Edit File: pdf_rbld.ps
% Copyright (C) 2001-2018 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_rbld.ps - Rebuilding of broken PDF files (xref errors) % This module contains routines that are used if we detect an error % while reading the xref tables. These routines will scan the file and % build an xref table by finding the objects. We also need to find the % appropriate trailer dictionary. Note: One procedure is also used % even if we do not need to rebuild a PDF file. % % This module cannot rebuild a PDF file which has had errors created inside % of objects or binary data streams. It often succeeds with files that % have had its end of lines converted between unix and dos versions. % if true --> we have an object with duplicate object and generation numbers. /dup_obj_gen_num //false def % Note: This procedure is also used by non-rebuild code. % Store a line in the xref array (Actually Objects and Generations arrays) % <obj num> (strm num> <obj loc> <gen num> <rebuild> % setxrefentry <obj num> strm num> <obj loc> <gen num> /setxrefentry { 5 1 roll dup 65535 or 65535 ne { ( **** Error: Generation number out of 0..65535 range, assuming 0.\n) pdfformaterror ( Output may be incorrect.\n) pdfformaterror pop 0 } if % We store generation numbers as value + 1 % We reserve 0 to indicate an free xref entry 1 add % increment generation number % To save space, generations numbers are stored in a string unless we % find a generation number greater than 255. If so then transfer to % an array. dup 255 gt { Generations type /stringtype eq { % Convert Generations to an array. Generations length array dup % Create new array 0 1 2 index length 1 sub { % Copy from old string to new array Generations 1 index get put dup } for pop /Generations exch store % Save new Generations array } if } if % Verify that the new values are for a new object. If the current % entry is null then we have a new entry. Objects 4 index get //null eq { ObjectStream 4 index 4 index cvx put % Save ObjectStream object number Objects 4 index 3 index cvx put % Save object location Generations 4 index 2 index put % Save geenration number } { % Verify that the new entry has at least as high a generaton number % We accept equal entry number because we have found PDF files in % which there are multiple objects with the same object and entry % numbers. The normal xref logic only accepts the first such % entry that it finds. However the 'rebuild PDF' logic can find % both such entries. The correct one is usually the last one. Generations 4 index get 1 index le { %% Check if the object we already found was at the locaton specified %% in the original xref (now stored in Orig_Objects). If so, prefer %% that offset, otherwise prefer the later object. %% NB check first to see that the object number is in the range of the original %% xref. If it isn't, set the 'original' Offset to 0. 3 index Orig_Objects length le {Orig_Objects 4 index get}{0}ifelse Objects 5 index get eq not { ObjectStream 4 index 4 index cvx put % Save ObjectStream object number Objects 4 index 3 index cvx put % Save object location Generations 4 index 2 index put % Save geenration number } if } if % Set error flag if we have equal object and generation numbers Generations 4 index get 1 index eq { /dup_obj_gen_num //true def } if } 8 -1 roll { ifelse } { pop if } ifelse % Run 'else' only when rebuilding. } bind executeonly def % Print the contents of the xref array. This actually consists of three % arrays (Objects, Generations, and ObjectStream). /print_xref % - print_xref - { 0 1 Objects length 1 sub % stack: 0 1 <number of objects - 1> { dup =only % print object number ( ) print dup Generations exch get 1 sub =only % print Generation number ( ) print dup ObjectStream exch get ==only % print ObjectStream object number ( ) print Objects exch get === % print object location } for flush } bind executeonly def % Get token from string and check its type % <string> <type> typed_token <false> % no token or not match % <string> <type> typed_token <obj> <last> <true> % matching token type % Where last is the string remainder /typed_token { exch token_nofail % get token { dup type % stack: type last token type 4 -1 roll eq { % stack: last token bool exch //true % desired object found - set exit status } { pop pop //false % not type - clear stack, set exit status } ifelse } { pop //false % no token - pop type, set exit status } ifelse % check if we got token } bind executeonly def % Allocate space for post_eof_count to be bound into procedures below. /post_eof_count 0 def % We want the location of the trailer dictionary at the start of file. % First we will find the xref. Then we will skip over the xref entries % to the trailer. /search_start_trailer % - search_start_trailer <trailer loc> { % Read the first 300 bytes and check for xref PDFfile 0 setfileposition PDFfile bytesavailable post_eof_count sub % location of end of data 300 .min % block size to read dup string 0 1 4 -1 roll 1 sub { 2 copy PDFfile read pop put pop } for (xref) search { % found 'xref' exch pop exch pop length 4 add PDFfile exch setfileposition PDFfile token pop % get starting entry - or 'trailer' (trailer) ne { % if we do not already have 'trailer' PDFfile token pop % get number of entries PDFfile token pop pop % this moves us into the middle of the first entry 25 string exch % define working string for readline { PDFfile 1 index readline pop pop } repeat % skip entries pop % pop working string PDFfile token pop pop % get 'trailer' PDFfile fileposition % get file position } if } { pop 0 % no xref, should not happen, report it upstrem } ifelse } bind executeonly def %% Searches backwards from a specified point looking for a 'trailer' keyword. %% position search_earlier_trailer position or 0 %% %% Its is just possible that a 'trailer' keyword could straddle a buffer, in which case %% we wouldn't find it. Given that this only executes for broken files anyway I don't %% propose to worry about it at the moment, if anyone ever turns up an example we may %% choose to enhance this routine further. %% /search_earlier_trailer { { % position dup 0 gt { % position bool dup 65535 .min exch % block_size position 1 index sub % block_size position-block_size dup % block_size new_position new_position PDFfile exch setfileposition % block_size new position exch dup % new position block_size block_size dup string 0 1 4 -1 roll 1 sub % {2 copy PDFfile read pop put pop } for % % new_position block size (...string from file....) (trailer) search { pop { search not { exit } if pop } loop % determine where the trailer is in the file % trailer loc = end loc - remaing string length length % new_position block size string length 3 1 roll % string length new_position block size add exch sub % string length - (new_position + block size) } { pop % discard old block size pop 0 } ifelse } { pop 0 exit }ifelse %% We either have a position for a trailer, or 0 if we failed to find one dup 0 eq not { exit }if pop % renove the zero leaving the new start position } loop } bind executeonly def % We want the location of the trailer dictionary at the end of file. % We will read the last block of data and search for the final occurance % of the word 'trailer' /search_end_trailer % - search_end_trailer <trailer loc> { % Position to read block of data from the end of the file. Note: We ignore % anything past the last %%EOF since this is not PDF data. PDFfile 0 setfileposition PDFfile bytesavailable post_eof_count sub % location of end of data dup 65535 .min % block size to read % stack: <file end pos> <block size> % move file position to the start of the block 2 copy sub PDFfile exch setfileposition % read block of data dup string 0 1 4 -1 roll 1 sub { 2 copy PDFfile read pop put pop } for % search for last occurance of 'trailer' (trailer) search { pop { search not { exit } if pop } loop % determine where the trailer is in the file % trailer loc = end loc - remaing string length length sub } { pop pop 0 } ifelse } bind executeonly def % We want to find the trailer dictionary. There is a trailer dictionary % for each xref object list. We only want the trailer dictionary associated % with the first xref object list. In theory this can be anywhere in the % file. However since we are trying to repair a broken file, we cannot simply % follow the xref links. So we are falling back to a simple strategy. We % find the specified location of the first xref list. If its location is in % the first half of the file then we search for the first trailer dictionary % at the start of the file. Otherwise we search for the last trailer at the % end of the file. /search_trailer % - search_trailer - { % Find the 'startxref' and associated position at the end of the file. % Position to read block of data from the end of the file. Note: We % actually end at the end of the last %%EOF since this is the end of the % useful PDF data. (Some files contain trailing garbage.) PDFfile 0 setfileposition PDFfile bytesavailable % size of file post_eof_count sub dup % location of end of last %%EOF dup 4096 .min % block size to read % stack: <useful file size> <useful file size file> <block size> % move file position to the start of the block 2 copy sub PDFfile exch setfileposition % read block of data dup string 0 1 4 -1 roll 1 sub { 2 copy PDFfile read pop put pop } for % search for last occurance of 'startxref' //false % Assume that startxref not present exch (startxref) { search not { exit } if % Exit loop when no more startxref's pop 3 -1 roll pop //true 3 1 roll % Idicate that we have found starxref } loop exch % Exch last string and 'found' flag { % determine where the startxref is in the file % 'startxref' loc = end loc - remaing string length - 9 bytes length sub 9 sub % move the file to this position and read startxref and position PDFfile exch setfileposition PDFfile token pop pop PDFfile token_no_close pop dup type /integertype eq not { pop % startxref not followed by integer. We will search the end of the file for trailer. PDFfilelen } if } { % startxref not found. We will search the end of the file for trailer. pop pop PDFfilelen } ifelse % compare xref position to 1/2 the length of the file and search for trailer exch 2 div lt { search_start_trailer dup 0 eq { pop search_end_trailer } if } { search_end_trailer dup 0 eq { pop search_start_trailer } if } ifelse dup 0 eq { pop ( **** Error: Trailer dictionary not found.\n) pdfformaterror ( Output may be incorrect.\n) pdfformaterror }{ % get the trailer dup PDFfile exch setfileposition % set to the specified trailer location /dictlevelcount 0 def PDFfile traileropdict .pdfrun % read trailer info { dup /Root known not { ( **** Warning: This trailer dictionary does not contain a /Root entry\n searching for a prior trailer.\n) pdfformatwarning %% remove trailer dict copy, duplicate file position, then remove the length of 'trailer' so we don't %% find the same one again.... pop dup 7 sub search_earlier_trailer }{exit}ifelse dup 0 eq { pop ( **** Error: Valid trailer not found.\n) pdfformaterror ( Output may be incorrect.\n) pdfformaterror }{ PDFfile exch setfileposition % set to the specified trailer location /dictlevelcount 0 def PDFfile traileropdict .pdfrun % read trailer info } ifelse } loop /Trailer exch def pop } ifelse } bind executeonly def % This routine will determine if there is stuff after the %%EOF. There is % supposed to be only a line termination. However many real life files % contain some garbage. This routine checks how much. We then ignore this % stuff when we are scanning for objects. /determine_post_eof_count % - determine_post_eof_count <count> { % Position to read block of data from the end of the file. PDFfilelen % size of file dup 4096 .min % file_size block_size dup 3 1 roll sub % block_size file_size-block_size PDFfile exch setfileposition % block_size string PDFfile exch readstring pop % () % search for last occurance of 'startxref', '%%EOF' is often damaged (startxref) search { pop { search not { exit } if pop } loop % how much is left = remaining string length % Now search for %%EO or try to read a number after 'startxref'. (%%EO) search { pop pop } { % Look for a number after startxref { dup token { pop exch pop } if } stopped pop } ifelse length } { % Can't even find startxref, assume it's all objects pop 0 } ifelse } bind executeonly def % This routine will scan a file searaching for object locations to build % an alternate version of the data in the xref tables. % Its purpose is to provide a basis for an xref fixing facility. /search_objects % - search_objects - { % Initialize the Objects, Generations, etc. arrays Objects dup length array copy /Orig_Objects exch def initPDFobjects % reset duplicate object and generation numbers error flag /dup_obj_gen_num //false def % Determine how many bytes are in the file after the final %%EOF /post_eof_count determine_post_eof_count def % Start at the beginning of the file PDFfile 0 setfileposition % Create a working string (and also store its length on stack). We are % using a maximum size string size the logic below wants a recovered object % to fit into our working string. 65535 dup string { % Now loop through the entire file looking for objects PDFfile fileposition % save current file position % When we get near the end of the file, we use a smaller interval of % our working string to prevent reading past the end. (See comments on % EOF testing below.) PDFfile bytesavailable post_eof_count sub 10 sub dup 4 index lt { 2 index 0 3 -1 roll getinterval % near EOF, use interval of string } { pop 1 index % not near end, use full working string } ifelse % Read a line from file. If the line does not fit into our working string, % or any other error, then we will discard it. PDFfile exch { readline } .internalstopped { pop pop //false } if % indicate no string if we stopped { % stack: <length> <working_str> <loc> <string> % Now that we have line, get obj num, ref num, and 'obj'. Verify that each % of these is correct type. dup (obj) search { % preliminary check for obj pop pop pop /integertype typed_token { % get obj number /integertype typed_token { % get ref number /nametype typed_token { % get 'obj' text pop % pop remaining string /obj eq { % verify name is 'obj' % make sure we have room in the arrays. We work in increments % of 20 each time we increase the size. 1 index 20 add 20 idiv 20 mul growPDFobjects % save xref parameters into ObjectStream, Objects and Generations 1 index 0 % rearrange parms for setxrefentry 4 index PDFoffset sub 3 index //true setxrefentry % save parameters pop pop pop pop % clear parameters } if % check if name is 'obj' } if % check if we got 'obj" string pop % remove ref number } if % check if we got ref number pop % remove obj number } if % check if we got object number } { pop pop } ifelse } if % check if got a string from readline pop % remove location % Check if we are approaching the end of the file. We do not want to % read past the end of the file since that closes it. We actually stop % 10-20 bytes early since there cannot be an object that close to the end. % (There is a Trailer dictionary, etc. at the end of the file.) PDFfile bytesavailable post_eof_count sub 20 lt { exit } if } loop % loop through the entire file pop pop % remove working string and its length % Output warning if we have two objects with the same object and generation % numbers. dup_obj_gen_num { ( **** Warning: There are objects with matching object and generation\n) pdfformatwarning ( **** numbers. The output may be incorrect.\n) pdfformatwarning } if currentdict /Orig_Objects undef } bind executeonly def % Print warning message because we found a problem while reading the xref % tables /print_xref_warning { ( **** Error: An error occurred while reading an XREF table.\n) pdfformaterror ( **** The file has been damaged. This may have been caused\n) pdfformaterror ( **** by a problem while converting or transfering the file.\n) pdfformaterror ( **** Ghostscript will attempt to recover the data.\n) pdfformaterror ( **** However, the output may be incorrect.\n) pdfformaterror } bind executeonly def % Attempt to recover the XRef data. This is called if we have a failure % while reading the normal XRef tables. This routine usually works % only for pre PDF1.5 versions of PDF files. /recover_xref_data % - recover_xref_data - { print_xref_warning % Print warning message count pdfemptycount sub { pop } repeat % remove anything left by readxref search_objects % Search for objects } bind executeonly def