My page load flickers the background image. Is there a way I can sync it so that the redraw is in one. There is basically a couple of paragraphs of text with an image bg, but the flicker happens. It's very odd.

is there anyway to modify the request header?

i need the signal "enable-default-context-menu", i will be implemented in the next release ?

Is there anyway to use proxy with it?

@gg...@live.com.ar

You need to use settings for that (http://webkitgtk.org/reference/webkitgtk-webkitwebview.html#WebKitWebView--settings) A webkit.WebSettings? object will have that as a property.

navigation-requested does not work for assets like images, JS, CSS, videos, etc.

Use resource-request-starting. Like this:

web_view.connect('resource-request-starting', resource_cb)

def resource_cb(view, frame, resource, request, response):
    print request.get_uri()
    request.set_uri('...whatever...')

Note that you can't feed back arbitrary data. You can only fiddle with the URL at this point. But that's okay, because through the URL you can do one of two things:

1. Encode your data and set a data: URI. 2. Write your data to a temp file and set a file: URI that points to the temp file.

I get the content of a web page by get_html() function, it shows like this: ... <script>document.location='http://www.someelse.com'</script> ...

Now, Can I get the content of the ready-to-redirect site www.someelse.com?

Where can I find a list of all the available signals on a webview widget?

http://webkitgtk.org/reference/webkitgtk-webkitwebview.html#webkitgtk-webkitwebview.signals

@sassanh

I use this code to use proxy::

import gtk, webkit
import ctypes
libgobject = ctypes.CDLL('/usr/lib/libgobject-2.0.so.0')
libwebkit = ctypes.CDLL('/usr/lib/libsoup-2.4.so.1')
libsoup = ctypes.CDLL('/usr/lib/libsoup-2.4.so.1')
libwebkit = ctypes.CDLL('/usr/lib/libwebkit-1.0.so')

proxy_uri = libsoup.soup_uri_new('http://127.0.0.1:8000')

session = libwebkit.webkit_get_default_session()
libgobject.g_object_set(session, "proxy-uri", proxy_uri, None)

import gtk, webkit; w = gtk.Window() ; s = gtk.ScrolledWindow() ; v = webkit.WebView() ; s.add(v); w.add(s) ; w.show_all()

v.open('http://www.youtube.com')

I can't figure out how to use the navigation_requested_cb hack :

import webkit
import gtk

window = gtk.Window()
view = webkit.WebView()

window.connect("navigation-requested",
		_navigation_requested_cb,
		view,
		view.get_main_frame(),
		?????)

How/where can I get the NetworkRequest??

How do you install pywebkitgtk on centos/redhat?

charlie,

in later versions (1.1.x I seem to remember) you need to guard against loops by using a class attribute to hold the currently requested uri and chacking for it... oh and just connect to the WebView? navigation-requested and the callback args will be... (view,frame,net_request) the net_request is your NetworkRequest? object.. see below.

class BrowserPage(webkit.WebView):
    def __init__(self):
        webkit.WebView.__init__(self)
        self.connect("navigation-requested",self._nav_request_cb)
        self.l_uri=None
    def _nav_request_cb(self,view,frame,net_req):
        uri=net_req.get_uri()
        if uri==self.l_uri:
            print "same uri"
            return 2
        print "getting %s" % uri
        self.l_uri=uri
        page=urllib.urlopen(uri)
        print "load to frame"
        frame.load_string(page.read(),"text/html","iso-8859-15",page.geturl())
        print "[done]"
        return 1

note: now uses frame.load_string instead of view.load_string to preserve frames..

or better still (as navigation-requested is deprecated...)

class BrowserPage(webkit.WebView):
    def __init__(self):
        webkit.WebView.__init__(self)
        self.connect("navigation-policy-decision-requested",self._nav_request_policy_decision_cb)
        self.l_uri=None
    def _nav_request_policy_decision_cb(self,view,frame,net_req,nav_act,pol_dec):
        uri=net_req.get_uri()
        if uri==self.l_uri:
            pol_dec.use()
            return True
        if uri.startswith('about:'):
            return False
        self.l_uri=uri
        page=urllib.urlopen(uri)
        frame.load_string(page.read(),"text/html","iso-8859-15",page.geturl())
        pol_dec.ignore()
        return True

Is there a way to store/remember cookies? For example, if I create a PyWebkitGTK app and use it to log into a website, I would like to remain logged in to that website the next time I start the app.

I would like to use webkit get webpage html. I try python code in console it work, but change to file it isn't. This is my browser.py

#!/usr/bin/python

import sys
import webkit

class WebView(webkit.WebView):
    def get_html(self):
        self.execute_script('oldtitle=document.title;document.title=document.documentElement.innerHTML;')
        html = self.get_main_frame().get_title()
        self.execute_script('document.title=oldtitle;')
        return html
    
if __name__ == '__main__':
    if len(sys.argv) < 2:
        print "plz use # browser.py <url>"
        sys.exit()
    
    print "url: ", sys.argv[1]
    browser = WebView()
    browser.open(sys.argv[1])
    print browser.get_html()

I don't need to show browser because it is console application. How can I do. Thank.

RE : Is there a way to store/remember cookies? For example, if I create a PyWebkitGTK app and use it to log into a website, I would like to remain logged in to that website the next time I start the app.

You can get cookies stored, from example above, read documnet.cookie instead document.documentElement.innerHTML;

-self.execute_script('oldtitle=document.title;document.title=document.documentElement.innerHTML;') +self.execute_script('oldtitle=document.title;document.title=document.cookie;')

to boatkrap, Q : I don't need to show browser because it is console application. How can I do ? A: use a virtual frame buffer Xvfb :1 -fbdir /var/tmp/ & export DISPLAY=:1 python webkit , it works I try myself.

thanks for the heads up on how to store cookies ...

but if i have a cookie stored on my machine, how do i make sure webkit knows to load those cookies when i point it to a web page?

I tried this other method to get the HTML of the document:

import jswebkit

  def get_html(self):
    frame = self.viewer.get_main_frame()
    ctx = jswebkit.JSContext(frame.get_global_context())
    text = ctx.EvaluateScript("document.body.innerHTML")
    return text

It works very well, and it's better than putting the whole HTML into the title. Also, with this method you can access only certain parts of the HTML code, and have some DOM interaction via javascript. You have to install the "python-jswebkit" package.

Using something like this from mbull:

class BrowserPage?(webkit.WebView?):

def init(self):

webkit.WebView?.init(self) self.connect("navigation-policy-decision-requested",self.nav_request_policy_decision_cb) self.l_uri=None

def nav_request_policy_decision_cb(self,view,frame,net_req,nav_act,pol_dec):

uri=net_req.get_uri() if uri==self.l_uri:

pol_dec.use() return True

if uri.startswith('about:'):

return False

self.l_uri=uri page=urllib.urlopen(uri) frame.load_string(page.read(),"text/html","iso-8859-15",page.geturl()) pol_dec.ignore() return True

The problem I'm running into is that this always does a GET. urlopen will do a POST if it's presented with data, but I'm not sure how to: a) find out if we've been asked to do a post b) get the user data from inside the navigation-policy-decision-requested callback

I'm having the same issue. I dug into the library and bindings and it looks like the info should be available via NetworkRequest?, but that the binding hasn't got those methods fleshed out yet.

dwil, et al, I think I've come up with a better cookie solution using ctypes- the details can be found on StackOverflow?? (Python webkit WebView remember cookies?, How to clear cookies in webkit) and tailored to your use case.