Reportlab: Converting Hundreds of Images Into PDFs

I was recently asked to convert a few hundred images into PDF pages. A friend of mine draws comics and my brother wanted to be able to read them on a tablet. Alas, if you had a bunch of files named something like this:


'Jia_01.Jpg', 'Jia_02.Jpg', 'Jia_09.Jpg', 'Jia_10.Jpg', 'Jia_11.Jpg', 'Jia_101.Jpg'


the Android tablet would reorder them into something like this:


'Jia_01.Jpg', 'Jia_02.Jpg', 'Jia_09.Jpg', 'Jia_10.Jpg', 'Jia_101.Jpg', 'Jia_11.Jpg'


And it got pretty confusing the more files you had that were out of order. Sadly, even Python sorts files this way. I tried using the glob module on the directly and then sorting the result and got the exact same issue. So the first thing I had to do was find some kind of sorting algorithm that could sort them correctly. It should be noted that Windows 7 can sort the files correctly in its file system, even though Python cannot.

After a little searching on Google, I found the following script on StackOverflow:

import re

#----------------------------------------------------------------------
def sorted_nicely( l ): 
    """     
    Sort the given iterable in the way that humans expect.
    """ 
    convert = lambda text: int(text) if text.isdigit() else text 
    alphanum_key = lambda key: [ convert(c) for c in re.split('([0-9]+)', key) ] 
    return sorted(l, key = alphanum_key)

That worked perfectly! Now I just had to find a way to put each comic page on their own PDF page. Fortunately, the reportlab library makes this pretty easy to accomplish. You just need to iterate over the images and insert them one at a time onto a page. It's easier to just look at the code, so let's do that:

import glob
import os
import re

from reportlab.lib.pagesizes import letter
from reportlab.platypus import SimpleDocTemplate, Paragraph, Image, PageBreak
from reportlab.lib.units import inch

#----------------------------------------------------------------------
def sorted_nicely( l ): 
    """ 
    # http://stackoverflow.com/questions/2669059/how-to-sort-alpha-numeric-set-in-python

    Sort the given iterable in the way that humans expect.
    """ 
    convert = lambda text: int(text) if text.isdigit() else text 
    alphanum_key = lambda key: [ convert(c) for c in re.split('([0-9]+)', key) ] 
    return sorted(l, key = alphanum_key)

#----------------------------------------------------------------------
def create_comic(fname, front_cover, back_cover, path):
    """"""
    filename = os.path.join(path, fname + ".pdf")
    doc = SimpleDocTemplate(filename,pagesize=letter,
                            rightMargin=72,leftMargin=72,
                            topMargin=72,bottomMargin=18)
    Story=[]
    width = 7.5*inch
    height = 9.5*inch    
    
    pictures = sorted_nicely(glob.glob(path + "\\%s*" % fname))
    
    Story.append(Image(front_cover, width, height))
    Story.append(PageBreak())
    
    x = 0
    page_nums = {100:'%s_101-200.pdf', 200:'%s_201-300.pdf',
                 300:'%s_301-400.pdf', 400:'%s_401-500.pdf',
                 500:'%s_end.pdf'}
    for pic in pictures:
        parts = pic.split("\\")
        p = parts[-1].split("%s" % fname)
        page_num = int(p[-1].split(".")[0])
        print "page_num => ", page_num
        
        im = Image(pic, width, height)
        Story.append(im)
        Story.append(PageBreak())
        
        if page_num in page_nums.keys():
            print "%s created" % filename 
            doc.build(Story)
            filename = os.path.join(path, page_nums[page_num] % fname)
            doc = SimpleDocTemplate(filename,
                                    pagesize=letter,
                                    rightMargin=72,leftMargin=72,
                                    topMargin=72,bottomMargin=18)
            Story=[]
        print pic
        x += 1
        
    Story.append(Image(back_cover, width, height))
    doc.build(Story)
    print "%s created" % filename
    
#----------------------------------------------------------------------
if __name__ == "__main__":
    path = r"C:\Users\Mike\Desktop\Sam's Comics"
    front_cover = os.path.join(path, "FrontCover.jpg")
    back_cover = os.path.join(path, "BackCover2.jpg")
    create_comic("Jia_", front_cover, back_cover, path) 

Let's break this down a bit. As usual, you have some necessary imports that are required for this code to work. You'll note we also have that sorted_nicely function that we talked about earlier is in this code too. The main function is called create_comic and takes four arguments: fname, front_cover, back_cover, path. If you have used the reportlab toolkit before, then you'll recognize the SimpleDocTemplate and the Story list as they're straight out of the reportlab tutorial.

Anyway, you loop over the sorted pictures and add the image to the Story along with a PageBreak object. The reason there's a conditional in the loop is because I discovered that if I tried to build the PDF with all 400+ images, I would run into a memory error. So I broke it up into a series of PDF documents that were 100 pages or less. At the end of the document, you have to call the doc object's build method to actually create the PDF document.

Now you know how I ended up writing a whole slew of images into multiple PDF documents. Theoretically, you could use PyPdf to knit all the resulting PDFs together into one PDF, but I didn't try it. You might end up with another memory error. I'll leave that as an exercise for the reader.

Source Code

• comic_maker.zip
• comic_maker.tar