Screenshot of multiple webpages in a headerless QtWebKit browser in Python

pyqt4, python, qt, qtwebkit

Solution

Pretty much as described in Luke's answer, I shuffled things around to avoid creating a `QApplication` instance for each `Render`

Not the tidiest, but works for me:

import re
import sys
import time

# Tested with PySide 1.0.9, changing imports to PyQt should work identically
from PySide.QtCore import Qt, QUrl
from PySide.QtGui import QApplication, QImage, QPainter
from PySide.QtWebKit import QWebPage, QWebSettings


def os_safe_name(url):
    url = re.sub("[^a-zA-Z0-9_-]+", "_", url)
    url = re.sub("_{2,}", "_", url)
    return url


class Render(QWebPage):
    def __init__(self, url):
        QWebPage.__init__(self)

        self.url = url
        self.finished = False

        # Settings
        s = self.settings()
        #s.setAttribute(QWebSettings.AutoLoadImages, False)
        s.setAttribute(QWebSettings.JavascriptCanOpenWindows, False)
        s.setAttribute(QWebSettings.PluginsEnabled, True)

        #self.mainFrame().setScrollBarPolicy(Qt.Vertical, Qt.ScrollBarAlwaysOff)
        self.mainFrame().setScrollBarPolicy(Qt.Horizontal, Qt.ScrollBarAlwaysOff)

        # When page is loaded, callback saves image to file
        self.loadFinished.connect(self._loadFinished)

        self.mainFrame().load(QUrl(url))

    def _loadFinished(self, result):
        frame = self.mainFrame()
        size = frame.contentsSize()
        size.setWidth(1366)
        self.setViewportSize(size)

        image = QImage(self.viewportSize(), QImage.Format_ARGB32)

        painter = QPainter(image)
        frame.render(painter)
        painter.end()

        self.filepath = "output/%s.png" % os_safe_name(self.url)
        image.save(self.filepath)

        self.finished = True


def run(url, app = None):
    if app is None:
        app = QApplication(sys.argv)

    r = Render(url)

    while not r.finished:
        app.processEvents()
        time.sleep(0.01)

    return r.filepath


if __name__ == '__main__':
    app = QApplication(sys.argv)

    print run("http://stackoverflow.com", app=app)
    print run("http://google.com", app=app)

Problem

Original source