ReportLab

Note: this code needs a running installation of reportlab as well as of pillow. The latter needs some image processing libraries to function properly. See the associated blogpost for installation instructions

The reportlab user guide is here and the Pillow guide is here

1. Tools and Assets

Display hook for `pillow` images

Normally, pillow images are in iPython only displayed as object description strings. This code inserts a hook that automatically draws them, akin to the matplotlib interactive backends. source



In [27]:

    
from io import BytesIO
from IPython.core import display
from PIL import Image


def display_pil_image(im):
   """Displayhook function for PIL Images, rendered as PNG."""

   b = BytesIO()
   im.save(b, format='png')
   data = b.getvalue()

   ip_img = display.Image(data=data, format='png', embed=True)
   return ip_img._repr_png_()

png_formatter = get_ipython().display_formatter.formatters['image/png']
dpi = png_formatter.for_type(Image.Image, display_pil_image)

Assets

we need to get the following assets into the local directory for the code to run (once the assets are there the wget statements can be commented out)



In [28]:

    
!wget http://blogs.oditorium.com/technology/wp-content/uploads/2014/08/logo-oditorium-whitebg.jpg
!wget http://blogs.oditorium.com/technology/wp-content/uploads/2014/08/graph.png









    



--2014-10-12 23:49:42--  http://blogs.oditorium.com/technology/wp-content/uploads/2014/08/logo-oditorium-whitebg.jpg
Resolving blogs.oditorium.com (blogs.oditorium.com)... 66.147.244.110
Connecting to blogs.oditorium.com (blogs.oditorium.com)|66.147.244.110|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 28539 (28K) [image/jpeg]
Saving to: ‘logo-oditorium-whitebg.jpg.3’

100%[======================================>] 28,539       155KB/s   in 0.2s   

2014-10-12 23:49:42 (155 KB/s) - ‘logo-oditorium-whitebg.jpg.3’ saved [28539/28539]

--2014-10-12 23:49:43--  http://blogs.oditorium.com/technology/wp-content/uploads/2014/08/graph.png
Resolving blogs.oditorium.com (blogs.oditorium.com)... 66.147.244.110
Connecting to blogs.oditorium.com (blogs.oditorium.com)|66.147.244.110|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 6091 (5.9K) [image/png]
Saving to: ‘graph.png.3’

100%[======================================>] 6,091       --.-K/s   in 0s      

2014-10-12 23:49:43 (166 MB/s) - ‘graph.png.3’ saved [6091/6091]

2. Reportlab - `canvas`

Writing text to the canvas

This code fragment generates a one-page file t1.pdf with some text lines written onto it using drawString



In [29]:

    
from reportlab.pdfgen import canvas

msg = "This is not a message"
c = canvas.Canvas("test1.pdf")
for i in range(1,10):
    c.drawString(50*i,50*i,msg)
c.showPage()
c.save()

!ls -l test1.pdf









    



-rw-r--r--  1 skl  staff  1595 12 Oct 23:49 test1.pdf

Writing images from disk to the canvas

The function drawImage writes an image from disk to the canvas. An image written that way is stored in the pdf file only once, which makes this important for multiple-use images, eg logos. There is another function drawInlineImage that places an image directly into the pdf flow. This can be slightly faster, but it will duplicate storage needs for repeating images



In [30]:

    
msg = "This is really not a message"
c = canvas.Canvas("test2.pdf")
for i in range(1,10):
    c.drawString(50*i,50*i,msg)
c.drawImage('graph.png',100,100)
c.showPage()
c.save()

!ls -l test2.pdf









    



-rw-r--r--  1 skl  staff  7615 12 Oct 23:49 test2.pdf

Writing images to the canvas on the fly

Often we generate images on the fly - eg using matplotlib - and dont want to save them to include them. The following code shows how this is done source



In [31]:

    
import matplotlib.pyplot as plt
from io import BytesIO
from reportlab.pdfgen import canvas
from reportlab.lib.units import inch, cm
from reportlab.lib.utils import ImageReader

fig = plt.figure(figsize=(4, 3))
plt.plot([1,4,3,2])
plt.ylabel('some numbers')

imgdata = BytesIO()
fig.savefig(imgdata, format='png')
imgdata.seek(0)  # rewind the data

Image = ImageReader(imgdata)

c = canvas.Canvas('test3.pdf')
c.drawImage(Image, cm, cm, inch, inch)
c.save()

!ls -l test3.pdf









    



-rw-r--r--  1 skl  staff  7914 12 Oct 23:49 test3.pdf

3. Reportlab - `platypus`

Platypus is reportlab's layout engine (see the user guide from page 59). It defines the following hierarchy of templates

DocTemplate - a DocTemplate describes the overall document
PageTemplate - a PageTemplate describes a page in the document; usually different pages of the same document share the same template, but for example the title page might be different from the rest
Frame - a Frame is a layout are within a page; in the most simple case there will be one frame per page, but eg for 2col layouts there will be 2 etc
Flowable - a Flowable is something that can flow into a frame, eg text or an image

Different page templates, different styles

The functions myTitlePage and myRegularPage define page templates. The command styles = getSampleStyleSheet() gets a whole list of styles that can be passed into the Paragraph function.



In [32]:

    
def lipsum():
    return "Lorem ipsum dolor sit amet, consectetur adipiscing elit. Aliquam ultrices ligula et libero tempus, ac pretium velit ultricies. Pellentesque sit amet vestibulum quam. Maecenas turpis ante, feugiat eu ultricies feugiat, ultricies ac elit. Praesent eleifend, nibh eu tempor consequat, nisi nunc hendrerit mi, at rhoncus massa sem quis nulla. Nunc ullamcorper mi a risus pretium, ac faucibus massa vehicula. Vestibulum venenatis aliquam felis eget hendrerit. Nulla porta massa placerat velit ultrices dictum. Curabitur mattis, lacus in convallis porta, ligula enim dignissim est, vel aliquam elit metus nec dolor. Vestibulum lacinia ac magna adipiscing iaculis. Suspendisse potenti. Nunc adipiscing magna id suscipit viverra. Sed tristique tortor ac erat mattis aliquam. Etiam nunc libero, iaculis non lectus quis, tincidunt adipiscing lacus. Aliquam in auctor dui."
#lipsum()



In [33]:

    
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer
from reportlab.lib.styles import getSampleStyleSheet
from reportlab.rl_config import defaultPageSize
from reportlab.lib.units import inch

PAGE_HEIGHT=defaultPageSize[1]; PAGE_WIDTH=defaultPageSize[0]
styles = getSampleStyleSheet()

Title = "<Document Title>"
def myTitlePage(canvas, doc):
    canvas.saveState()
    canvas.setFont('Times-Bold',48)
    canvas.drawCentredString(PAGE_WIDTH/2.0, PAGE_HEIGHT-108, Title)
    canvas.restoreState()

pageinfo = " - <some document info here>"
def myRegularPage(canvas, doc):
    canvas.saveState()
    canvas.setFont('Times-Roman',9)
    canvas.drawString(inch, 0.75 * inch, "page %d %s" % (doc.page, pageinfo))
    canvas.restoreState()



In [34]:

    
!rm test4.pdf
doc = SimpleDocTemplate("test4.pdf")
Story = [Spacer(1,2*inch)]
style = styles["Normal"]
styleH1 = styles["Heading1"]
styleH2 = styles["Heading2"]

for i in range(100): 
    if i % 10 == 0: 
        p = Paragraph("Headline 1", styleH1)
        Story.append(p)
    
    if i % 3 == 0: 
        p = Paragraph("Headline 1", styleH2)
        Story.append(p)
        
    bogustext = lipsum()
    p = Paragraph(bogustext, style)
    Story.append(p)
    Story.append(Spacer(1,0.2*inch))

doc.build(Story, onFirstPage=myTitlePage, onLaterPages=myRegularPage)

!ls -l test4.pdf









    



-rw-r--r--  1 skl  staff  29097 12 Oct 23:49 test4.pdf

Flowables and frames

It is possible to construct a document solely based on flowables and frames



In [35]:

    
!rm test5.pdf

from reportlab.pdfgen.canvas import Canvas
from reportlab.lib.styles import getSampleStyleSheet
from reportlab.lib.units import inch
from reportlab.platypus import Paragraph, Frame
styles = getSampleStyleSheet()
styleN = styles['Normal']
styleH = styles['Heading1']

story = []
story.append(Paragraph("This is a Heading",styleH))
for i in range(100):
    story.append(Paragraph(lipsum(),styleN))
    story.append(Spacer(1,0.2*inch))
c = Canvas('test5.pdf')
f = Frame(3*inch, inch, 3*inch, 9*inch, showBoundary=1)
f.addFromList(story,c)
c.save()

!ls -l test5.pdf









    



-rw-r--r--  1 skl  staff  2388 12 Oct 23:49 test5.pdf

Two column layout

source



In [36]:

    
from reportlab.platypus import BaseDocTemplate, Frame, Paragraph, PageBreak, PageTemplate
from reportlab.lib.styles import getSampleStyleSheet
import random

words = "lorem ipsum dolor sit amet consetetur sadipscing elitr sed diam nonumy eirmod tempor invidunt ut labore et".split()

styles=getSampleStyleSheet()
story=[]

doc = BaseDocTemplate('test6.pdf',showBoundary=1)

frame1 = Frame(doc.leftMargin, doc.bottomMargin, doc.width/2-6, doc.height, id='col1')
frame2 = Frame(doc.leftMargin+doc.width/2+6, doc.bottomMargin, doc.width/2-6, doc.height, id='col2')

story.append(Paragraph(" ".join([random.choice(words) for i in range(1000)]),styles['Normal']))
doc.addPageTemplates([PageTemplate(id='TwoCol',frames=[frame1,frame2]), ])

doc.build(story)

!ls -l test6.pdf









    



-rw-r--r--  1 skl  staff  4015 12 Oct 23:49 test6.pdf

Mixed layout, one and two cols, title and back