Misc
defund submitted a math paper to a research conference and received a few comments from the editors. Unfortunately, we only have a fragment of the returned paper.
Author: defund
We have a PDF with no trailer / xref table
$ exiftool paper_cut.pdf
ExifTool Version Number : 11.11
File Name : paper_cut.pdf
Directory : .
File Size : 32 kB
File Modification Date/Time : 2019:04:21 22:23:44+08:00
File Access Date/Time : 2019:04:22 00:00:00+08:00
File Inode Change Date/Time : 2019:04:21 22:24:06+08:00
File Permissions : rw-r--r--
File Type : PDF
File Type Extension : pdf
MIME Type : application/pdf
PDF Version : 1.3
Linearized : No
Warning : Invalid xref table
Usually some corrupted PDFs can be rendered by Ghostscript. But in this case it did not process the file.
$ gs -o repaired.pdf -sDEVICE=pdfwrite -dPDFSETTINGS=/prepress paper_cut.pdf
GPL Ghostscript 9.25 (2018-09-13)
Copyright (C) 2018 Artifex Software, Inc. All rights reserved.
This software comes with NO WARRANTY: see the file PUBLIC for details.
**** Error: Cannot find a 'startxref' anywhere in the file.
Output may be incorrect.
**** Error: An error occurred while reading an XREF table.
**** The file has been damaged. This may have been caused
**** by a problem while converting or transfering the file.
**** Ghostscript will attempt to recover the data.
**** However, the output may be incorrect.
**** Error: Trailer dictionary not found.
Output may be incorrect.
No pages will be processed (FirstPage > LastPage).
**** This file had errors that were repaired or ignored.
**** Please notify the author of the software that produced this
**** file that it does not conform to Adobe's published PDF
**** specification.
**** The rendered output from this file may be incorrect.
If we look in the PDF file, we realise that there is a stream but no endstream.
This means that we can try to manually parse the stream of whatever content is left.
We also see it has a encoding of FlateDecode.
Reference for PDF file format (content stream):
- https://pspdfkit.com/blog/2018/pdf-text-extraction/
- http://what-when-how.com/itext-5/parsing-pdfs-part-1-itext-5/
- https://commandlinefanatic.com/cgi-bin/showarticle.cgi?article=art019
- https://labs.appligent.com/pdfblog/pdf-hello-world/
- https://resources.infosecinstitute.com/pdf-file-format-basic-structure/
Extracting FlateDecode content stream from PDF:
- http://binaryhax0r.blogspot.com/2010/06/flatedecode-decoder.html
- https://stackoverflow.com/questions/20620374/how-to-inflate-a-partial-zlib-file
- https://stackoverflow.com/questions/42791781/uncompress-and-save-zlib-data-in-pdf-with-python
- https://forums.adobe.com/thread/1663978
- https://gist.github.com/averagesecurityguy/ba8d9ed3c59c1deffbd1390dafa5a3c2
So putting it into a Python script, I extracted the content stream and also all the text.
This leaves me with the following text.
However, there is no flag to be seen.
This was when I realised that there are a whole bunch of coordinates after all the texts have been rendered.
Content Stream Operators
- https://www.adobe.com/content/dam/acom/en/devnet/pdf/pdfs/pdf_reference_archives/PDFReference.pdf (Page 700)
- https://github.com/galkahana/HummusJS/wiki/Use-the-pdf-drawing-operators
From the operators, it appears to be drawing operators at the defined coordinates.
With this, I extracted all coordinates and then simply plotted it on a graph
From this, we can somewhat see the flag. But then it is all in discrete points which is hard to make out.
Here I had to look properly at the operators in the content stream.
// Extracted from Adobe PDF reference document
m moveto Begin new subpath
l lineto Append straight line segment to path
c curveto Append curved segment to path (three control points)
I implemented the line and curves on a Canvas in the script and thus got the following
Final script
actf{worthy_of_the_fields_medal}