How do I render *parts* of a svg file?

Question

Hello good people! :)

I want to render parts of a svg file by name but for the life of me I cannot figure out how to do so (using python + gtk).

Here's the svg file in question: http://david.bellot.free.fr/svg-cards/files/SVG-cards-2.0.1.tar.gz (Update: this file no longer exists, but you can track it down at http://svg-cards.sourceforge.net/)

On his site, David, says:

You can draw a card either by rendering the file onto a pixmap and clipping each card manually or by using the card's name through a DOM interface. All cards are embedded into a SVG group.

I don't know what he means by a DOM interface. I have done some searching and the best result I found that seems to fit what I want to do is:

QSvgRenderer *renderer = new QSvgRenderer(QLatin1String("SvgCardDeck.svg"));
QGraphicsSvgItem *black = new QGraphicsSvgItem();
QGraphicsSvgItem *red   = new QGraphicsSvgItem();

black->setSharedRenderer(renderer);
black->setElementId(QLatin1String("black_joker"));

red->setSharedRenderer(renderer);
red->setElementId(QLatin1String("red_joker"));

Notice however that it is for Qt and is not even written in python.

This is what I have so far:

#!/usr/bin/env python

from __future__ import absolute_import

import cairo
import gtk
import rsvg

from xml import xpath
from xml.dom import minidom

window = gtk.Window()
window.set_title("Foo")
window.set_size_request(256, 256)
window.set_property("resizable", False)
window.set_position(gtk.WIN_POS_CENTER)
window.connect("destroy", gtk.main_quit)
window.show()

document = minidom.parse("cards.svg")
element = xpath.Evaluate("//*[@id='1_club']", document)[0]
xml = element.toxml()

svg = rsvg.Handle()
svg.write(xml)

pixbuf = svg.get_pixbuf()

image = gtk.Image()
image.set_from_pixbuf(pixbuf)
image.show()

window.add(image)

gtk.main()

It doesn't work, of course.

What am I missing?

Answer 1

I believe that what he means by 'through a DOM interface' is that since SVG is XML, you could load the SVG file in minidom, or some other Python XML parser, and extract the XML node with the specific name that you are looking for. That XML node should represent an item that can be rendered.

Answer 2

You can do it by editing the tag. Edit width and height, set the viewBox attribute on the main svg element to the rectangle you want, render, repeat.

See http://stackoverflow.com/questions/1098317/svg-slicing and http://dingoskidneys.com/~dholth/svg/

Answer 3

The GTK library for rendering SVG is called RSVG. It has python bindings, but they are undocumented, and they don't wrap the rsvg_handle_get_pixbuf_sub() and rsvg_handle_render_cairo_sub() functions which you would normally use for that purpose in C. Here's what you have to do as far as I can tell. You extract the XML node as Adam Crossland suggested. To render it, you have to do something like this:

import gtk
import rsvg
handle = rsvg.Handle()
handle.write(buffer=xml_data) 
# xml_data is the XML string for the object you want
image = gtk.Image()
image.set_from_pixbuf(handle.get_pixbuf())

That's if you want it in a gtk.Image, otherwise do something else with the pixbuf. You can also render it to a Cairo context with handle.render_cairo(cr) where cr is your Cairo context.

EDIT:

Sorry, I didn't read the python bindings closely enough at first. The _sub() functions are implemented using the id= argument, so your program can boil down to this:

#!/usr/bin/env python

import gtk
import rsvg

window = gtk.Window()
window.set_title("Foo")
window.connect("destroy", gtk.main_quit)
window.show()

svg = rsvg.Handle(file='cards.svg')
pixbuf = svg.get_pixbuf(id='#3_diamond')

image = gtk.Image()
image.set_from_pixbuf(pixbuf)
image.show()

window.add(image)

gtk.main()

I tested this and it works. However, the window is the size of the entire SVG canvas, and is clipped to the screen size (which is why I rendered the 3 of diamonds instead of the ace of clubs which is up in the corner.) So you'll still have to find some way to crop the pixbuf around the card that you want, but that shouldn't be too hard.

Answer 4

Here's my answer to the cropping blank space problem. It's a rough hack but it worked great. This would also serve as a good starting point to get cards for anyone making a card game in python.

import gtk
import rsvg
svg = rsvg.Handle(file="/usr/share/gnome-games-common/cards/gnomangelo_bitmap.svg")
w, h = 202.5, 315
card_names = map(str, range(1,11)) + ["jack", "queen", "king"]
suites = ["club", "diamond", "heart", "spade"]
specials = [{"name":"black_joker","x":0, "y":4}, {"name":"red_joker","x":1, "y":4}, {"name":"back","x":2, "y":4}]
for suite_number, suite in enumerate(suites):
    for card_number, card in enumerate(card_names):
        print "processing", suite, card, '#'+card+'_'+suite
        pixbuf = svg.get_pixbuf(id='#'+card+'_'+suite)
        pixbuf.subpixbuf(int(w*card_number), int(h*suite_number), int(w), int(h)).save("./"+card+"_"+suite+".png","png", {})
for special in specials:
    print "processing", special["name"]
    pixbuf = svg.get_pixbuf(id='#'+special["name"])
    card_number = special["x"]
    suite_number = special["y"]
    pixbuf.subpixbuf(int(w*card_number), int(h*suite_number), int(w), int(h)).save("./"+special["name"]+".png","png", {})

Answer 5

QB89Dragon,

How did you figure this out. I have heard (and couldn't find) documentation for rsvg for python doesn't exist. How did you know these methods/functions and how they work within Python.

Because of this lack of documentation, I don't know what to do with them after the call to the subpixbug method (since there is a save, I'm sure there is a load, but how to do it). Also, it would also be great to output the png files to the hard drive.

Please give some tips on what to do to use these once buffered, how to save them, and how did you find out how to do this in the first place.

With thanks, Narnie