22

✏️ Editing: pdf_base.ps

% Copyright (C) 2001-2019 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 the license contained in the file LICENSE in this distribution.
%
% Refer to licensing information at http://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.
%

% pdf_base.ps
% Basic parser for PDF reader.

% This handles basic parsing of the file (including the trailer
% and cross-reference table), as well as objects, object references,
% streams, and name/number trees; it doesn't include any facilities for
% making marks on the page.

/.setlanguagelevel where { pop 2 .setlanguagelevel } if
.currentglobal //true .setglobal
pdfdict begin

% Define the name interpretation dictionary for reading values.
/valueopdict mark
  (<<) cvn { mark } bind executeonly	% don't push an actual mark!
  (>>) cvn { { .dicttomark } stopped {
      (   **** Error: File has an unbalanced >> \(close dictionary\).\n)
      pdfformaterror
      (                Output may be incorrect.\n) pdfformaterror
    } if
  } bind executeonly
  ([) cvn { mark } bind executeonly		% ditto
  (]) cvn dup load
%  /true true		% see .pdfexectoken below
%  /false false		% ibid.
%  /null null		% ibid.
  /F dup cvx		% see Objects section below
  /R dup cvx		% see Objects section below
  /stream dup cvx	% see Streams section below
.dicttomark readonly def

% ------ Utilities ------ %

% Define a scratch string. The PDF language definition says that
% no line in a PDF file can exceed 255 characters, but this string
% is also used to search for %PDF-, which needs 1024 characters.
/pdfstring 1024 string def

% Handle the PDF 1.2 #nn escape convention when reading from a file.
% This should eventually be done in C.
/.pdffixname {			% <execname> .pdffixname <execname'>
  PDFversion 1.2 ge {
    dup .namestring (#) search {
      name#escape cvn exch pop
    } {
      pop
    } ifelse
  } if
} bind executeonly def
/name#escape			% <post> <(#)> <pre> name#escape <string>
{ exch pop
  1 index 2 () /SubFileDecode filter dup (x) readhexstring
                % Stack: post pre stream char t/f
  not {	% tolerate, but complain about bad syntax
    pop closefile (#) concatstrings exch
    (   **** Error: Invalid hex following '#' name escape, using literal '#' in name.\n)
    pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
  } {
    exch closefile concatstrings
    exch 2 1 index length 2 sub getinterval
  } ifelse
  (#) search { name#escape } if concatstrings
} bind executeonly def

/num-chars-dict mark (0123456789-.) {dup} forall .dicttomark readonly def

% Execute a file, interpreting its executable names in a given
% dictionary.  The name procedures may do whatever they want
% to the operand stack.
/.pdftokenerror {		% <count> <opdict> <errtoken> .pdftokenerror -
  % if we're called prior to actually drawing the page contents
  % (i.e. for pageusestransparency) BXlevel may not be defined, yet.
  /BXlevel where {/BXlevel get 0 le}{//true} ifelse
  {
    (   **** Error: Unknown operator: ') pdfformaterror
    dup =string cvs pdfformaterror
    (') pdfformaterror
    % Attempt a retry scan of the element after changing to PDFScanInvNum
    << /PDFScanInvNum //true >> setuserparams
    =string cvs
    token pop exch pop dup type
    dup /integertype eq exch /realtype eq or {
      exch pop exch pop
      (, processed as number, value: ) pdfformaterror
      dup =string cvs pdfformaterror (\n) pdfformaterror
     << /PDFScanInvNum //null >> setuserparams    % reset to default scanning rules
      //false 	% suppress any stack cleanup
    } {
      % error was non-recoverable with modified scanning rules, continue.
      dup type /nametype eq {
      //true 1 index .namestring {
        //num-chars-dict exch known and
      } forall { % perhaps, it's a malformed number.
        PDFSTOPONERROR {//true}{
        pop pop pop 0 //false
        ( looks like a malformed number, replacing with 0.) pdfformaterror
        } ifelse
      } {
        //true % punt
      } ifelse
    } {
      //true   % punt
    } ifelse
    (\n) pdfformaterror
    } ifelse
  } {
    //true
  } ifelse
  { % clean up the operand stack if this was non-recoverable
  pop pop count exch sub { pop } repeat	% pop all the operands
  } if
  (                Output may be incorrect.\n) pdfformaterror
} bind executeonly def

currentdict /num-chars-dict .undef

/.pdfexectoken {		% <count> <opdict> <exectoken> .pdfexectoken ?
  PDFDEBUG {
    //pdfdict /PDFSTEPcount known not { //pdfdict /PDFSTEPcount 1 .forceput } executeonly if
    PDFSTEP {
      //pdfdict /PDFtokencount 2 copy .knownget { 1 add } { 1 } ifelse .forceput
      PDFSTEPcount 1 gt {
        //pdfdict /PDFSTEPcount PDFSTEPcount 1 sub .forceput
      } executeonly
      {
        dup ==only
        (    step # ) print PDFtokencount =only
        ( ? ) print flush 1 //false .outputpage
        (%stdin) (r) file 255 string readline {
          token {
            exch pop //pdfdict /PDFSTEPcount 3 -1 roll .forceput
          } executeonly
          {
            //pdfdict /PDFSTEPcount 1 .forceput
          } executeonly ifelse % token
        } {
          pop /PDFSTEP //false def	 % EOF on stdin
        } ifelse % readline
      } ifelse % PDFSTEPcount > 1
    } executeonly
    {
      dup ==only () = flush
    } ifelse % PDFSTEP
  } executeonly if % PDFDEBUG
  2 copy .knownget {
    exch pop exch pop exch pop exec
  } {
                % Normally, true, false, and null would appear in opdict
                % and be treated as "operators".  However, there is a
                % special fast case in the PostScript interpreter for names
                % that are defined in, and only in, systemdict and/or
                % userdict: putting these three names in the PDF dictionaries
                % destroys this property for them, slowing down their
                % interpretation in all PostScript code.  Therefore, we
                % check for them explicitly here instead.
    dup dup dup /true eq exch /false eq or exch /null eq or {
      exch pop exch pop //systemdict exch get
    } {
      % Hackish fix to detect missing whitespace after "endobj". Yet another
      % problem that (you guessed it!) Adobe Acrobat ignores silently
      dup .namestring (endobj) anchorsearch {
        (   **** Error: Missing whitespace after 'endobj'.\n) pdfformaterror
        (               Output may be incorrect.\n) pdfformaterror
        pop pop pop exch pop /endobj get exec
      } {
        %% First, lets try and see if this 'might' be a broken number
        %% we look for 0-9 as well as '.' ',' and '-' to permit those
        %% locales in which the separator is a comma, as well as negative
        %% numbers. We've seena t least one tool replace '0' with '-' *sometimes*
        //true exch
        dup length 1 sub 0 1 3 -1 roll {
          1 index exch get
          dup 44 lt {
            pop exch pop //false exch exit
          }{
            dup 57 gt {
              pop exch pop //false exch exit
            }{
              dup 47 eq {
                pop exch pop //false exch exit
              }{
                pop
              } ifelse
            } ifelse
          } ifelse
        } for
        pop
        {
          %% If it looks like a number, try to deal with it as such
          PDFSTOPONERROR {
            dup .pdftokenerror
          }{
            dup {.pdftokenerror} stopped
          } ifelse
          { pop
            %% But if it fails, fall back to converting into a anem.
            %% This will propagate through and cause different
            %% fallback code to try and take care ot if.
            cvlit exch pop exch pop
          }{
            exch pop
          } ifelse
        }{
          %% Doesn't look like a number, convert it into a name
          cvlit exch pop exch pop
        } ifelse
      } ifelse
    } ifelse
  } ifelse
} bind executeonly odef
/PDFScanRules_true << /PDFScanRules //true >> def
/PDFScanRules_null << /PDFScanRules //null >> def
/.pdfrun {			% <file> <opdict> .pdfrun -
        % Construct a procedure with the stack depth, file and opdict
        % bound into it.
  1 index cvlit                  % file <<>> file
  count 2 sub                    % file <<>> file cnt
  3 1 roll mark                  % file cnt <<>> file [
  /PDFScanRules .getuserparam //null eq {
    //PDFScanRules_true { setuserparams } 0 get % force PDF scanning mode
    mark 7 4 roll
  } {
    mark 5 2 roll                % file [ [ cnt <<>> file
  } ifelse
  {	% Stack: ..operands.. count opdict file
    { token } stopped {
      dup type /filetype eq { pop } if
      pop pop stop
    } if {
      dup type /nametype eq {
        dup xcheck {
          .pdfexectoken
        } {
          .pdffixname
          exch pop exch pop PDFDEBUG {
            PDFSTEPcount 1 le {
              dup ==only ( ) print flush
            } if
          } if
        } ifelse
      } {
        exch pop exch pop PDFDEBUG {
          PDFSTEPcount 1 le {
            dup ==only ( ) print flush
          } if
        } if
      } ifelse
    } {
      pop pop exit
    } ifelse
  }
  aload pop //.packtomark exec cvx             % file [ {cnt <<>> file ... }
  { loop } 0 get 2 packedarray cvx      % file [ { {cnt <<>> file ... } loop }
  PDFSTOPONERROR { {exec //false} } { {stopped} } ifelse
  aload pop                             % file [ { {cnt <<>> file ... } loop } stopped
  /PDFScanRules .getuserparam //null eq {
    //PDFScanRules_null { setuserparams } 0 get % reset PDF scannig mode if it was off
  } if
  /PDFsource PDFsource                  % file [ { {cnt <<>> file ... } loop } stopped /PDFsource PDFsource
  { store {
      /StreamRunAborted //true store
      (   **** Error reading a content stream. The page may be incomplete.\n)
      pdfformaterror
      (               Output may be incorrect.\n) pdfformaterror
    } if
  } aload pop                           % file [ { {cnt <<>> file ... } loop } stopped /PDFsource PDFsource store {...} if
  //.packtomark exec cvx                       % file { { {cnt <<>> file ... } loop } stopped /PDFsource PDFsource store {...} if}
  /PDFsource 3 -1 roll store            % {...}
  exec
} bind executeonly def

% Execute a file, like .pdfrun, for a marking context.
% This temporarily rebinds LocalResources and DefaultQstate.
/.pdfruncontext {		% <resdict> <file> <opdict> .pdfruncontext -
  /.pdfrun load LocalResources DefaultQstate
  /LocalResources 7 -1 roll
  dup /ParentResources LocalResources put % save the parent LocalResources
  store		% store new LocalResources
  /DefaultQstate qstate store
  3 .execn
  /DefaultQstate exch store
  /LocalResources exch store
} bind executeonly def

% Get the depth of the PDF operand stack.  The caller sets pdfemptycount
% before calling .pdfrun or .pdfruncontext.  It is initially set by
% pdf_main, and is also set by any routine which changes the operand
% stack depth (currently .pdfpaintproc, although there are other callers
% of .pdfrun{context} which have not been checked for opstack depth.
/.pdfcount {		% - .pdfcount <count>
  count pdfemptycount sub
} bind executeonly def

% Read a token, but simply return false (no token read) in the case of an
% error.  This is messy because 'token' either may or may not pop its operand
% if an error occurs, and because the return values are different depending
% on whether the source is a file or a string. To avoid closing the file
% check for '{' before trying 'token'.
/token_nofail_dict mark
  ( )  { dup ( ) readstring pop pop } bind executeonly
  (\t) 1 index
  (\r) 1 index
  (\n) 1 index
  (\000) 1 index
  ({) { //null //true exit } bind executeonly
.dicttomark def

/token_nofail {		% <file|string> token_nofail false
                        % <file> token_nofail <token> true
                        % <string> token_nofail <post> <token> true
  dup type /filetype eq {
    { dup ( ) .peekstring not { ({) } if
      //token_nofail_dict exch .knownget not {
        //null 1 index { token } //.internalstopped exec exit
      } if
      exec
    } loop
    {                                   % stack: source null [source]
      //null ne { pop } if pop //false
    } {			         	% stack: source null ([post] token true | false)
      { 3 1 roll pop pop //true }
      { pop pop //false }
      ifelse
    } ifelse
  } {
    //null 1 index			% stack: source null source
    { token } //.internalstopped exec {	% stack: source null [source]
      //null ne { pop } if pop //false
    } {			         	% stack: source null ([post] token true | false)
      { 4 2 roll pop pop //true }
      { pop pop //false }
      ifelse
    } ifelse
   } ifelse
} bind executeonly def

currentdict /token_nofail_dict .undef

% ================================ Objects ================================ %

% We keep track of PDF objects using the following PostScript variables:
%
%	Generations (string): Generations[N] holds 1+ the current
%	    generation number for object number N.  (As far as we can tell,
%	    this is needed only for error checking.)  For free objects,
%	    Generations[N] is 0.
%
%	Objects (array): If object N is loaded, Objects[N] is the actual
%	    object; otherwise, Objects[N] is an executable integer giving
%	    the file offset of the object's location in the file.  If
%	    ObjectStream[N] is non-zero then Objects[N] contains the index
%	    into the object stream instead of the file offset of the object.
%
%	ObjectStream (array): If object N is in an object stream then
%	    ObjectStream[N] holds the object number of the object stream.
%	    Otherwise ObjectStream[N] contains 0.  If ObjectStream[N]
%	    is non-zero then Objects[N] contains  the index into the object
%	    stream.
%
%	GlobalObjects (dictionary): If object N has been resolved in
%	    global VM, GlobalObjects[N] is the same as Objects[N]
%	    (except that GlobalObjects itself is stored in global VM,
%	    so the entry will not be deleted at the end of the page).
%
%	IsGlobal (string): IsGlobal[N] = 1 iff object N was resolved in
%	    global VM.  This is an accelerator to avoid having to do a
%	    dictionary lookup in GlobalObjects when resolving every object.

% Initialize the PDF object tables.
/initPDFobjects {		% - initPDFobjects -
  /ObjectStream 0 array def
  /Objects 0 array def
  /Generations 0 string def
  .currentglobal //true .setglobal
  /GlobalObjects 20 dict def
  .setglobal
  /IsGlobal 0 string def
} bind executeonly def

% Grow the tables to a specified size.
/growPDFobjects {		% <minsize> growPDFobjects -
  dup ObjectStream length gt {
    dup ObjectStream exch array dup 3 1 roll copy pop /ObjectStream exch def
  } if
  dup Objects length gt {
    dup Objects exch array dup 3 1 roll copy pop /Objects exch def
  } if
  dup Generations length gt {
    %% Initiallly Generations is a string, but the xref rebuilding code
    %% can convert it to an array (/setxrefentry in pdf_rbld.ps). So we
    %% need to be able to deal with either case.
    Generations type /stringtype eq {
      dup Generations exch string dup 3 1 roll copy pop /Generations exch def
    } {
      dup Generations exch array dup 3 1 roll copy pop /Generations exch def
    } ifelse
  } if
  dup IsGlobal length gt {
    dup IsGlobal exch string dup 3 1 roll copy pop /IsGlobal exch def
  } if
  pop
} bind executeonly def

% We represent an unresolved object reference by a procedure of the form
% {obj# gen# resolveR}.  This is not a possible PDF object, because PDF has
% no way to represent procedures.  Since PDF in fact has no way to represent
% any PostScript object that doesn't evaluate to itself, we can 'force'
% a possibly indirect object painlessly with 'exec'.
% Note that since we represent streams by executable dictionaries
% (see below), we need both an xcheck and a type check to determine
% whether an object has been resolved.
/resolved? {		% <object#> resolved? <value> true
                        % <object#> resolved? false
  Objects 1 index get dup xcheck {	% Check if executable
    dup type /integertype eq {		% Check if an integer
                % Check whether the object is in GlobalObjects.
      pop IsGlobal 1 index get 0 eq {	% 0 --> Not in GlabalObjects
        pop //false			% The object is not resolved
      } {				% The object is in GlobalObjects
                % Update Objects from GlobalObjects
        PDFDEBUG { (%Global=>local: ) print dup //== exec } if
        GlobalObjects 1 index get dup Objects 4 1 roll put //true
      } ifelse
    } {				% Else object is executable but not integer
      exch pop //true		% Therefore must be executable dict. (stream)
    } ifelse
  } {				% Else object is not executable.
    exch pop //true		% Therefore it must have been resolved.
  } ifelse
} bind executeonly def
/oforce /exec load def
/oget {		% <array> <index> oget <object>
                % <dict> <key> oget <object>
                % Before release 6.20, this procedure stored the resolved
                % object back into the referring slot.  In order to support
                % PDF linearization, we no longer do this.
  get oforce
} bind executeonly def

/oforce_array { % <array> oforce_array <array>
  [ exch { oforce } forall ]
} bind executeonly def

/oforce_elems { % <array> oforce_elems <first> ... <last>
  { oforce } forall
} bind executeonly def

%% Safe recursion maintain a dicitonary in which we record all the object numbers
%% of objects which we resolve. Whenever it resolves a new one, it checks all the
%% existing ones to see if its already present, in which case we assume recursion
%% has taken place. Whenever this procedure calls itself it copies the existing
%% dictionary, and throws it away afterwards. This means that we won't falsely
%% detect recursion if two elements at the same level indirect to the same
%% object.
/safe_recursive { % <<recursion dict>> <any> oforce_recursive <recursion dict> <any>
  dup type dup /arraytype eq exch /packedarraytype eq or {
    dup rcheck {              % protect tint transform functions, etc.
      dup length 3 eq {
        dup 2 get /resolveR eq {
          dup 0 get             % <recursion dict> {x y resolveR} x
          dup                   % <recursion dict> {x y resolveR} x x
          3 index exch known {
            (   **** Error: detected circular reference in object number ) pdfformaterror pdfstring cvs pdfformaterror
            (\n) pdfformaterror
            pop //null
          } {
            dup 3 index          % <recursion dict> {x y resolveR} x x <recursion dict>
            3 1 roll put         % <recursion dict> {x y resolveR}
          } ifelse
        } if
      } if
      oforce                  % but dereference {1 0 R}
      dup type dup /arraytype eq exch /packedarraytype eq or {
        [ exch 2 index <<>> copy  exch { safe_recursive exch } forall pop]
      } {
        dup type /dicttype eq {
          << exch 2 index <<>> copy  exch { 3 -1 roll exch safe_recursive 3 -1 roll oforce 3 1 roll exch } forall pop>>
        } if
      } ifelse
    } if
  } {
    dup type /dicttype eq {
      << exch 2 index <<>> copy  exch { 3 -1 roll exch safe_recursive 3 -1 roll oforce 3 1 roll exch } forall pop>>
    } if
  } ifelse
} bind executeonly def

/oforce_recursive {
  << >> exch safe_recursive exch pop
} bind executeonly def

% A null value in a dictionary is equivalent to an omitted key;
% we must check for this specially.
/knownoget {	% <dict> <key> knownoget <value> true
                % <dict> <key> knownoget false
                % See oget above regarding this procedure.
  .knownget {
    oforce dup //null eq { pop //false } { //true } ifelse
  } {
    //false
  } ifelse
} bind executeonly def

% See /knownoget above.
/oknown {	% <dict> <key> oknown <bool>
  .knownget { oforce //null ne } { //false } ifelse
} bind executeonly def

/knownogetdict { % <dict> <key> knownogetdict <dict> true
                 % <dict> <key> knownogetdict false
  //knownoget exec dup {
    1 index type /dicttype ne { pop pop //false } if
  } if
} bind executeonly def

% PDF 1.1 defines a 'foreign file reference', but not its meaning.
% Per the specification, we convert these to nulls.
/F {		% <file#> <object#> <generation#> F <object>
                % Some PDF 1.1 files use F as a synonym for f!
   .pdfcount 3 lt { f } { pop pop pop //null } ifelse
} bind executeonly def

% Verify the generation number for a specified object
% Note:  The values in Generations is the generation number plus 1.
% If the value in Generations is zero then the object is free.
/checkgeneration {  % <object#> <generation#> checkgeneration <object#> <OK>
  Generations 2 index get 1 sub 1 index eq {	% If generation # match ...
    pop //true					% Then return true
  } {					% Else not a match ...
    QUIET not {				% Create warning message if not QUIET
      Generations 2 index get 0 eq {	% Check if object is free ...
        (   **** Warning: reference to free object: )
        2 index =string cvs concatstrings ( ) concatstrings	% put obj #
        exch =string cvs concatstrings ( R\n) concatstrings	% put gen #
      } {
        (   **** Warning: wrong generation: )
        2 index =string cvs concatstrings ( ) concatstrings	% put obj #
        exch =string cvs concatstrings				% put gen #
        (, xref gen#: ) concatstrings 1 index Generations  % put xref gen #
        exch get 1 sub =string cvs concatstrings (\n) concatstrings
      } ifelse
      pdfformatwarning			% Output warning message
    } {					% Else QUIET ...
      pop				% Pop generation number
    } ifelse
    % We should return false for an incorrect generation number, however
    % we are simply printing a warning and then returning true.  This makes
    % Ghostscript tolerant of of bad generation numbers.
    //true
  } ifelse
} bind executeonly def
/R {		% <object#> <generation#> R <object>
  %% Parameter validation; this should not be required but we have found files
  %% (Bug 697351) which are corrupted in ways that mean we see a 'R' in a stream
  %% even though it is not an indirect object reference. Because a stream
  %% may reference objetcs which have not been resolved, we cannot simply
  %% 'undef' teh R operator, the only solution is to have it chekc the
  %% types of its operands. This is poterntially slow of course, we may
  %% need to remove this in future if the performance penalty is too great.
  %% We cannot slow down all files significantly purely to work-around
  %% files which are invalid.
  1 index type /integertype eq 1 index type /integertype eq and {
    /resolveR cvx 3 packedarray cvx
  } {
    (   **** Error: indirect object reference \(R\) encountered with invalid arguments.) pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
  } ifelse
} bind executeonly def

% If we encounter an object definition while reading sequentially,
% we just store it away and keep going.
/objopdict mark
  valueopdict { } forall
  /endobj dup cvx
.dicttomark readonly def

/obj {			% <object#> <generation#> obj <object>
  PDFfile objopdict .pdfrun
} bind executeonly def

/endobj {		% <object#> <generation#> <object> endobj <object>

%% If we have a stream wioth a broken endstream we can get here without
  %% an object definition on nthe stack. If we simply pop the 'extraneous'
  %% data we will break the stream definition and throw an error. So we look
  %% to see if this appears to be this specific case.
  2 index type /marktype eq 2 index type /marktype eq or {
    {
      dup type /marktype eq {
        exit
      }{
        pop
      }if
    }loop
  }{
    {
      2 index type /integertype ne 2 index type /integertype ne or {
        count 3 gt {
          (   **** Error: obj definition followed by multiple tokens, attempting to recover.\n) pdfformaterror
          (               Output may be incorrect.\n) pdfformaterror
          pop
        } {
          //null
          (   **** Error: ignoring obj followed by multiple tokens.\n) pdfformaterror
          (               Output may be incorrect.\n) pdfformaterror
          exit
        } ifelse
      } {
        exit
      } ifelse
    } loop

3 1 roll
                  % Read the xref entry if we haven't yet done so.
                  % This is only needed for generation # checking.
    1 index resolved? {
      pop
    } if
    checkgeneration {
                  % The only global objects we bother to save are
                  % (resource) dictionaries.
      1 index dup gcheck exch type /dicttype eq and {
        PDFDEBUG { (%Local=>global: ) print dup //== exec } if
        GlobalObjects 1 index 3 index put
        IsGlobal 1 index 1 put
      } if
      Objects exch 2 index put
    } {
      pop pop //null
    } ifelse
  }ifelse
} bind executeonly def

% When resolving an object reference in an object stream, we stop at
% the end of file.  Note:  Objects in an object stream do not have either
% a starting 'obj' or and ending 'endobj'.
/resolveobjstreamopdict mark
  valueopdict { } forall
  (%%EOF) cvn { exit } bind executeonly
  /endobj {  % bug 689795
    (   **** Error: Objects in an object stream should not have 'endobj'.\n)
    pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
  } bind executeonly
.dicttomark readonly def

% Note: This version of this function is not currently being used.
% Resolve all objects in an object stream
/resolveobjectstream {		% <object stream #> resolveobjectstream -
  PDFDEBUG { (%Resolving object stream: ) print } if
  0 resolveR	% Get the objectstream dict, all objstrms use 0 as the gen #
  dup /First get		% Save location of first object onto the stack
  1 index /N get		% Save number of objects onto the stack
  2 index //false resolvestream	% Convert stream dict into a stream
  /ReusableStreamDecode filter	% We need to be able to position stream
                % Objectstreams begin with list of object numbers and locations
                % Create two arrays to hold object numbers and stream location
  1 index array			% Array for holding object number
  2 index array			% Array for holding stream object location
                % Get the object numbers and locations.
  0 1 5 index 1 sub {		% Loop and collect obj # and locations
                % Stack: objstreamdict First N objectstream [obj#] [loc] index
    2 index 1 index 		% Setup to put obj# into object number array
    5 index token pop put	% Get stream, then get obj# and put into array
    1 index 1 index 		% Setup to put object loc into location array
    5 index token pop put	% Get stream, get obj loc and put into array
    pop				% Remove loop index
  } for
                % Create a bytestring big enough for reading any object data
                % Scan for the size of the largest object
  0 0				% Init max object size and previous location
  2 index {			% Loop through all object locations
                                % Stack:  ... maxsize prevloc currentloc
    dup 4 1 roll		% Save copy of object location into stack
    exch sub				% Object size = currentloc - prevloc
    .max			% Determine maximum object size
    exch			% Put max size under previous location
  } forall
  pop				% Remove previous location
  .bigstring			% Create bytestring based upon max obj size
                % Move to the start of the object data
  3 index 6 index		% Get objectstream and start of first object
  setfileposition		% Move to the start of the data
                % Read the data for all objects except the last.  We do
                % not know the size of the last object so we need to treat
                % it as a special case.
  0 1 6 index 2 sub {
    dup 4 index exch get 	% Get our current object number
                % Stack: objstreamdict First N objectstream [obj#] [loc]
                %        bytestring loopindex object#
    dup resolved? {		% If we already have this object
      pop pop			% Remove object and object number
      1 add 2 index exch get	% Get location of next object
      6 index add 6 index exch	% Form location of next object and get stream
      setfileposition		% Move to the start of the next object data
    } {				% Else this is a new object ...
                % We are going to create a string for reading the object
      2 index 0			% use our working string
                % Determine the size of the object
      5 index 4 index 1 add get	% Get location of the next object
      6 index 5 index get	% Get location of this object
      sub			% Size of object = next loc - this loc
      getinterval		% Create string for reading object
      6 index exch readstring pop	% Read object
      /ReusableStreamDecode filter 	% Convert string into a stream
      resolveobjstreamopdict .pdfrun	% Get PDF object
      Objects exch 2 index exch put	 % Put object into Objects array
      pop pop			% Remove object # and loop index
    } ifelse
  } for
  pop pop			% Remove our working string and loc array
                % Now read the last object in the object stream.  Since it
                % is the last object, we can use the original stream and
                % terminate when we hit the end of the stream
                % Stack: objstreamdict First N objectstream [obj#]
  2 index 1 sub get	 	% Get our current object number
  dup resolved? not {		% If we do not already have this object
    exch	 		% Get our object stream
    resolveobjstreamopdict .pdfrun	% Get PDF object
    Objects exch 2 index exch put	% Put object into Objects array
  } if
  pop pop pop pop		% Clear stack
} bind executeonly def

/no_debug_dict <<
  /PDFDEBUG //false
>> readonly def

% Resolve all objects in an object stream
/resolveobjectstream {		% <object stream #> resolveobjectstream -
  PDFDEBUG { (%Resolving object stream: ) print } if
  dup 0 resolveR % Get the objectstream dict, all objstrms use 0 as the gen #
  dup /Type get /ObjStm ne {	% Verify type is object stream
    (   **** Error: Incorrect Type in object stream dictionary.\n) pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
    /resolveobjectstream cvx /typecheck signalerror
  } if
  dup /N get			% Save number of objects onto the stack
  1 index //false resolvestream	% Convert stream dict into a stream
  /ReusableStreamDecode filter	% We need to be able to position stream
                % Objectstreams begin with list of object numbers and locations
  1 index array			% Create array for holding object number
                % Get the object numbers
  0 1 4 index 1 sub {		% Loop and collect obj numbers
                % Stack: strm# objstreamdict N PDFDEBUG objectstream [obj#] loopindex
    1 index 1 index 		% Setup to put obj# into object number array
    4 index token pop put	% Get stream, then get obj# and put into array
    2 index token pop pop pop	% Get stream, get obj loc and clear stack
  } for
                % Move to the start of the object data
  1 index 4 index /First get	% Get objectstream and start of first object
  setfileposition		% Move to the start of the data
                % We disable PDFDEBUG while reading the data stream.  We will
                % print the data later
  PDFDEBUG { //no_debug_dict begin } if
                % Read the data for all objects.  We check to see if we get
                % the number of objects that we expect.
                % Stack: strm# objstreamdict N objectstream [obj#] PDFDEBUG
  mark 3 -1 roll 		% Get objectstream
  count 4 index add		% Determine stack depth with objects
  3 1 roll
  resolveobjstreamopdict .pdfrun % Get PDF objects
  count counttomark 1 add index ne { % Check stack depth
    (   **** Error: Incorrect object count in object stream.\n) pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
    /resolveobjectstream cvx /rangecheck signalerror
  } if
                % We have the object data
  counttomark array astore	% Put objects into an array
  exch pop exch pop		% Remove mark and count
  currentdict //no_debug_dict eq { end } if % Restore debug context
                % Save the objects into Objects
  0 1 2 index length 1 sub {	% Loop through all objects
                % Stack: strm# objstreamdict N [obj#] [objects] loopindex
    dup 3 index exch get 	% Get our current object number
                % Stack: strm# objstreamdict N [obj#] [objects] loopindex obj#
    dup ObjectStream exch get 7 index eq {
      dup resolved? {		% If we already have this object
        pop pop			% Remove object and object number
      } {			% Else if we do not have this object
        PDFDEBUG { (%Resolving compressed object: [) print dup =only ( 0]) = } if
        Objects exch 3 index	% Put the object into Objects
        3 index get
        PDFDEBUG { dup === flush } if
        put
      } ifelse
    } {
      pop	% Ignore old object; remove object number.
    } ifelse
    pop 			% Remove loop index
  } for
  pop pop pop pop pop		% Remove strm# objstream, N, (obj#], and [objects]
} bind executeonly def

currentdict /no_debug_dict undef

% When resolving an object reference, we stop at the endobj or endstream.
/resolveopdict mark
  valueopdict { } forall
  /xref { % Bug 697761
     (   **** Error: Encountered 'xref' while expecting 'endobj'.\n) pdfformaterror
     (               Treating this as a missing 'endobj', output may be incorrect.\n) pdfformaterror
     endobj exit
  } bind executeonly
  /endstream { endobj exit } bind
  /endobj { endobj exit } bind
  /endjobj { % Bug 689876.
     (   **** Error: Operator 'endobj' is misspelled as 'endjobj'.\n) pdfformaterror
     (               Output may be incorrect.\n) pdfformaterror
     endobj exit
  } bind executeonly
  /enbobj { % Bug 690397.
     (   **** Error: Operator 'endobj' is misspelled as 'enbobj'.\n) pdfformaterror
     (               Output may be incorrect.\n) pdfformaterror
     endobj exit
  } bind executeonly
  /obj {
                % OmniForm generates PDF file with endobj missing in some
                % objects. AR ignores this. So we have to do it too.
     (   **** Error: Encountered 'obj' while expecting 'endobj'.\n) pdfformaterror
     (               Treating this as a missing 'endobj', output may be incorrect.\n) pdfformaterror
     pop pop endobj exit
  } bind executeonly
.dicttomark readonly def

/resolveR {             % <object#> <generation#> resolveR <object>
  1 index 3 1 roll       % copy the object for the code after pdf_run_resolve (may store it)
  PDFDEBUG {
    PDFSTEPcount 1 le {
      (%Resolving: ) print 2 copy 2 array astore //== exec
    } if
  } if
  1 index dup 0 le exch NumObjects ge or {
    (   **** Error: Considering object with an invalid number )
    2 index 20 string cvs concatstrings
    ( as null.\n) concatstrings pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
    pop pop //null
  } {
    1 index resolved? {           % If object has already been resolved ...
      exch pop exch pop           % then clear stack and return object
    } {                           % Else if not resolved ...
      PDFfile fileposition 3 1 roll       % Save current file position
      1 index Objects exch get           % Get location of object from xref
      3 1 roll checkgeneration {          % Verify the generation number
                        % Stack: savepos objpos obj#
         ObjectStream 1 index get dup 0 eq { % Check if obj in not an objstream
           pop exch PDFoffset add PDFfile exch setfileposition
           PDFfile token pop 2 copy ne
            { (   **** Error: Unrecoverable error in xref!\n) pdfformaterror
              (               Output may be incorrect.\n) pdfformaterror
              /resolveR cvx /rangecheck signalerror
            }
           if pop PDFfile token pop
           PDFfile token pop /obj ne
            { (   **** Error: Unrecoverable error in xref!\n) pdfformaterror
              (               Output may be incorrect.\n) pdfformaterror
              /resolveR cvx /rangecheck signalerror
            }
           if
           pdf_run_resolve        % PDFfile resolveopdict .pdfrun
        } {                       % Else the object is in an ObjectStream
                  % Process an objectstream object.  We are going to resolve all
                  % of the objects in sthe stream and place them into the Objects
                  % array.
                  % Stack: savepos objpos obj# objectstream#
          resolveobjectstream
          resolved? {             % If object has already been resolved ...
            exch pop              % Remove object pos from stack.
          } {
            pop pop //null        % Pop objpos and obj#, put null for object
          } ifelse
        } ifelse
      } {                         % Else the generation number is wrong
              % Don't cache if the generation # is wrong.
          pop pop //null          % Pop objpos and obj#, put null for object
      } ifelse                    % ifelse generation number is correct
      exch PDFfile exch setfileposition   % Return to original file position
    } ifelse
  } ifelse

dup type /dicttype eq  % For dictionaries only
   {
     dup /Type known  % make sure the dictionary has a type
     {
       %% We must not add to *all* dictionaries, because some of them are handled by
       %% executing all the key/value pairs, and adding a integer causes problems
       %% Since we only use this number for font identification, or checking self-references,
       %% we only need it for dictionaries with a Type, and these are always handled more carefully.
       dup 3 -1 roll /.gs.pdfobj# exch put
     }
     {exch pop}
     ifelse
   }
   {exch pop}
   ifelse
} bind executeonly def

% ================================ Streams ================================ %

% We represent a stream by an executable dictionary that contains,
% in addition to the contents of the original stream dictionary:
%	/File - the file or string where the stream contents are stored,
%	  if the stream is not an external one.
%	/FilePosition - iff File is a file, the position in the file
%	  where the contents start.
%	/StreamKey - the key used to decrypt this stream, if any.
% We do the real work of constructing the data stream only when the
% contents are needed.

% Construct a stream.  The length is not reliable in the face of
% different end-of-line conventions, but it's all we've got.
%
% PDF files are inconsistent about what may fall between the 'stream' keyword
% and the actual stream data, and it appears that no one algorithm can
% detect this reliably.  We used to try to guess whether the file included
% extraneous \r and/or \n characters, but we no longer attempt to do so,
% especially since the PDF 1.2 specification states flatly that the only
% legal terminators following the 'stream' keyword are \n or \r\n, both of
% which are properly skipped and discarded by the token operator.
% Unfortunately, this doesn't account for other whitespace characters that
% may have preceded the EOL, such as spaces or tabs. Thus we back up one
% character and scan until we find the \n terminator.
/stream {	% <dict> stream <modified_dict>
  dup /F known dup PDFsource PDFfile eq or {
    not {
      dup /File PDFfile put
      % make sure that we are just past the EOL \n character
      PDFfile dup fileposition 1 sub setfileposition	% back up one
      % Skip spaces till \n or a non-space character is found.
      //false
      {
        PDFfile read pop
        dup 32 eq { pop //true or } { exit } ifelse
      } loop
      dup 13 eq {
        pop
        PDFfile read pop
      } if
      10 ne {
        PDFfile dup fileposition 1 sub setfileposition
        //true or
      } if {
         (   **** Error: stream operator isn't terminated by valid EOL.\n) pdfformaterror
         (               Output may be incorrect.\n) pdfformaterror
      } if

dup /FilePosition PDFfile fileposition put
      PDFDEBUG {
        PDFSTEPcount 1 le {
          (%FilePosition: ) print dup /FilePosition get //== exec
        } if
      } if
    } if
    % Some (bad) PDf files have invalid stream lengths.  This causes problems
    % if we reposition beyond the end of the file.  So we compare the given
    % length to number of bytes left in the file.
    dup /Length knownoget {
      dup PDFfile bytesavailable lt {	% compare to to bytes left in file
        PDFfile fileposition 		% reposition to the end of stream
        add PDFfile exch setfileposition
      } {
        pop				% bad stream length - do not reposition.
                                        % This will force a length warning below
      } ifelse
    } if
  } {
    pop
        % We're already reading from a stream, which we can't reposition.
        % Capture the sub-stream contents in a string.
    dup /Length oget string PDFsource exch readstring
    not {
      (   **** Error: Unexpected EOF in stream!\n) pdfformaterror
      (               Output may be incorrect.\n) pdfformaterror
      /stream cvx /rangecheck signalerror
    } if
    1 index exch /File exch put
  } ifelse
  PDFsource token_nofail not { //null } if
  dup /endobj eq {
    % Another case that Acrobat Reader handles -- 'endobj' without 'endstream'.
    (   **** Error: stream missing 'endstream'.\n) pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
    pop /endstream		% fake a valid endstream
  } if
  /endstream ne {
    (   **** Error: stream Length incorrect.\n) pdfformaterror
    (               Output may be incorrect.\n) pdfformaterror
    dup /Length undef % prevent the use of the incorrect length.

%% Bug #696560 If we get a stream whose length is incorrect we cannot
    %% safely *assume* it has an object definition as we can get here when
    %% reading an XRef stream, and in that case we will not have the object
    %% and generation numbers on the stack!
    %% So before we call endobj and exit pdfrun, try and see if ths looks like
    %% a regular object or not. If it isn't then don't try and ahndle it as
    %% one, just return the object.
    count 3 ge {
      1 index type /integertype eq 2 index type /integertype eq and {
        cvx endobj exit   % exit from .pdfrun now.
      } if
    } if
  } {
    PDFsource (??) .peekstring pop (>>) eq { % Bug 690161, sample #1
      (   **** Error: Spurious '>>' after 'endstream' ignored.\n) pdfformaterror
      (               Output may be incorrect.\n) pdfformaterror
      PDFsource (12) readstring pop pop
    } if
  } ifelse
  cvx
} bind executeonly def

/endstream {
  exit
} bind executeonly def

% Contrary to the published PDF (1.3) specification, Acrobat Reader
% accepts abbreviated filter names everywhere, not just for in-line images,
% and some applications (notably htmldoc) rely on this.
/unabbrevfilterdict mark
  /AHx /ASCIIHexDecode  /A85 /ASCII85Decode  /CCF /CCITTFaxDecode
  /DCT /DCTDecode  /Fl /FlateDecode  /LZW /LZWDecode  /RL /RunLengthDecode
.dicttomark readonly def

% Extract and apply filters.
/filterparms {		% <dict> <DPkey> <Fkey> filterparms
                        %   <dict> <parms> <filternames>
  2 index exch knownoget {
    oforce_recursive
    exch 2 index exch knownoget {
      dup length 0 eq {
        pop //false
      } {
        //true
      } ifelse
    } {
      //false
    } ifelse {
                % Both filters and parameters.
      oforce_recursive
      exch dup type /nametype eq {
        1 array astore exch
        dup type /arraytype ne { 1 array astore } if exch
      } if
    } {
                % Filters, but no parameters.
      //null exch
      dup type /nametype eq { 1 array astore } if
    } ifelse
  } {
                % No filters: ignore parameters, if any.
    pop //null { }
  } ifelse
} bind executeonly def
/filtername {		% <filtername> filtername <filtername'>
  //unabbrevfilterdict 1 index .knownget { exch pop } if
  dup /Filter resourcestatus { pop pop } {
    Repaired exch	% this error is not the creator's fault
    RepiredAnError exch
    (   **** ERROR: Unable to process ) pdfformaterror
    64 string cvs pdfformaterror
    ( data. Page will be missing data.\n) pdfformaterror
    /RepairedAnError exch store
    /Repaired exch store % restore the previous "Repaired" state
    % provide a filter that returns EOF (no data)
    /.EOFDecode
  } ifelse
} bind executeonly def

/pdf_rules_dict << /PDFRules //true >> readonly def

% Add PDF option to ASCII85Decode filter
% <source>        <name> add_A85_param <source> <dict'> <name>
% <source> <dict> <name> add_A85_param <source> <dict'> <name>
/add_A85_param {
  dup /ASCII85Decode eq {
    1 index type /dicttype eq {
      exch dup length 1 add dict copy
      dup /PDFRules //true put
      exch
     } {
      //pdf_rules_dict exch
    } ifelse
  } if
} bind executeonly def

currentdict /pdf_rules_dict undef

/applyfilters {		% <parms> <source> <filternames> applyfilters <stream>
  2 index //null eq
  {
    { filtername add_A85_param filter }
  }
  {
    dup length 3 index length ne         %% compare lengths of DecodeParmas and Filter arrays
    {                                    %% if they aren't the same, ignore the decode params
      (   **** Error: ignoring stream /DecodeParams array as its length is different to the Filters array.\n) pdfformaterror
      (               Output may be incorrect.\n) pdfformaterror
      3 -1 roll pop //null 3 1 roll
      { filtername add_A85_param filter }
    }
    {
      {		% Stack: parms source filtername
        2 index 0 oget dup type /dicttype ne
        { pop }
        {
          exch filtername dup /JBIG2Decode eq { exch jbig2cachectx exch } if
        } ifelse
        add_A85_param filter
        exch dup length 1 sub 1 exch getinterval exch
      }
    } ifelse
  } ifelse
  forall exch pop
} bind executeonly def

% JBIG2 streams have an optional 'globals' stream obj for
% sharing redundant data between page images. Here we resolve
% that stream reference (if any) and run it through the decoder,
% creating a special -jbig2globalctx- postscript object our
% JBIG2Decode filter implementation looks for in the parm dict.
/jbig2cachectx { % <parmdict> jbig2cachectx <parmdict>
  dup /JBIG2Globals knownoget {
    % make global ctx
    PDFfile fileposition exch % resolvestream is not reentrant
    //true resolvestream 		% stack after: PDFfileposition -file-
    % Read the data in a loop until EOF to so we can move the strings into a bytestring
    [ { counttomark 1 add index 60000 string readstring not { exit } if } loop ]
    exch pop 0 1 index { length add } forall	% compute the total length
    % now copy the data from the array of strings into a bytestring
    .bytestring exch 0 exch { 3 copy putinterval length add } forall pop
    .jbig2makeglobalctx
    PDFfile 3 -1 roll setfileposition
    1 index exch
    /.jbig2globalctx exch put
  } if
} bind executeonly def

% Resolve a stream dictionary to a PostScript stream.
% Streams with no filters require special handling:
%     - Whether we are going to interpret the stream, or If we are just
%       going to read data from them, we impose a SubFileDecode filter
%         that reads just the requisite amount of data.
% Note that, in general, resolving a stream repositions PDFfile.
% Clients must save and restore the position of PDFfile themselves.
/resolvestream {	% <streamdict> <readdata?> resolvestream <stream>
  1 index /F knownoget {
                % This stream is stored on an external file.
    (r) file 3 -1 roll
    /FDecodeParms /FFilter filterparms
                % Stack: readdata? file dict parms filternames
    4 -1 roll exch
    pdf_decrypt_stream
    applyfilters
  } {
    exch
    dup /Length knownoget { 0 eq } { //false } ifelse {
      () 0 () /SubFileDecode filter
    } {
      dup /FilePosition .knownget {
        1 index /File get exch setfileposition
      } if
                % Stack: readdata? dict
      /DecodeParms /Filter filterparms
                % Stack: readdata? dict parms filternames
      2 index /File .knownget not {
        (   **** Error: Stream object is missing the stream data.\n)
        (               Output may be incorrect.\n) pdfformaterror
        pdfformaterror
        ()
      } if exch
                % Stack: readdata? dict parms file/string filternames
      dup length 0 eq {
                % All the PDF filters have EOD markers, but in this case
                % there is no specified filter.
        exch dup type /filetype eq 5 index or {
              % Use length for any files or reading data from any source.
          3 index /Length knownoget not { 0 } if
        } {
        0       % Otherwise length of 0 for whole string
        } ifelse
        4 index /IDFlag known { pop } { () /SubFileDecode filter } ifelse
        exch
        pdf_decrypt_stream		% add decryption if needed
        pop exch pop
      } {
                % Stack: readdata? dict parms source filternames
        exch 3 index /Length knownoget {
          () /SubFileDecode filter
        } if exch
        pdf_decrypt_stream		% add decryption if needed
        applyfilters
      } ifelse
    } ifelse
  } ifelse
                % Stack: readdata? dict file
  exch pop exch pop
} bind executeonly def

% ============================ Name/number trees ============================ %

/nameoget {		% <nametree> <key> nameoget <obj|null>
  exch /Names exch .treeget
} bind executeonly def

/numoget {		% <numtree> <key> numoget <obj|null>
  exch /Nums exch .treeget
} bind executeonly def

/.treeget {		% <key> <leafkey> <tree> .treeget <obj|null>
  dup /Kids knownoget {
    exch pop .branchget
  } {
    exch oget .leafget
  } ifelse
} bind executeonly def

/.branchget {		%  <key> <leafkey> <kids> .branchget <obj|null>
  dup length 0 eq {
    pop pop pop //null
  } {
    dup length -1 bitshift 2 copy oget
                        % Stack: key leafkey kids mid kids[mid]
    dup /Limits oget aload pop
                        % Stack: key leafkey kids mid kids[mid] min max
    6 index lt {
      pop pop
      1 add 1 index length 1 index sub getinterval .branchget
    } {
      5 index gt {
        pop
        0 exch getinterval .branchget
      } {
        exch pop exch pop .treeget
      } ifelse
    } ifelse
  } ifelse
} bind executeonly def

/.leafget {		% <key> <pairs> .leafget <obj|null>
  dup length 2 eq {
    dup 0 get 2 index eq { 1 oget } { pop //null } ifelse
    exch pop
  } {
    dup length -1 bitshift -2 and 2 copy oget
                        % Stack: key pairs mid pairs[mid]
    3 index gt { 0 exch } { 1 index length 1 index sub } ifelse
    getinterval .leafget
  } ifelse
} bind executeonly def

% The following variants return tree entry whose key is closest but
% less or equal to the given key.

/numogetle {		% <numtree> <key> numogetle <key obj true|false>
  exch /Nums exch .treegetle
} bind executeonly def

/.treegetle {		% <key> <leafkey> <tree> .treegetle <key obj true|false>
  dup /Kids knownoget {
    exch pop .branchgetle
  } {
    exch oget .leafgetle
  } ifelse
} bind executeonly def

/.branchgetle {		%  <key> <leafkey> <kids> .branchgetle  <key obj true|false>
    dup length 0 eq {
        pop pop pop //false
    } {
        dup length -1 bitshift 2 copy oget
        dup /Limits oget aload pop
        % Stack: key leafkeyb kids mid kids[mid] min max
        6 index ge {
            5 index le {
                exch pop exch pop .treegetle
            } {
                pop 0 exch getinterval .branchgetle
            } ifelse
        } {
            pop 2 index length 2 index sub 1 gt {
                pop 1 index length 1 index sub getinterval .branchgetle
            } {
                exch pop exch pop .treegetle
            } ifelse
        } ifelse
    } ifelse
} bind executeonly def

/.leafgetle {		% <key> <pairs> .leafget <obj|null>
    dup length 2 eq {
        dup 0 get
        2 index le { exch pop aload pop //true } { pop pop //false } ifelse
    } {
        dup length -1 bitshift -2 and 2 copy oget
        % Stack: key pairs mid pairs[mid]
        3 index gt { 0 exch } { 1 index length 1 index sub } ifelse
        getinterval .leafgetle
    } ifelse
} bind executeonly def

end			% pdfdict
.setglobal

MR SKB