Issue 127 - rst2pdf - Error if images already in PDF format

Issue 127:	Error if images already in PDF format
	1 person starred this issue and may be notified of changes.	Back to list

Status:	Fixed
Owner:	----
Closed:	Oct 13
Type-Enhancement
Priority-Low

Reported by ryankdale, Jun 05, 2009

If an .. image:: or .. figure:: directive refers to an image already in PDF
format, an error is returned and the image does not appear in the output.

Step-by-step test . . .

1. Create plot

from pylab import *
plot([1,2,3])
savefig('image.pdf')

2. Write test.rst file referring to image

title
=====

.. figure:: image.pdf
  
   here's the figure caption 

3. Run rst2pdf test.rst.  Results in the error,

[ERROR] createpdf.py:890 Error opening "image.pdf": cannot identify image file

It looks like the error is due to neither PIL nor UniConvertor being able
to read PDFs.

One workaround to have vector figures is to convert to SVG. But for my use
case (PDFs from the Genome Browser (http://genome.ucsc.edu) converting to
SVGs causes UniConvertor to fail (which is a separate issue I'm trying to
solve).

Best case scenario would be to put PDF-format images directly in the output.

This is using version 0.10.1 on Ubuntu 8.04

Comment 1 by roberto.alsina, Jun 05, 2009

I have asked in the reportlab list if there is a way to do this.

So far: 

* PIL (which we use for raster images) doesn't support reading PDF files

* uniconvertor (which we use for vectors) doesn't support reading PDF files

So, so far, I see no way to do this.

Even converting the pdf to ps or eps doesn't seem to work.

Status: Accepted

Comment 2 by roberto.alsina, Jun 05, 2009

After some research, I could cobble something using pythonmagick. I am reluctant to 
add a 4th possible dependency for images, but if I can make this work with and 
without it, I will implement it.

Comment 3 by roberto.alsina, Jun 13, 2009

I have an evaluation copy of PageCatcher, I will use it to implement this. 
PageCatcher is not free software, however, so I will try to do the pythonmagick one, 
too.

Comment 4 by roberto.alsina, Jun 17, 2009

As of r677, this is working via pythonmagick.

I have not been able to make pageCatcher do it yet, but maybe soon.

Status: Started

Comment 5 by ryankdale, Jun 18, 2009

Thanks for this.  For large PDF images, changing the image size with :width: 50% does
not seem to have an effect (r692).  Not sure whether to open another issue for this.

Just to clarify: pythonmagick converts PDF to raster before inserting into the
document.  In contrast, PageCatcher would insert the vector PDF directly.  Is this
correct?

Comment 6 by roberto.alsina, Jun 18, 2009

Yes, using PageCatcher it should work just like if it was all one seamless PDF, 
while with pythonmagick you are embedding just a picture of the PDF.

I'll take a look at the problem with :width:

Comment 7 by roberto.alsina, Jun 18, 2009

The :width: 50% does work for me, check test_image_units_pdf.txt in r694

Comment 8 by roberto.alsina, Aug 19, 2009

This seems to work correctly, although the PageCatcher support is broken.
Since PageCatcher is proprietary, this issue is now downgraded to "I keep it open 
just in case I am very very very bored someday".

Labels: -Priority-Medium Priority-Low

Comment 9 by roberto.alsina, Sep 22, 2009

(No comment was entered for this change.)

Labels: -Type-Defect Type-Enhancement

Comment 10 by roberto.alsina, Oct 13, 2009

Closing it, I will not work on this issue anymore.

If someone is familiar with pagecatcher, of course, feel free to fix it :-)

Status: Fixed