Table of contents

For other information, see the Ghostscript overview.


Ghostscript's capabilities in relation to PostScript

The Ghostscript interpreter, except as noted below, is intended to execute properly any source program written in the (LanguageLevel 3) PostScript language as defined in the PostScript Language Reference, Third Edition (ISBN 0-201-37922-8) published by Addison-Wesley in mid-1999. However, the interpreter is configurable in ways that can restrict it to various subsets of this language. Specifically, the base interpreter accepts the Level 1 subset of the PostScript language, as defined in the first edition of the PostScript Language Reference Manual (ISBN 0-201-10174-2) Addison-Wesley 1985, plus the file system, version 25.0 language, and miscellaneous additions listed in sections A.1.6, A.1.7, and A.1.8 of the Second Edition respectively, including allowing a string operand for the "status" operator. The base interpreter may be configured (see the documentation on building Ghostscript for how to configure it) by adding any combination of the following:

  • The ability to process PostScript Type 1 fonts. This facility is normally included in the interpreter.
  • The CMYK color extensions listed in section A.1.4 of the Second Edition (including colorimage). These facilities are available only if the color, dps, or level2 feature was selected when Ghostscript was built.
  • The Display PostScript extensions listed in section A.1.3 of the Second Edition, but excluding the operators listed in section A.1.2. These facilities are available only if the dps feature or the level2 feature was selected when Ghostscript was built.
  • The composite font extensions listed in section A.1.5 of the Second Edition, and the ability to handle Type 0 fonts. These facilities are available only if the compfont feature or the level2 feature was selected when Ghostscript was built.
  • The ability to load TrueType fonts and to handle PostScript Type 42 (encapsulated TrueType) fonts. These facilities are available only if the ttfont feature was selected when Ghostscript was built.
  • The PostScript Level 2 "filter" facilities except the DCTEncode and DCTDecode filters. These facilities are available only if the filter, dps, or level2 feature was selected when Ghostscript was built.
  • The PostScript Level 2 DCTEncode and DCTDecode filters. These facilities are available only if the dct or level2 feature was selected when Ghostscript was built.
  • All the other PostScript Level 2 operators and facilities listed in section A.1.1 of the Second Edition and not listed in any of the other A.1.n sections. These facilities are available only if the level2 feature was selected when Ghostscript was built.
  • All PostScript LanguageLevel 3 operators and facilities listed in the Third Edition, except as noted below. These facilities are available only if the psl3 feature was selected when Ghostscript was built.
  • The ability to recognize DOS EPSF files and process only the PostScript part, ignoring bitmap previews or other information. This facility is available only if the epsf feature was selected when Ghostscript was built.

Ghostscript currently does not implement the following PostScript LanguageLevel 3 facilities:

  • Settable ProcessColorModel for page devices, except for a very few special devices.
  • IODevices other than %stdin, %stdout, %stderr, %lineedit, %statementedit, %os%, and (if configured) %pipe% and %disk0% through %disk0%.

Ghostscript can also interpret files in the Portable Document Format (PDF) 1.7 format defined in the PDF Reference Version 1.7, distributed by Adobe Systems Incorporated, except as noted below. This facility can be disabled by deselecting the pdf feature when Ghostscript is built.

Ghostscript currently implements the majority of non-interactive features defined in the PDF reference.

Ghostscript also includes a number of additional operators defined below that are not in the PostScript language defined by Adobe.


Implementation limits

The implementation limits show here correspond to those in Tables B.1 and B.2 of the Second and Third Editions, which describe the quantities fully. Where Ghostscript's limits are different from those of Adobe's implementations (as shown in the Third Edition), Adobe's limits are also shown.

Architectural limits

Architectural limits (corresponds to Adobe table B.1)
Quantity    Limit    Type    Adobe
integer   32-bit   twos complement integer    
real   single-precision   IEEE float    
array   16777216   elements   65535
dictionary   16777215   elements   65535
string   16777216   characters   65535
name   16383   characters   127
filename   128*   characters    
save level   none   (capacity of memory)   15
gsave level   none   (capacity of memory)   13

* The limit on the length of a file name is 128 characters if the name starts with a %...% IODevice designation, or 124 characters if it does not.

Typical memory limits in LanguageLevel 1

Memory limits (corresponds to Adobe table B.2)
Quantity    Limit    Type    Adobe
userdict   200        
FontDirectory   100        
operand stack   800       500
dictionary stack   20        
execution stack   250    
interpreter level   none   (capacity of memory)   10
path   none   (capacity of memory)   1500
dash   11        
VM   none   (capacity of memory)   240000
file   none   (determined by operating system)   6
image   65535   values (samples × components)
for 1-, 2-, 4-, or 8-bit samples
  3300
    32767   values for 12-bit samples   3300

Other differences in VM consumption

In 32-bit builds packed array elements occupy either 2 bytes or 12 bytes. The average element size is probably about 7 bytes. Names occupy 16 bytes plus the space for the string.

In 64-bit builds packed array elements occupy either 2 bytes or 16 bytes. The average element size is probably about 9 bytes. Names occupy 24 bytes plus the space for the string.

The garbage collector doesn't reclaim portions of arrays obtained with getinterval, rather it collects entire arrays.


Additional operators in Ghostscript

Graphics and text operators

Transparency

NOTE: The following paragraphs describe non-standard operators for accessing the PDF 1.4 and later transparent imaging model through Postscript. If used incorrectly, they can have unexpected side effects and result in undefined behavior. As a result, these operators are disabled when SAFER is in force (as it is by default from version 9.50 onwards). To utilise these operators you will either have to disable SAFER (-dNOSAFER) or use the command line parameter -dALLOWPSTRANSPARENCY. The latter will make the custom operators available, but leave the file access controls active.

Ghostscript provides a set of operators for implementing the transparency and compositing facilities of PDF 1.4. These are defined only if the transpar option was selected when Ghostscript was built. We do not attempt to explain the underlying graphics model here: for details, see Adobe Technical Note #5407, "Transparency in PDF". Previously (in 9.52 and earlier), Ghostscript's model maintained separate alpha and mask values for opacity and shape. This model has been changed (as of 9.53) and instead Ghostscript maintains separate float values for stroke and fill alpha values with a boolean that indicates if these should be interpreted as shape or alpha values to be more in line with the PDF specification.

What follows is a subset of all the custom operators related to transparency, but covers the most useful, most common requirements.

Graphics state operators

Pushing the compositor device must be done before any other marking operations are made on the current page, and must be done per page. Popping the compositor should be done after the last marking operation of the page, and before the call to showpage. Any marking operations made after the compositor is popped will bypass the transparent imaging model, and may produce unexpected output.

<depth> .pushpdf14devicefilter -
Installs the transparency compositor device into the graphics state. At present the depth parameter should always be zero (Subject To Change.)
- .popdf14devicefilter -
Removes (or, more accuracately, disables) the transparency compositor in graphics state.
<modename> .setblendmode -
Sets the blending mode in the graphics state. If the mode name is not recognized, causes a rangecheck error. The initial value of the blending mode is /Compatible.
- .currentblendmode <modename>
Returns the graphics state blend mode on the stack.
[Deprecated as of 9.53] <0..1> .setopacityalpha -
Sets the opacity alpha value in the graphics state. The initial opacity alpha value is 1. Note, it is strongly suggested that this method not be used as it currently may give inconsistent results when mixed with methods that set stroke and fill alpha values.
[Deprecated as of 9.53] - .currentopacityalpha <0..1>
Returns the graphics state opacity alpha on the stack. Note, it is strongly suggested that this method not be used as it currently may give inconsistent results when mixed with methods that set stroke and fill alpha values.
[Deprecated as of 9.53] <0..1> .setshapealpha -
Sets the shape alpha value in the graphics state. The initial shape alpha value is 1. Note, it is strongly suggested that this method not be used as it currently may give inconsistent results when mixed with methods that set stroke and fill alpha values.
[Deprecated as of 9.53] - .currentshapealpha <0..1>
Returns the graphics state shape alpha on the stack. Note, it is strongly suggested that this method not be used as it currently may give inconsistent results when mixed with methods that set stroke and fill alpha values.
<0..1> .setstrokeconstantalpha -
Sets the stroke alpha value in the graphics state. The initial stroke alpha value is 1.
- .currentstrokeconstantalpha <0..1>
Returns the graphics state stroke alpha value on the stack.
<0..1> .setfillconstantalpha -
Sets the fill alpha value in the graphics state. The initial fill alpha value is 1.
- .currentfillconstantalpha <0..1>
Returns the graphics state fill alpha value on the stack.
<bool> .setalphaisshape -
If true, the values set by setstrokeconstantalpha and setfillconstantalpha are interpreted as shape values. The initial value of the AIS flag is false.
- .currentalphaisshape <0..1>
Returns the graphics state alpha is shape (AIS) on the stack.
<bool> .settextknockout -
Sets the text knockout flag in the graphics state. The initial value of the text knockout flag is true.
- .currenttextknockout <bool>
Returns the graphics state text knockout on the stack..
Rendering stack operators

The interpreter state is extended to include a (per-context) rendering stack for handling transparency groups and masks (generically, "layers"). Groups accumulate a full value for each pixel (paint plus transparency); masks accumulate only a coverage value. Layers must be properly nested, i.e., the 'end' or 'discard' operator must match the corresponding 'begin' operator.

Beginning and ending groups must nest properly with respect to save and restore: save and restore do not save and restore the layer stack. Currently, layers are not required to nest with respect to gsave and grestore, except that the device that is current in the graphics state when ending a layer must be the same as the device that was current when beginning the layer. THIS AREA IS SUBJECT TO CHANGE.

<paramdict> <llx> <lly> <urx> <ury> .begintransparencygroup -
Begins a new transparency group. The ll/ur coordinates are the bounding box of the group in the current user coordinate system. paramdict has the following keys:
/Isolated
(optional) Boolean; default value = false.
/Knockout
(optional) Boolean; default value = false.
- .endtransparencygroup -
Ends the current transparency group, compositing the group being ended onto the group that now becomes current.
<cs_set?> <paramdict> <llx> <lly> <urx> <ury> .begintransparencymaskgroup -
Begins a new transparency mask, which is represented as a group. The ll/ur coordinates are the bounding box of the mask in the current user coordinate system. paramdict has the following keys:
/Subtype
(required) Name, either /Alpha or /Luminosity.
/Background
(optional) Array of number.
/TransferFunction
(optional) Function object (produced by applying .buildfunction to a Function dictionary).
The cs_set parameter is a boolean indicating whether the color space for the mask group is the current color space in the graphics state, or whether mask group color space should be inherited from the previous group in the transparency group stack. In general, for the most consistent results, it is recommended that this be set to true, and the intended color space set in the graphics state prior to the .begintransparencymaskgroup call.
<mask#> .endtransparencymask -
Ends the current transparency mask group, compositing the mask group being ended and setting it as the current soft mask in the graphics state.

The mask# parameter indicates whether the mask should be treated as as opacity mask (0) or shape (1).

New ImageType

The transparency extension defines a new ImageType 103, similar to ImageType 3 with the following differences:

  • The required MaskDict is replaced by two optional dictionaries, OpacityMaskDict and ShapeMaskDict. If present, these dictionaries must have a BitsPerComponent entry, whose value may be greater than 1. Note that in contrast to ImageType 3, where any non-zero chunky mask value is equivalent to 1, ImageType 103 simply takes the low-order bits of chunky mask values.
  • A Matte entry may be present in one or both mask dictionaries, indicating premultiplication of the data values. If both MaskDicts have a Matte entry and the values of the two Matte entries are different, a rangecheck error occurs.
  • InterleaveType appears in the MaskDicts, not the DataDict, because each mask has its own InterleaveType. InterleaveType 2 (interlaced scan lines) is not supported.

Other graphics state operators

<int> .setoverprintmode -
Sets the overprint mode in the graphics state. Legal values are 0 or 1. Per the PDF 1.3 specification, if the overprint mode is 1, then when the current color space is DeviceCMYK, color components whose value is 0 do not write into the target, rather than writing a 0 value. THIS BEHAVIOR IS NOT IMPLEMENTED YET. The initial value of the overprint mode is 0.
- .currentoverprintmode <int>
Returns the current overprint mode.

Character operators

<font> <charcode> %Type1BuildChar -
This is not a new operator: rather, it is a name known specially to the interpreter. Whenever the interpreter needs to render a character (during a ...show, stringwidth, or charpath), it looks up the name BuildChar in the font dictionary to find a procedure to run. If it does not find this name, and if the FontType is 1, the interpreter instead uses the value (looked up on the dictionary stack in the usual way) of the name %Type1BuildChar.

The standard definition of %Type1BuildChar is in the initialization file gs_type1.ps. Users should not need to redefine %Type1BuildChar, except perhaps for tracing or debugging.

<font> <charname> %Type1BuildGlyph -
Provides the Type 1 implementation of BuildGlyph.

Other operators

Mathematical operators

<number> arccos <number>
Computes the arc cosine of a number between -1 and 1.
<number> arcsin <number>
Computes the arc sine of a number between -1 and 1.

Dictionary operators

mark <key1> <value1> <key2> <value2> ... .dicttomark <dict>
Creates and returns a dictionary with the given keys and values. This is the same as the PostScript Level 2 >> operator, but is available even in Level 1 configurations.
<dict> <key> .knownget <value> true
<dict> <key> .knownget false
Combines known and get in the obvious way.

Relational operators

<number|string> <number|string> max <number|string>
Returns the larger of two numbers or strings.
<number|string> <number|string> min <number|string>
Returns the smaller of two numbers or strings.

File operators

<file> .fileposition <integer> true
Returns the position of file. Unlike the standard fileposition operator, which causes an error if the file is not positionable, .fileposition works on all files, including filters: for non-positionable files, it returns the total number of bytes read or written since the file was opened.
<string> findlibfile <foundstring> <file> true
<string> findlibfile <string> false
Opens the file of the given name for reading, searching through directories as described in the usage documentation. If the search fails, findlibfile simply pushes false on the stack and returns, rather than causing an error.
<prefix_string|null> <access_string> .tempfile <string> <file>
Creates and opens a temporary file like the file operator, also returning the file name. There are three cases for the <prefix_string|null> operand:
  • null: create the file in the same directory and with the same name conventions as other temporary files created by the Ghostscript implementation on this platform. E.g., the temporary file might be named /tmp/gs_a1234.
  • A string that contains only alphanumeric characters, underline, and dash: create the file in the standard temporary directory, but use the <prefix_string> as the first part of the file name. E.g., if <prefix_string> is xx, the temporary file might be named /tmp/xxa1234.
  • A string that is the beginning of an absolute file name: use the <prefix_string> as the first part of the file name. E.g., if <prefix_string> is /my/tmpdir/zz, the temporary file might be named /my/tmpdir/zza1234.
  • When running in SAFER mode, the absolute path must be one of the strings on the permit file writing list (see -dSAFER) .

Ghostscript also supports the following IODevice in addition to a subset of those defined in the Adobe documentation:

  • %pipe%command, which opens a pipe on the given command. This is supported only on operating systems that provide popen (primarily Unix systems, and not all of those).
  • %disk#%, which emulates the %disk0 through %disk9 devices on some Adobe PostScript printers. This pseudo device provides a flat filenaming system with a user definable location for the files (/Root). These devices will only be present if the diskn.dev feature is specified during the build.
  • This feature is intended to allow compatibility with font downloaders that expect to store fonts on the %disk device of the printer.

    Use of the %disk#% devices requires that the location of files be given by the user setting the /Root device parameter. The syntax for setting the /Root parameter is:

        mark /Root (directory_specification) (%disk#) .putdevparams
    
    For example, to store the files of the %disk0 device on the directory /tmp/disk0, use:
        mark /Root (/tmp/disk0/) (%disk0) .putdevparams
    

    The files will be stored in the specified directory with arbitrary names. A mapping file is used to store the association between the file names given for the file operations on the %diskn# device and the file that resides in the /Root directory.

Miscellaneous operators

<array> bind <array>
Depending on the command line parameters bind is redefined as:
Flag     Definition
DELAYBIND   returns the argument, stores the argument for later use by .bindnow
<array> .bind <array>
Performs standard bind operation as defined in PLRM regardless of the DELAYBIND flag.
- .bindnow -
Applies bind operator to all saved procedures after binding has been deferred through -dDELAYBIND. Note that idiom recognition has no effect for the deferred binding because the value returned from bind is discarded.

Since v. 8.12 .bindnow undefines itself and restores standard definition of bind operator. In earlier versions after calling .bindnow, the postscript bind operator needs to be rebound to the internal implementation .bind, as in this fragment from the ps2ascii script:

DELAYBIND {
  .bindnow
  /bind /.bind load def
} if
This is necessary for correct behavior with later code that uses the bind operator.

<string> getenv <string> true
<string> getenv false
Looks up a name in the shell environment. If the name is found, returns the corresponding value and true; if the name is not found, returns false.
<string> <boolean> .setdebug -
Sets or clears any subset of the debugging flags included in <string> based on the value of <boolean>. These correspond to the debug flags set by -Z on the command line and enable debug and tracing output from various internal modules.

Note that most tracing output is only produced if the Ghostscript interpreter was built with the DEBUG preprocessor symbol defined.

The zsetdebug() C function, which implements this operator, is a useful breakpoint for debuggers. Inserting '() true .setdebug' in the interpreted code will trigger a breakpoint at that location without side effects. The current flag state is available in C as the gs_debug[] array, indexed by character value. The zsetdebug function will be entered, and gs_debug[] updated, whether or not Ghostscript is built with the DEBUG preprocessor symbol defined, so this is useful even with release builds.

- .setsafe -
If Ghostscript is started with -dNOSAFER or -dDELAYSAFER, this operator can be used to enter SAFER mode (see -dSAFER)

The following is deprecated, see -dSAFER

Since SAFER mode is implemented with userparameters and device parameters, it is possible to use save and restore before and after .setsafe to return to NOSAFER mode, but note that such a save object is accessible to any procedures or file run in SAFER mode. A malicious file with an unbalanced restore could potentially restore back to a point where SAFER was not in operation.

Note: This uses setpagedevice to change .LockSafetyParams, so the page will be erased as a side effect of this operator

- .locksafe -

The following is deprecated, see -dSAFER

This operator sets the current device's .LockSafetyParams and the LockFilePermissions userparameter true as well as adding the paths on LIBPATH and FONTPATH and the paths given by the system params /GenericResourceDir and /FontResourceDir to the current PermitFileReading list of paths.

If Ghostscript is started with -dNOSAFER or -dDELAYSAFER, this operator can be used to enter SAFER mode with the current set of PermitFile... user parameters in effect. Since .setsafe sets the PermitFile... user parameters to empty arrays, a script or job server that needs to enable certain paths for file Reading, Writing and/or Control can use this operator to perform the locking needed to enter SAFER mode.

For example, to enable reading everywhere, but disallow writing and file control (deleting and renaming files), the following can be used:

    { << /PermitFileReading [ (*) ]
         /PermitFileWriting [ ]
         /PermitFileControl [ ]
      >> setuserparams
      .locksafe
    } stopped pop

In the above example, use of stopped will allow the use of this sequence on older versions of Ghostscript where .locksafe was not an operator.

Note: This uses setpagedevice to change .LockSafetyParams, so the page will be erased as a side effect of this operator

See also .LockSafetyParams and User Parameters.

<name> <string> .addcontrolpath
Adds a single path to the file access control lists.

The <name> parameter can be one of:

  • /PermitFileReading

  • /PermitFileWriting

  • /PermitFileControl

Whilst the string parameter is the path to be added to the requested list.

NOTE: Any attempt to call this operator after .activatepathcontrol has been called will result in a Fatal error, and the interpreter will immediately exit.

.activatepathcontrol
Activates file access controls. Once activated, these access controls remain in place until the interpreter shuts down.
.currentpathcontrolstate
Returns true on the operand stack if file access control has been activated, false if not.
<dict> .genordered <dict> (default: /OutputType /Type3).
<dict> .genordered <string> (/OutputType /ThreshString).
<dict> .genordered <array> (/OutputType /TOSArray).
This operator creates an ordered dither screening pattern with the parameters from the dictionary, returning (by default) a PostScript HalftoneType 3 (threshold array based) dictionary suitable for use with sethalftone or as a component Halftone of a HalftoneType 5 Halftone dictionary. The /OutputType parameter can also select other than Halftone Type 3 as the return paramter, <dict> has the following keys (all are optional):
/Frequency
Integer; default value = 75
/Angle
Integer; default value = 0
/HResolution
Real or Integer; default value is device X resolution.
/VResolution
Real or Integer; default value is device Y resolution.
/DotShape
Integer; default value = 0 (CIRCLE). Other shapes available are:
1=REDBOOK, 2=INVERTED, 3=RHOMBOID, 4=LINE_X, 5=LINE_Y, 6=DIAMOND1, 7=DIAMOND2, 8=ROUNDSPOT,
/SuperCellSize
Integer; default value = 1 -- actual cell size determined by Frequency, Angle, H/V Resolution.
A larger value will allow more levels to be attained.
/Levels
Integer; default value = 1 -- actual number of gray levels is determined by Frequency and H/V Resolution.
SuperCellSize may need to be specified large enough to achieve the requested number of gray levels.
/OutputType
Name; default value = /Type3 (HalftoneType 3 dictionary). Other shapes available are:
/ThreshString
First two bytes are width (high byte first), next two bytes are height, followed by the threshold array bytes (same as /Thresholds of the Type3 dictionary).
/TOSArray
First element is the width, next is the height, followed by pairs X, then Y, of the turn-on-sequence of the threshold array. This information can be used to construct a threshold array with a transfer function "pickled into" the threshold array, which is useful if the turn-on-sequence has more than 256 pairs. Refer to toolbin/halftone/thresh_remap for more information.
.shellarguments

This operator is used to access the ARGUMENTS command line option.

Relies on Ghostscript being called with the "--" command line option - see Input Control

See examples in lib for more information.

Device operators

<device> copydevice <device>
Copies a device. The copy is writable and installable. The copy is created in the current VM (local or global), usually local VM for executing ordinary PostScript files.
<devicename> finddevice <device>
Creates a default instance of a device specified by name. The instance is created in global VM. If finddevice is called more than once with the same device name, it creates the default instance the first time, and returns the same instance thereafter.
<devicename> findprotodevice <device>
Finds the prototype of a device specified by name. A prototype can be used with getdeviceprops or other parameter-reading operators, but it is read-only and cannot be set with setdevice: it must be copied first.
<matrix> <width> <height> <palette> makeimagedevice <device>
Makes a new device that accumulates an image in memory. matrix is the initial transformation matrix: it must be orthogonal (that is, [a 0 0 b x y] or [0 a b 0 x y]). palette is a string of 2^N or 3 × 2^N elements, specifying how the 2^N possible pixel values will be interpreted. Each element is interpreted as a gray value, or as RGB values, multiplied by 255. For example, if you want a monochrome image for which 0=white and 1=black, the palette should be <ff 00>; if you want a 3-bit deep image with just the primary colors and their complements (ignoring the fact that 3-bit images are not supported), the palette might be <000000 0000ff 00ff00 00ffff ff0000 ff00ff ffff00 ffffff>. At present, the palette must contain exactly 2, 4, 16, or 256 entries, and must contain an entry for black and an entry for white; if it contains any entries that aren't black, white, or gray, it must contain at least the six primary colors (red, green, blue, and their complements cyan, magenta, and yellow); aside from this, its contents are arbitrary.

Alternatively, palette can be 16, 24, 32, or null (equivalent to 24). These are interpreted as:

Palette     Bits allocated per color
16   5 red, 6 green, 5 blue
24   8 red, 8 green, 8 blue
32   8C, 8M, 8Y, 8K

Note that one can also make an image device (with the same palette as an existing image device) by copying a device using the copydevice operator.

<device> <index> <string> copyscanlines <substring>
Copies one or more scan lines from an image device into a string, starting at a given scan line in the image. The data is in the same format as for the image operator. It is an error if the device is not an image device or if the string is too small to hold at least one complete scan line. Always copies an integral number of scan lines.
<device> setdevice -

Sets the current device to the specified device. Also resets the transformation and clipping path to the initial values for the device. Signals an invalidaccess error if the device is a prototype or if .LockSafetyParams is true for the current device.

Some device properties may need to be set with putdeviceprops before setdevice is called. For example, the pdfwrite device will try to open its output file, causing an undefinedfilename error if OutputFile hasn't been set to a valid filename. Another method in such cases is to use the level 2 operator instead: << /OutputDevice /pdfwrite /OutputFile (MyPDF.pdf) >> setpagedevice.

- currentdevice <device>
Gets the current device from the graphics state.
<device> getdeviceprops <mark> <name1> <value1> ... <namen> <valuen>
Gets the properties of a device. See the section on device parameters below for details.
<mark> <name1> <value1> ... <namen> <valuen> <device> putdeviceprops <device>
Sets properties of a device. May cause undefined, invalidaccess, typecheck, rangecheck, or limitcheck errors.

Filters

Standard filters

In its usual configuration, Ghostscript supports all the standard PostScript LanguageLevel 3 filters, both encoding and decoding, except that it does not currently support:

  • the EarlyChange key in the LZWEncode filter.

Ghostscript also supports additional keys in the optional dictionary operands for some filters. For the LZWDecode filter:

InitialCodeLength <integer> (default 8)
An integer between 2 and 11 specifying the initial number of data bits per code. Note that the actual initial code length is 1 greater than this, to allow for the reset and end-of-data code values.
FirstBitLowOrder <boolean> (default false)
If true, codes appear with their low-order bit first.
BlockData <boolean> (default false)
If true, the data is broken into blocks in the manner specified for the GIF file format.

For the CCITTFaxEncode and CCITTFaxDecode filters:

DecodedByteAlign <integer> (default 1)
An integer N with the value 1, 2, 4, 8, or 16, specifying that decoded data scan lines are always a multiple of N bytes. The encoding filter skips data in each scan line from Columns to the next multiple of N bytes; the decoding filter pads each scan line to a multiple of N bytes.

Non-standard filters

In addition to the standard PostScript LanguageLevel 3 filters, Ghostscript supports the following non-standard filters. Many of these filters are used internally to implement standard filters or facilities; they are almost certain to remain, in their present form or a backward-compatible one, in future Ghostscript releases.

<target> /BCPEncode filter <file>
<source> /BCPDecode filter <file>
Create filters that implement the Adobe Binary Communications Protocol. See Adobe documentation for details.
<target> <seed_integer> /eexecEncode filter <file>
Creates a filter for encrypting data into the encrypted format described in the Adobe Type 1 Font Format documentation. The seed_integer must be 55665 for the eexec section of a font, or 4330 for a CharString. Note that for the eexec section of a font, this filter produces binary output and does not include the initial 4 (or lenIV) garbage bytes.
<source> <seed_integer> /eexecDecode filter <file>
<source> <dict> /eexecDecode filter <file>
Creates a filter for decrypting data encrypted as described in the Adobe Type 1 Font Format documentation. The seed_integer must be 55665 or 4330 as described just above. PDF interpreters don't skip space characters after operator eexec. Use keep_spaces = true for decoding embedded PDF fonts. Recognized dictionary keys are:
seed <16-bit integer> (required)
lenIV <non-negative integer> (default=4)
eexec <bool> (default=false)
keep_spaces <bool> (default=false)
<target> /MD5Encode filter <file>
Creates a filter that produces the 16-byte MD5 digest of the input. Note that no output is produced until the filter is closed.
<source> <hex_boolean> /PFBDecode filter <file>
Creates a filter that decodes data in .PFB format, the usual semi-binary representation for Type 1 font files on IBM PC and compatible systems. If hex_boolean is true, binary packets are converted to hex; if false, binary packets are not converted.
<target> <dict> /PixelDifferenceEncode filter <file>
<source> <dict> /PixelDifferenceDecode filter <file>
Implements the Predictor=2 pixel-differencing option of the LZW filters. Recognized keys are:
Colors <integer> (1 to 4, default=1)
BitsPerComponent <integer> (1, 2, 4, or 8, default=8)
Columns <integer> (>= 0, required)

See the Adobe PDF Reference Manual for details.

<target> <dict> /PNGPredictorEncode filter <file>
<source> <dict> /PNGPredictorDecode filter <file>

Implements the "filter" algorithms of the Portable Network Graphics (PNG) graphics format. Recognized keys are:

Keys recognized in PNG filter algorithms
Key    Range    Default
Colors <integer>     1 to 16    16
BitsPerComponent <integer>   1, 2, 4, 8, or 16    8
Columns <integer>   >= 0    1
Predictor <integer>   10 to 15    15

The Predictor is the PNG algorithm number + 10 for the Encoding filter; the Decoding filter ignores Predictor. 15 means the encoder attempts to optimize the choice of algorithm. For more details see the PNG specification

http://www.w3.org/TR/WD-png-960128.html
<target> /TBCPEncode filter <file>
<source> /TBCPDecode filter <file>
Create filters that implement the Adobe Tagged Binary Communications Protocol. See Adobe documentation for details.
<target> /zlibEncode filter <file>
<source> /zlibDecode filter <file>
Creates filters that use the data compression method variously known as 'zlib' (the name of a popular library that implements it), 'Deflate' (as in RFC 1951, which is a detailed specification for the method), 'gzip' (the name of a popular compression application that uses it), or 'Flate' (Adobe's name). Note that the PostScript Flate filters are actually a combination of this filter with an optional predictor filter.

Unstable filters

Some versions of Ghostscript may also support other non-standard filters for experimental purposes. The current version includes the following such filters, which are not documented further. No code should assume that these filters will exist in compatible form, or at all, in future versions.

<target/source> <string> ByteTranslateEncode/Decode filter <file>
string must be a string of exactly 256 bytes. Creates a filter that converts each input byte b to string[b]. Note that the Encode and Decode filters operate identically: the client must provide a string for the Decode filter that is the inverse mapping of the string for the Encode filter.
<target/source> <dict> BoundedHuffmanEncode/Decode filter <file>
These filters encode and decode data using Huffman codes. Since these filters aren't used anywhere, we don't document them further, except to note the recognized dictionary keys, which must be set identically for encoding and decoding:
FirstBitLowOrder <bool> (default=false)
MaxCodeLength <int> (default=16)
EndOfData <bool> (default=true)
EncodeZeroRuns <int> (default=256)
Tables <int_array>
<target/source> <dict> BWBlockSortEncode/Decode filter <file>
This filter implements the Burroughs-Wheeler block sorting compression method, which we've heard is also used in the popular bzip2 compression application. See http://sources.redhat.com/bzip2/ for more information. The only recognized dictionary key is:
BlockSize <integer> (default=16384)

Device parameters

Ghostscript supports the concept of device parameters for all devices, not just page devices. (For non-page devices, these are accessible through getdeviceprops and putdeviceprops, as indicated above.) Here are the currently defined parameters for all devices:
.LockSafetyParams <boolean>
This parameter allows for improved system security by preventing PostScript programs from being able to change potentially dangerous device paramters such as OutputFile. This parameter cannot be set false if it is already true.

If this parameter is true for the current device, attempt to set a new device that has .LockSafetyParams false will signal an invalidaccess error.

BitsPerPixel <integer> (usually read-only)
Number of bits per pixel.
.HWMargins [<four floats>]
Size of non-imageable regions around the edges of the page, in points (units of 1/72in; see the notes on measurements in the documentation on devices).
HWSize [<integer> <integer>]
X and Y size in pixels.
%MediaSource <integer>
The input tray key as determined by setpagedevice. PostScript language programs don't set this parameter directly; they can request a particular tray through the MediaPosition setpagedevice parameter, but the setpagedevice logic need not necessarily honor the request. Devices which support switchable trays should implement %MediaSource in their put_params device procedure, but (unlike most other such parameters) need not implement corresponding reading logic in get_params.
%MediaDestination <integer>
The output tray key as determined by setpagedevice. Handling by devices should be parallel to %MediaSource.
.IgnoreNumCopies <boolean>
Some page description languages support a NumCopies parameter. This parameter instructs the device to ignore this, producing only one copy of the document on output. Note that some devices ignore NumCopies regardless because of limitation of the output format or the implementation.
Name <string> (read-only)
The device name. Currently the same as OutputDevice.
Colors, GrayValues, RedValues, GreenValues, BlueValues, ColorValues (usually read-only)
As for the deviceinfo operator of Display PostScript. Red, Green, Blue, and ColorValues are only defined if Colors > 1.
TextAlphaBits, GraphicsAlphaBits (usually read-only)
The number of bits of anti-aliasing information for text or graphics respectively. Legal values are 1 (no anti-aliasing, the default for most devices), 2, or 4.

Because this feature relies upon rendering the input it is incompatible, and will generate an error on attempted use, with any of the vector output devices.

Ghostscript also supports the following read-only parameter that is not a true device parameter:

.EmbedFontObjects <integer>
If non-zero, indicates that the device may embed font objects (as opposed to bitmaps for individual characters) in the output. The purpose of this parameter is to disable third-party font renderers for such devices. (This is zero for almost all devices.)

In addition, the following are defined per Adobe's documentation for the setpagedevice operator:

Duplex (if supported)
HWResolution
ImagingBBox
Margins
LeadingEdge
MediaPosition
NumCopies (for printers only)
Orientation (if supported)
OutputDevice
PageOffset (write-only)
PageSize
ProcessColorModel (usually read-only)

Some devices may only allow certain values for HWResolution and PageSize. The null device ignores attempts to set PageSize; its size is always [0 0].

It should be noted that calling setpagedevice with one of the above keys may reset the effects of any pdfmark commands up to that point. In particular this is true of HWResolution, a behavior that differs from Adobe Distiller.

For raster printers and image format (jpeg*, tiff*, png* ...) devices these page device parameters are also defined:

MaxBitmap <integer>
Maximum space for a full page raster image (bitmap) in memory.

This value includes the space for padding raster lines and for an array of pointers for each raster line, thus the MaxBitmap value to allow a given PageSize of a specific number of bits per pixel to be rendered in a full page buffer may be somewhat larger than the bitmap size alone.

BandListStorage <file|memory>
The default is determined by the make file macro BAND_LIST_STORAGE. Since memory is always included, specifying -sBandListStorage=memory when the default is file will use memory based storage for the band list of the page. This is primarily intended for testing, but if the disk I/O is slow, band list storage in memory may be faster.
BufferSpace <integer>
Size of the buffer space for band lists, if the full page raster image (bitmap) is larger than MaxBitmap (see above.)

The buffer space is used to collect display list (clist) commands for the bands and then to consolidate those commands when writing the clist to the selected BAND_LIST_STORAGE device (memory or file) set when Ghostscript is compiled.

If MaxBitmap (above) forces banding mode, and if BufferSpace is large enough, the display list (clist) will consist of a single band.

The BufferSpace will determine the size of the 'consolidation' buffer (above) even if the MaxBitmap value is low enough to force banding/clist mode.

BGPrint <boolean>
With many printer devices, when the display list (clist) banding mode is being used, the page rendering and output can be performed in a background thread. The default value, false, causes the rendering and printing to be done in the same thread as the parser. When -dBGPrint=true, the page output will be overlapped with parsing and writing the clist for the next page.

If the device does not support background printing, rendering and printing will be performed as if -dBGPrint=false.

Note that the background printing thread will allocate a band buffer (size determined by the BufferSpace or BandBufferSpace values) in addition to the band buffer in the 'main' parsing thread.

If NumRenderingThreads is > 0, then the background printing thread will use the specified number of rendering threads as children of the background printing thread. The background printing thread will perform any processing of the raster data delivered by the rendering threads. Note that BGPrint is disabled for vector devices such as pdfwrite and NumRenderingThreads has no effect on these devices eitehr.

GrayDetection <boolean>
When true, and when the display list (clist) banding mode is being used, during writing of the clist, the color processing logic collects information about the colors used before the device color profile is applied. This allows special devices that examine dev->icc_struct->pageneutralcolor with the information that all colors on the page are near neutral, i.e. monochrome, and converting the rendered raster to gray may be used to reduce the use of color toners/inks.

Since the determination of whether or not the page uses colors is determined before the conversion to device colors, this information is independent of the device output profile. The determination has a small delta (DEV_NEUTRAL and AB_NEUTRAL in base/gscms.h) to allow colors close to neutral to be detected as neutral. Changing this value requires rebuilding.

Among the devices distributed with the source, currently only the pnmcmyk device supports this parameter and will produce either a P7 PAM CMYK output or a P5 PGM Gray output depending on the use of color on the page.

Also, the 'pageneutralcolor' status can be interrogated as a device parameter of the same name. Using PostScript there are several methods:

    currentpagedevice /pageneutralcolor get

    mark currentdevice getdeviceprops .dicttomark /pageneutralcolor get

    /pageneutralcolor /GetDeviceParam .special_op { exch pop }{ //false } ifelse

Note that the pageneutralcolor state is reset to false after the page is output, so this parameter is only valid immediately before showpage is executed, although the setpagedevice EndPage procedure can be used to check the state just prior to the actual output of the page that resets pagenuetralcolor. For example:

    << /EndPage {
      exch pop 2 ne dup {
        currentpagedevice /pageneutralcolor get (pageneutralcolor: ) print = flush
      } if
    }
    >> setpagedevice
Notes:

Since -dGrayDetection=true requires extra checking during writing of the clist, this option should only be used for devices that support the optimization of pages to monochrome, otherwise performance may be degraded for no benefit.

Since GrayDetection=true is only effective when in clist (banding) mode, it is recommended to also force banding. For example: -dGrayDetection=true -dMaxBitmap=0

NumRenderingThreads <integer>
When the display list (clist) banding mode is being used, bands can be rendered in separate threads. The default value, 0, causes the rendering of bands to be done in the same thread as the parser and device driver. NumRenderingThreads of 1 or higher results in bands rendering in the specified number of 'background' threads.

The number of threads should generally be set to the number of available processor cores for best throughput.

Note that each thread will allocate a band buffer (size determined by the BufferSpace or BandBufferSpace values) in addition to the band buffer in the 'main' thread.

Additoinally note that ths parameter has no effect with devices which do not generally render to a bitmap output, such as the vector devices (eg pdfwrite) and has no effect when rendering, but not using a clist. See Improving_performance

OutputFile <string>
An empty string means "send to printer directly", otherwise specifies the file name for output; %d is replaced by the page number for page-oriented output devices; on Unix systems %pipe%command writes to a pipe. (|command also writes to a pipe, but is now deprecated). Also see the -o parameter.

Attempts to set this parameter if .LockSafetyParams is true will signal an invalidaccess error.

OpenOutputFile <boolean>
If true, open the device's output file when the device is opened, rather than waiting until the first page is ready to print.
PageCount <integer> (read-only)
Counts the number of pages printed on the device.

The following parameters are for use only by very specialized applications that separate band construction from band rasterization. Improper use may cause unpredictable errors. In particular, if you only want to allocate more memory for banding, to increase band size and improve performance, use the BufferSpace parameter, not BandBufferSpace.

BandHeight <integer>
The height of bands when banding. 0 means use the largest band height that will fit within the BandBufferSpace (or BufferSpace, if BandBufferSpace is not specified). If BandHeight is larger than the number of lines that will fit in the buffer, opening the device will fail.
BandWidth <integer>
The width of bands in the rasterizing pass, in pixels. 0 means use the actual page width. A BandWidth value smaller than the width of the page will be ignored, and the actual page width will be used instead.
BandBufferSpace <integer>
The size of the band buffer in the rasterizing pass, in bytes. 0 means use the same buffer size as for the interpretation pass.

Ghostscript supports the following parameter for setpagedevice and currentpagedevice that is not a device parameter per se:

ViewerPreProcess <procedure>
Specifies a procedure to be applied to the page device dictionary before any other processing is done. The procedure may not alter the dictionary, but it may return a modified copy. This "hook" is provided for use by viewing programs such as GSview.

User parameters

Ghostscript supports the following non-standard user parameters:
ProcessDSCComment <procedure|null>
If not null, this procedure is called whenever the scanner detects a DSC comment (comment beginning with %% or %!). There are two operands, the file and the comment (minus any terminating EOL), which the procedure must consume.
ProcessComment <procedure|null>
If not null, this procedure is called whenever the scanner detects a comment (or, if ProcessDSCComment is also not null, a comment other than a DSC comment). The operands are the same as for ProcessDSCComment.
LockFilePermissions <boolean>
If true, this parameter and the three PermitFile... parameters cannot be changed. Attempts to change any of the values when LockFilePermissions is true will signal invalidaccess. Also, when this value is true, the file operator will give invalidaccess when attempting to open files (processes) using the %pipe device.

Also when LockFilePermissions is true, strings cannot reference the parent directory (platform specific). For example (../../xyz) is illegal on unix, Windows and Macintosh, and ([.#.#.XYZ]) is illegal on VMS.

This parameter is set true by the .setsafe and .locksafe operators.

PermitFileReading <array of strings>
PermitFileWriting <array of strings>
PermitFileControl <array of strings>
These parameters specify paths where file reading, writing and the 'control' operations are permitted, respectively. File control operations are deletefile and renamefile. For renamefile, the filename for the current filename must match one of the paths on the PermitFileControl list, and the new filename must be on both the PermitFileControl and the PermitFileWriting lists of paths.

The strings can contain wildcard characters as for the filenameforall operator and unless specifying a single file, will end with a * for directories (folders) to allow access to all files and sub-directories in that directory.

Note: The strings are used for stringmatch operations similar to filenameforall, thus on MS Windows platforms, use the '/' character to separate directories and filenames or use '\\\\' to have the string contain '\\' which will match a single '\' in the target filename (use of '/' is strongly recommended).

The SAFER mode and the .setsafe operator set all three lists to empty arrays, thus the only files that can be read are the %stdin device and on LIBPATH or FONTPATH or the Resource paths specified by the /FontResourceDir or /GenericResourceDir system params. Files cannot be opened for writing anywhere and cannot be deleted or renamed except for files created with the .tempfile operator).

Note: Limiting file reading as above is NOT compatible with SAFER mode in release versions before 7.11 and corresponds to the use of -dPARANOIDSAFER in version 7.04 (up to and not including version 7.10) and GPL versions 6.53 (up to and not including 6.60).

AlignToPixels <integer>
Control sub-pixel positioning of character glyphs (where applicable). A value of 1 specifies alignment of text characters to pixels boundaries. A value of 0 to subpixels where the division factor is set by the device parameter TextAlphaBits. If the latter is 1, the same rendering results regardless of the value of AlignToPixels. The initial value defaults to 1, but this may be overridden by the command line argument -dAlignToPixels.
GridFitTT <integer>
Control the use of True Type grid fitting. Ghostscript, by default, uses Freetype for rendering Truetype (and most other) glyphs (but other scaler/renderer libraries can be used), thus has access to a complete Truetype bytecode interpreter.

This parameter controls the hinting of Truetype glyphs.

  • A value of 0 disables grid fitting for all True Type fonts (not generally recommended).
  • A value of 1 enables the grid fitting using the native Truetype hinting bytecode program(s). Fonts or glyphs with faulty bytecode program(s) will be rendered unhinted.
  • A value 2 is scaler/renderer dependent (generally, if no alternative hinting engine is available this will be equivalent to 1). With the Freetype (our default) this enables Freetype's built-in autohinter.
  • With Freetype, a value of 3 is effectively equivalent to 1.

This parameter defaults to 1, but this may be overridden on the command line with -dGridFitTT=n.


Miscellaneous additions

Extended semantics of 'run'

The operator run can take either a string or a file as its argument. In the latter case, it just runs the file, closing it at the end, and trapping errors just as for the string case.

Decoding resources

Decoding is a Ghostscript-specific resource category. It contains various resources for emulating PostScript fonts with other font technologies. Instances of the Decoding category are tables which map PostScript glyph names to character codes used with TrueType, Intellifont, Microtype and other font formats.

Currently Ghostscript is capable of PostScript font emulation in 2 ways :

  • 1. Through FAPI plugins, and
  • 2. With TrueType font files, using the native font renderer, by specifying TrueType font names or files in Resource/Init/Fontmap.GS.

Decoding resources are not currently used by the native font renderer.

An instance of the Decoding resource category is a dictionary. The dictionary keys are PostScript glyph names and the values are either character codes, or arrays of character codes. Arrays are used when a single name may be mapped to various character codes - in this case Ghostscript tries all alternatives until a success. The name of the resource instance should reflect the character set for which it maps. For example, /Unicode /Decoding resource maps to Unicode UTF-16.

The rules for using Decoding resources in particular cases are specified in the configuration file Resource/Init/xlatmap. See the file itself for more information.

The file format for Decoding resource files is generic PostScript. Users may want to define custom Decoding resources. The ParseDecoding procset defined in Resource/Init/gs_ciddc.ps allows representation of the table in a comfortable form.

CIDDecoding resources

CIDDecoding resources are similar to Decoding resources, except they map Character Identifiers (CIDs) rather than glyph names. Another difference is that the native Ghostscript font renderer uses CIDDecoding resources while emulate CID fonts with TrueType or OpenType fonts.

An instance of the CIDDecoding resource category is a dictionary of arrays. Keys in the dictionary are integers, which correspond to high order byte of a CID. Values are 256-element arrays, and their indices correspond to the low order byte of a CID. Each elemet of an array is either null, or character code (integer), or an array of character codes (integers). The zero code represents mapping to the default character.

The dictionary includes the additional key CIDCount. Its value is the maximal CID defined, plus one.

The Ghostscript library is capable of generating some CIDDecoding instances automatically, using the appropriate CMap (character map) resources. This covers most of practical cases if the neccessary CMap resources are provided. See the table .CMapChooser in Resource/Init/gs_ciddc.ps for the names of automatically gerenated resources and associated CMaps. They allow to mapping CNS1, GB1, Japan1, Japan2 and Korea1 CID sets to TrueType character sets known as Unicode (exactly UTF-16), Big5, GB1213, ShiftJIS, Johab and Wansung.

The file format for CIDDecoding resource file is generic PostScript. Users may want to define custom resources to CIDDecoding resource category.

GlyphNames2Unicode

GlyphNames2Unicode is an undocumented dictionary which Adobe PostScript printer driver uses to communicate with Adobe Distiller. In this dictionary the keys are glyph names, the values are Unicode UTF-16 codes for them. The dictionaly is stored in the FontInfo dictionary under the key GlyphNames2Unicode. Ghostscript recognises it and uses to generate ToUnicode CMaps with pdfwrite.

Multiple Resource directories

Since 8.10 release Ghostscript maintains multiple resource directories.

Ghostscript does not distinguish lib and Resource directories. There is no file name conflicts because lib does not contain subdirectories, but Resource always store files in subdirectories.

The search method with multiple resource directories appears not fully conforming to PLRM. We cannot unconditionally call ResourceFileName while executing findresource or resourcestatus, resourceforall, because per PLRM it always returns a single path. Therefore Ghostscript implements an extended search method in findresource, resourcestatus and resourceforall, which first calls ResourceFileName and checks whether the returned path points to an existing file. If yes, the file is used, othervise Ghostscript searches all directories specified in LIB_PATH. With a single resource directory it appears conforming to PLRM and equivalent to Adobe implementations.

ResourceFileName may be used for obtaining a path where a resource file to be installed. In this case Ghostscript to be invoked with -sGenericResourceDir=path, specifying an absolute path. The default value for GenericResourceDir is a relative path. Therefore a default invocation with a PostScript installer will install resource files into /gs/Resource.


Copyright © 2000-2021 Artifex Software, Inc. All rights reserved.

This software is provided AS-IS with no warranty, either express or implied. This software is distributed under license and may not be copied, modified or distributed except as expressly authorized under the terms of that license. Refer to licensing information at https://www.artifex.com or contact Artifex Software, Inc., 1305 Grant Avenue - Suite 200, Novato, CA 94945, U.S.A., +1(415)492-9861, for further information.

Ghostscript version 9.54.0, 30 March 2021