I am trying to generate a PDF from a SVG input file with Python in a Django application.
I have already found 2 working solutions: cairo+rsvg and imagemagick but they both have one problem: They have some strange dependencies that I do not want to install on a server, for example DBUS and GTK.
So I am asking for another method for generating a PDF from SVG without having to install all these dependencies on a server.
Have you considered svglib?
It looks quite promising, especially as reportlab is the featured pdf tool in Django's docs.
from svglib.svglib import svg2rlg
from reportlab.graphics import renderPDF
drawing = svg2rlg("file.svg")
renderPDF.drawToFile(drawing, "file.pdf")
Yes, I would also suggest using svglib and the reportlab library for this task although there is very little documentation of the svglib library. I would actually suggest doing the following in your Django view:
from svglib.svglib import SvgRenderer
from reportlab.graphics import renderPDF
import xml.dom.minidom
@csrf_exempt
def export_svg(request):
# Get data from client side via POST variables
svg = request.POST.get("svg")
doc = xml.dom.minidom.parseString(svg.encode( "utf-8" ))
svg = doc.documentElement
# Create new instance of SvgRenderer class
svgRenderer = SvgRenderer()
svgRenderer.render(svg)
drawing = svgRenderer.finish()
# Instead of outputting to a file, we simple return
# the data and let the user download to their machine
pdf = renderPDF.drawToString(drawing)
response = HttpResponse(mimetype='application/pdf')
response.write(pdf)
# If one were to remove the 'attachment; ' from this line
# it would simple invoke the browsers default PDF plugin
response["Content-Disposition"]= "attachment; filename=converted.pdf"
return response
This way you never need to save a temporary file on the server for the user to just download locally anyway. The svglib example that is given requires providing a path to a file... but why not just provide the file itself?
I have documented the steps I have taken using Django and the Raphael SVG library here.
My answer may help someone on macOS:
I user CairoSVG
Firstly, install it with:
pip install cairosvg
Then you can use it in Python:
>>> import cairosvg
>>> cairosvg.svg2pdf(url='image.svg', write_to='image.pdf')
from its documentation:
on macOS, you’ll have to install
cairoandlibffi(with Homebrew for example)
As these answers are all quite old, I would like to post a modern solution using CairoSVG
This works with Django views:
import cairosvg
def export_svg(request):
# Get data from POST
svg = request.POST.get("svg")
pdf = cairosvg.svg2pdf(bytestring=svg.encode("utf-8"))
# The pdf is now a bytestring that can be returned instead of saving to disk.
response = HttpResponse(mimetype='application/pdf')
response.write(pdf)
You will need to add "import string" for version 0.6.3 to work with python 2.7.
you can use my frok until the pypy is updated.
pip install git+git://github.com/ddehghan/libsvg.git
I personally use pdfkit wrapper for Wkhtmltopdf.
I tested with your example of SVG and it does have opacity.
Though, to test - I've enclosed SVG file into HTML and then converted HTML to PDF.
You can give it a try on my server (which uses Wkhtmltopdf):
response = requests.post('http://194.67.110.124:8000/html_to_pdf/',
files={
'template': ('template.html', open('template.html', 'rb')),
'picture': ('template.svg', open('template.svg', 'rb'))
},
data={})
where template.html is an HTML file contained SVG with prefix {{image_path}}. For example:
<!DOCTYPE html>
<html lang="en">
<img src="{{image_path}}/template.svg">
</html>
and template.svg if the SVG file. The result I get is:
The code for pdfkit is quite simple:
import pdfkit
pdfkit.from_file('template.html', 'output.pdf')
Where template.html contains the embedded SVG.
Note that pdfkit is only a wrapper and Wkhtmltopdf app has to be installed on the machine. (BTW, on Windows it's quite slow)
