf == NULL) {
TIFFError(filename, "No space for scanline buffer");
return;
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
* libtiff/tif_dirread.c (TIFFFetchStripThing): Free "data" also upon failure to allocate "resizeddata". * tools/tiff2ps.c (PSDataBW): Zero buffer *after* checking for allocation failure, not before.
2011-05-31 13:10:18 -04:00
// FIXME
memset(tf_buf, 0, stripsize);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
if ( ascii85 ) {
/*
* Allocate a buffer to hold the ASCII85 encoded data. Note
* that it is allocated with sufficient room to hold the
* encoded data (5*stripsize/4) plus the EOD marker (+8)
* and formatting line breaks. The line breaks are more
* than taken care of by using 6*stripsize/4 rather than
* 5*stripsize/4.
*/
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
ascii85_p = limitMalloc( (stripsize+(stripsize/2)) + 8 );
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( !ascii85_p ) {
_TIFFfree( tf_buf );
TIFFError( filename, "Cannot allocate ASCII85 encoding buffer." );
return;
}
}
#endif
Initial revision
1999-07-27 17:50:26 -04:00
if (ascii85)
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
Ascii85Init();
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
Initial revision
1999-07-27 17:50:26 -04:00
for (s = 0; s < TIFFNumberOfStrips(tif); s++) {
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tmsize_t cc = TIFFReadEncodedStrip(tif, s, tf_buf, stripsize);
Initial revision
1999-07-27 17:50:26 -04:00
if (cc < 0) {
TIFFError(filename, "Can't read strip");
break;
}
cp = tf_buf;
if (photometric == PHOTOMETRIC_MINISWHITE) {
for (cp += cc; --cp >= tf_buf;)
*cp = ~*cp;
cp++;
}
Added support 16-bit images as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=1566 Patch from William Bader.
2007-07-03 12:00:53 -04:00
/*
* for 16 bits, the two bytes must be most significant
* byte first
*/
if (bitspersample == 16 && !HOST_BIGENDIAN) {
PS_FlipBytes(cp, cc);
}
Initial revision
1999-07-27 17:50:26 -04:00
if (ascii85) {
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
if (alpha) {
int adjust, i;
for (i = 0; i < cc; i+=2) {
adjust = 255 - cp[i + 1];
cp[i / 2] = cp[i] + adjust;
}
cc /= 2;
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
ascii85_l = Ascii85EncodeBlock( ascii85_p, 1, cp, cc );
if ( ascii85_l > 0 )
fwrite( ascii85_p, ascii85_l, 1, fd );
#else
Initial revision
1999-07-27 17:50:26 -04:00
while (cc-- > 0)
Ascii85Put(*cp++, fd);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif /* EXP_ASCII85_ENCODER */
Initial revision
1999-07-27 17:50:26 -04:00
} else {
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
unsigned char c;
if (alpha) {
int adjust;
* tools/tiff2ps.c: fix 2 heap-based buffer overflows (in PSDataBW and PSDataColorContig). Reported by Agostino Sarubbo. Fixes http://bugzilla.maptools.org/show_bug.cgi?id=2633 and http://bugzilla.maptools.org/show_bug.cgi?id=2634.
2016-12-17 14:45:28 -05:00
while (cc-- > 1) {
Added support for alfa channel. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=428
2003-12-19 10:30:27 -05:00
DOBREAK(breaklen, 1, fd);
/*
* For images with alpha, matte against
* a white background; i.e.
* Cback * (1 - Aimage)
* where Cback = 1.
*/
adjust = 255 - cp[1];
c = *cp++ + adjust; PUTHEX(c,fd);
cp++, cc--;
}
} else {
while (cc-- > 0) {
c = *cp++;
DOBREAK(breaklen, 1, fd);
PUTHEX(c, fd);
}
Initial revision
1999-07-27 17:50:26 -04:00
}
}
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( !ascii85 )
{
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
if ( level2 || level3)
fputs(">\n", fd);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
}
#if !defined( EXP_ASCII85ENCODER )
else
Ascii85Flush(fd);
#else
if ( ascii85_p )
_TIFFfree( ascii85_p );
#endif
Initial revision
1999-07-27 17:50:26 -04:00
_TIFFfree(tf_buf);
}
void
PSRawDataBW(FILE* fd, TIFF* tif, uint32 w, uint32 h)
{
* tools/{tiff2pdf.c, tiff2ps.c, tiffinfo.c}: Offset and length tags now require 64-bit parameter rather than 32-bit. * libtiff/tif_dirread.c: Fixed issues with unaligned access to 64-bit values.
2008-12-30 22:06:27 -05:00
uint64 *bc;
Initial revision
1999-07-27 17:50:26 -04:00
uint32 bufsize;
fix 64bit type issues
2008-01-07 10:47:54 -05:00
int breaklen = MAXLINE;
tmsize_t cc;
Initial revision
1999-07-27 17:50:26 -04:00
uint16 fillorder;
unsigned char *tf_buf;
unsigned char *cp, c;
tstrip_t s;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t ascii85_l; /* Length, in bytes, of ascii85_p[] data */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
uint8 * ascii85_p = 0; /* Holds ASCII85 encoded data */
#endif
Initial revision
1999-07-27 17:50:26 -04:00
(void) w; (void) h;
TIFFGetFieldDefaulted(tif, TIFFTAG_FILLORDER, &fillorder);
TIFFGetField(tif, TIFFTAG_STRIPBYTECOUNTS, &bc);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
/*
* Find largest strip:
*/
avoid 32/64bit casting warnings
2009-01-22 16:12:45 -05:00
bufsize = (uint32) bc[0];
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
tiff2ps: fix use of wrong data type that caused issues (/Height being written as 0) on 64-bit big endian platforms
2019-08-18 04:52:45 -04:00
for ( s = 0; ++s < tf_numberstrips; ) {
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( bc[s] > bufsize )
avoid 32/64bit casting warnings
2009-01-22 16:12:45 -05:00
bufsize = (uint32) bc[s];
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
}
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
tf_buf = (unsigned char*) limitMalloc(bufsize);
Initial revision
1999-07-27 17:50:26 -04:00
if (tf_buf == NULL) {
TIFFError(filename, "No space for strip buffer");
return;
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
if ( ascii85 ) {
/*
* Allocate a buffer to hold the ASCII85 encoded data. Note
* that it is allocated with sufficient room to hold the
* encoded data (5*bufsize/4) plus the EOD marker (+8)
* and formatting line breaks. The line breaks are more
* than taken care of by using 6*bufsize/4 rather than
* 5*bufsize/4.
*/
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
ascii85_p = limitMalloc( (bufsize+(bufsize/2)) + 8 );
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( !ascii85_p ) {
_TIFFfree( tf_buf );
TIFFError( filename, "Cannot allocate ASCII85 encoding buffer." );
return;
}
}
#endif
tiff2ps: fix use of wrong data type that caused issues (/Height being written as 0) on 64-bit big endian platforms
2019-08-18 04:52:45 -04:00
for (s = 0; s < tf_numberstrips; s++) {
avoid 32/64bit casting warnings
2009-01-22 16:12:45 -05:00
cc = TIFFReadRawStrip(tif, s, tf_buf, (tmsize_t) bc[s]);
Initial revision
1999-07-27 17:50:26 -04:00
if (cc < 0) {
TIFFError(filename, "Can't read strip");
break;
}
if (fillorder == FILLORDER_LSB2MSB)
TIFFReverseBits(tf_buf, cc);
if (!ascii85) {
for (cp = tf_buf; cc > 0; cc--) {
DOBREAK(breaklen, 1, fd);
c = *cp++;
PUTHEX(c, fd);
}
fputs(">\n", fd);
breaklen = MAXLINE;
} else {
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
Ascii85Init();
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
ascii85_l = Ascii85EncodeBlock( ascii85_p, 1, tf_buf, cc );
if ( ascii85_l > 0 )
fwrite( ascii85_p, ascii85_l, 1, fd );
#else
Initial revision
1999-07-27 17:50:26 -04:00
for (cp = tf_buf; cc > 0; cc--)
Ascii85Put(*cp++, fd);
Ascii85Flush(fd);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#endif /* EXP_ASCII85ENCODER */
Initial revision
1999-07-27 17:50:26 -04:00
}
}
_TIFFfree((char *) tf_buf);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER )
if ( ascii85_p )
_TIFFfree( ascii85_p );
#endif
Initial revision
1999-07-27 17:50:26 -04:00
}
void
Ascii85Init(void)
{
ascii85breaklen = 2*MAXLINE;
ascii85count = 0;
}
static char*
Ascii85Encode(unsigned char* raw)
{
static char encoded[6];
uint32 word;
word = (((raw[0]<<8)+raw[1])<<16) + (raw[2]<<8) + raw[3];
if (word != 0L) {
uint32 q;
uint16 w1;
q = word / (85L*85*85*85); /* actually only a byte */
Avoid warnings.
2004-06-04 10:16:14 -04:00
encoded[0] = (char) (q + '!');
Initial revision
1999-07-27 17:50:26 -04:00
word -= q * (85L*85*85*85); q = word / (85L*85*85);
Avoid warnings.
2004-06-04 10:16:14 -04:00
encoded[1] = (char) (q + '!');
Initial revision
1999-07-27 17:50:26 -04:00
word -= q * (85L*85*85); q = word / (85*85);
Avoid warnings.
2004-06-04 10:16:14 -04:00
encoded[2] = (char) (q + '!');
Initial revision
1999-07-27 17:50:26 -04:00
w1 = (uint16) (word - q*(85L*85));
Avoid warnings.
2004-06-04 10:16:14 -04:00
encoded[3] = (char) ((w1 / 85) + '!');
encoded[4] = (char) ((w1 % 85) + '!');
Initial revision
1999-07-27 17:50:26 -04:00
encoded[5] = '\0';
} else
encoded[0] = 'z', encoded[1] = '\0';
return (encoded);
}
void
Ascii85Put(unsigned char code, FILE* fd)
{
ascii85buf[ascii85count++] = code;
if (ascii85count >= 4) {
unsigned char* p;
int n;
for (n = ascii85count, p = ascii85buf; n >= 4; n -= 4, p += 4) {
char* cp;
for (cp = Ascii85Encode(p); *cp; cp++) {
putc(*cp, fd);
if (--ascii85breaklen == 0) {
putc('\n', fd);
ascii85breaklen = 2*MAXLINE;
}
}
}
_TIFFmemcpy(ascii85buf, p, n);
ascii85count = n;
}
}
void
Ascii85Flush(FILE* fd)
{
if (ascii85count > 0) {
char* res;
_TIFFmemset(&ascii85buf[ascii85count], 0, 3);
res = Ascii85Encode(ascii85buf);
fwrite(res[0] == 'z' ? "!!!!" : res, ascii85count + 1, 1, fd);
}
fputs("~>\n", fd);
}
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
#if defined( EXP_ASCII85ENCODER)
#define A85BREAKCNTR ascii85breaklen
#define A85BREAKLEN (2*MAXLINE)
/*****************************************************************************
*
* Name: Ascii85EncodeBlock( ascii85_p, f_eod, raw_p, raw_l )
*
* Description: This routine will encode the raw data in the buffer described
* by raw_p and raw_l into ASCII85 format and store the encoding
* in the buffer given by ascii85_p.
*
* Parameters: ascii85_p - A buffer supplied by the caller which will
* contain the encoded ASCII85 data.
* f_eod - Flag: Nz means to end the encoded buffer with
* an End-Of-Data marker.
* raw_p - Pointer to the buffer of data to be encoded
* raw_l - Number of bytes in raw_p[] to be encoded
*
* Returns: (int) < 0 Error, see errno
* >= 0 Number of bytes written to ascii85_p[].
*
* Notes: An external variable given by A85BREAKCNTR is used to
* determine when to insert newline characters into the
* encoded data. As each byte is placed into ascii85_p this
* external is decremented. If the variable is decrement to
* or past zero then a newline is inserted into ascii85_p
* and the A85BREAKCNTR is then reset to A85BREAKLEN.
* Note: for efficiency reasons the A85BREAKCNTR variable
* is not actually checked on *every* character
* placed into ascii85_p but often only for every
* 5 characters.
*
* THE CALLER IS RESPONSIBLE FOR ENSURING THAT ASCII85_P[] IS
* SUFFICIENTLY LARGE TO THE ENCODED DATA!
* You will need at least 5 * (raw_l/4) bytes plus space for
* newline characters and space for an EOD marker (if
* requested). A safe calculation is to use 6*(raw_l/4) + 8
* to size ascii85_p.
*
*****************************************************************************/
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t Ascii85EncodeBlock( uint8 * ascii85_p, unsigned f_eod, const uint8 * raw_p, tsize_t raw_l )
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
{
char ascii85[5]; /* Encoded 5 tuple */
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t ascii85_l; /* Number of bytes written to ascii85_p[] */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
int rc; /* Return code */
uint32 val32; /* Unencoded 4 tuple */
ascii85_l = 0; /* Nothing written yet */
if ( raw_p )
{
--raw_p; /* Prepare for pre-increment fetches */
for ( ; raw_l > 3; raw_l -= 4 )
{
tiff2ps.c: fix warning caused by integer promotion uint8 value is promoted to int in (value << 24) so -fsanitize yield runtime errors : tiff2ps.c:2969:33: runtime error: left shift of 246 by 24 places cannot be represented in type 'int'
2019-02-11 17:56:05 -05:00
val32 = (uint32)*(++raw_p) << 24;
val32 += (uint32)*(++raw_p) << 16;
val32 += (uint32)*(++raw_p) << 8;
val32 += (uint32)*(++raw_p);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
if ( val32 == 0 ) /* Special case */
{
ascii85_p[ascii85_l] = 'z';
rc = 1;
}
else
{
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[4] = (char) ((val32 % 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[3] = (char) ((val32 % 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[2] = (char) ((val32 % 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[1] = (char) ((val32 % 85) + 33);
ascii85[0] = (char) ((val32 / 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
_TIFFmemcpy( &ascii85_p[ascii85_l], ascii85, sizeof(ascii85) );
rc = sizeof(ascii85);
}
ascii85_l += rc;
if ( (A85BREAKCNTR -= rc) <= 0 )
{
ascii85_p[ascii85_l] = '\n';
++ascii85_l;
A85BREAKCNTR = A85BREAKLEN;
}
}
/*
* Output any straggler bytes:
*/
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
if ( raw_l > 0 )
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
{
fix 64bit type issues
2008-01-07 10:47:54 -05:00
tsize_t len; /* Output this many bytes */
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
len = raw_l + 1;
val32 = *++raw_p << 24; /* Prime the pump */
Fixed bug in Ascii85EncodeBlock() function: array index was out of range.
2003-05-07 04:12:30 -04:00
if ( --raw_l > 0 ) val32 += *(++raw_p) << 16;
if ( --raw_l > 0 ) val32 += *(++raw_p) << 8;
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[3] = (char) ((val32 % 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[2] = (char) ((val32 % 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
val32 /= 85;
Avoid warnings.
2004-06-04 10:16:14 -04:00
ascii85[1] = (char) ((val32 % 85) + 33);
ascii85[0] = (char) ((val32 / 85) + 33);
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
_TIFFmemcpy( &ascii85_p[ascii85_l], ascii85, len );
ascii85_l += len;
}
}
/*
* If requested add an ASCII85 End Of Data marker:
*/
if ( f_eod )
{
ascii85_p[ascii85_l++] = '~';
ascii85_p[ascii85_l++] = '>';
ascii85_p[ascii85_l++] = '\n';
}
return ( ascii85_l );
} /* Ascii85EncodeBlock() */
#endif /* EXP_ASCII85ENCODER */
Initial revision
1999-07-27 17:50:26 -04:00
char* stuff[] = {
"usage: tiff2ps [options] input.tif ...",
"where options are:",
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
" -1 generate PostScript Level 1 (default)",
" -2 generate PostScript Level 2",
" -3 generate PostScript Level 3",
" -8 disable use of ASCII85 encoding with PostScript Level 2/3",
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
" -a convert all directories in file (default is first), Not EPS",
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
" -b # set the bottom margin to # inches",
" -c center image (-b and -l still add to this)",
Sync tool usage printouts and man pages with reality.
2013-03-05 22:35:09 -05:00
" -C name set postscript document creator name",
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
" -d # set initial directory to # counting from zero",
Initial revision
1999-07-27 17:50:26 -04:00
" -D enable duplex printing (two pages per sheet of paper)",
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
" -e generate Encapsulated PostScript (EPS) (implies -z)",
More improvements from Richard Nolde.
2010-11-26 18:26:46 -05:00
" -h # set printed page height to # inches (no default)",
" -w # set printed page width to # inches (no default)",
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
" -H # split image if height is more than # inches",
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
" -W # split image if width is more than # inches",
New functionality for tiff2ps utility: splitting long images in several pages. See http://bugzilla.remotesensing.org/show_bug.cgi?id=142 for explanation. Patch granted by John Williams <williams@morinda.com>.
2002-06-21 06:24:40 -04:00
" -L # overLap split images by # inches",
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
" -i # enable/disable (Nz/0) pixel interpolation (default: enable)",
tools/tiff2ps.c, man/tiff2ps.1: Applied patch form Sebastian Eken (thanks, Sebastian!). New switches: -b # for a bottom margin of # inches -c center image -l # for a left margin of # inches -r rotate the image by 180 degrees New features merged with code for shrinking/overlapping. Previously added -c and -n switches (for overriding PS units) renamed in -x and -y respectively. http://bugzilla.remotesensing.org/show_bug.cgi?id=200
2002-10-07 08:49:32 -04:00
" -l # set the left margin to # inches",
Broad changes to tiff2ps by Bruce A. Mallett. Fixed PS_Lvl2page() code which outputs non-ASCII85 raw data. Moved test for when to output a line break to *after* the output of a character. This just serves to fix an eye-nuisance where the first line of raw data was one character shorter than subsequent lines. Added an experimental ASCII85 encoder which can be used only when there is a single buffer of bytes to be encoded. This version is much faster at encoding a straight-line buffer of data because it can avoid alot of the loop overhead of the byte-by-bye version. To use this version you need to define EXP_ASCII85ENCODER (experimental ...). Added bug fix given by Michael Schmidt to PS_Lvl2page() in which an end-of-data marker ('>') was not being output when producing non-ASCII85 encoded PostScript Level 2 data. Fixed PS_Lvl2colorspace() so that it no longer assumes that a TIFF having more than 2 planes is a CMYK. This routine no longer looks at the samples per pixel but instead looks at the "photometric" value. This change allows support of CMYK TIFFs. Modified the PostScript L2 imaging loop so as to test if the input stream is still open before attempting to do a flushfile on it. This was done because some RIPs close the stream after doing the image operation. Got rid of the realloc() being done inside a loop in the PSRawDataBW() routine. The code now walks through the byte-size array outside the loop to determine the largest size memory block that will be needed. Added "-m" switch to ask tiff2ps to, where possible, use the "imagemask" operator instead of the "image" operator. Added the "-i #" switch to allow interpolation to be disabled. Unrolled a loop or two to improve performance.
2001-03-28 20:34:33 -05:00
" -m use \"imagemask\" operator instead of \"image\"",
tiff2ps: enforce memory allocation limit fixes #153 / http://bugzilla.maptools.org/show_bug.cgi?id=2845
2020-03-21 10:50:38 -04:00
" -M size set the memory allocation limit in MiB. 0 to disable limit",
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
" -o # convert directory at file offset # bytes",
Initial revision
1999-07-27 17:50:26 -04:00
" -O file write PostScript to file instead of standard output",
Sync tool usage printouts and man pages with reality.
2013-03-05 22:35:09 -05:00
" -p generate regular (non-encapsulated) PostScript",
" -P L or P set optional PageOrientation DSC comment to Landscape or Portrait",
* tools/tiff2ps.c: Incorporated significant functionality update from Richard Nolde. In particular, support for rotating the image by 90, 180, 270, and 'auto' has been added. Still waiting for documentation patch to man/tiff2ps.1. * man/tiffcrop.1: Incorporated documentation updates from Richard Nolde. * tools/tiffcrop.c: Incorporated significant functionality update from Richard Nolde.
2009-01-11 15:04:22 -05:00
" -r # or auto rotate by 90, 180, 270 degrees or auto",
Initial revision
1999-07-27 17:50:26 -04:00
" -s generate PostScript for a single image",
See: http://bugzilla.maptools.org/show_bug.cgi?id=1983 * tools/tiff2ps.c: improvements and enhancements from Richard Nolde with additional command line options for Document Title, Document Creator, and Page Orientation
2010-09-26 01:39:31 -04:00
" -t name set postscript document title. Otherwise the filename is used",
Initial revision
1999-07-27 17:50:26 -04:00
" -T print pages for top edge binding",
Fixed help formatting.
2002-10-10 10:19:26 -04:00
" -x override resolution units as centimeters",
" -y override resolution units as inches",
Add ability to generate PS Level 3. It basically allows one to use the /flateDecode filter for ZIP compressed TIFF images. Patch supplied by Tom Kacvinsky. Fixes http://bugzilla.remotesensing.org/show_bug.cgi?id=328 Force deadzone printing when EPS output specified as per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=325
2003-04-29 05:52:16 -04:00
" -z enable printing in the deadzone (only for PostScript Level 2/3)",
Initial revision
1999-07-27 17:50:26 -04:00
NULL
};
static void
usage(int code)
{
char buf[BUFSIZ];
int i;
setbuf(stderr, buf);
Added library version reporting to all tools.
2003-03-12 09:05:05 -05:00
fprintf(stderr, "%s\n\n", TIFFGetVersion());
Initial revision
1999-07-27 17:50:26 -04:00
for (i = 0; stuff[i] != NULL; i++)
fprintf(stderr, "%s\n", stuff[i]);
exit(code);
}
Interpret the -w and -h options independently. As per bug http://bugzilla.remotesensing.org/show_bug.cgi?id=689
2005-01-12 07:51:26 -05:00
Reference in New Issue Copy Permalink