def download(self, url, outfd=None):
"""
Download a given URL using current cookies.
@param url: URL or path to download
@param outfd: Output file-like stream. If None, return data string.
@return: Bytes downloaded (None if something went wrong)
@note: If url is a path, the current base URL will be pre-appended.
"""
def _on_reply(reply):
url = unicode(reply.url().toString())
self._download_reply_status = not bool(reply.error())
self._download_reply_status = None
if not urlparse.urlsplit(url).scheme:
url = urlparse.urljoin(self.url, url)
request = QNetworkRequest(QUrl(url))
# Create a new manager to process this download
manager = QNetworkAccessManager()
reply = manager.get(request)
if reply.error():
raise SpynnerError("Download error: %s" % reply.errorString())
reply.downloaded_nbytes = 0
manager.setCookieJar(self.manager.cookieJar())
manager.connect(manager, SIGNAL('finished(QNetworkReply *)'), _on_reply)
outfd_set = bool(outfd)
if not outfd_set:
outfd = StringIO()
self._start_download(reply, outfd)
while self._download_reply_status is None:
self._events_loop()
if outfd_set:
return (reply.downloaded_nbytes if not reply.error() else None)
else:
return outfd.getvalue()
def download(self, url, outfd=None):
"""
Download a given URL using current cookies.
@param url: URL or path to download
@param outfd: Output file-like stream. If None, return data string.
@return: Bytes downloaded (None if something went wrong)
@note: If url is a path, the current base URL will be pre-appended.
"""
def _on_reply(reply):
url = unicode(reply.url().toString())
self._download_reply_status = not bool(reply.error())
self._download_reply_status = None
if not urlparse.urlsplit(url).scheme:
url = urlparse.urljoin(self.url, url)
request = QNetworkRequest(QUrl(url))
# Create a new manager to process this download
manager = QNetworkAccessManager()
reply = manager.get(request)
if reply.error():
raise SpynnerError("Download error: %s" % reply.errorString())
reply.downloaded_nbytes = 0
manager.setCookieJar(self.manager.cookieJar())
manager.connect(manager, SIGNAL('finished(QNetworkReply *)'),
_on_reply)
outfd_set = bool(outfd)
if not outfd_set:
outfd = StringIO()
self._start_download(reply, outfd)
while self._download_reply_status is None:
self._events_loop()
if outfd_set:
return (reply.downloaded_nbytes if not reply.error() else None)
else:
return outfd.getvalue()
class Browser:
"""
Stateful programmatic web browser class based upon QtWebKit.
>>> browser = Browser()
>>> browser.load("http://www.wordreference.com")
>>> browser.runjs("console.log('I can run Javascript!')")
>>> browser.runjs("_jQuery('div').css('border', 'solid red')") # and jQuery!
>>> browser.select("#esen")
>>> browser.fill("input[name=enit]", "hola")
>>> browser.click("input[name=b]", wait_load=True)
>>> print browser.url, len(browser.html)
>>> browser.close()
"""
ignore_ssl_errors = True
"""@ivar: If True, ignore SSL certificate errors."""
user_agent = None
"""@ivar: User agent for requests (see QWebPage::userAgentForUrl for details)"""
jslib = "_jQuery"
"""@ivar: Library name for jQuery library injected by default to pages."""
download_directory = "."
"""@ivar: Directory where downloaded files will be stored."""
debug_stream = sys.stderr
"""@ivar: File-like stream where debug output will be written."""
debug_level = ERROR
"""@ivar: Debug verbose level (L{ERROR}, L{WARNING}, L{INFO} or L{DEBUG})."""
event_looptime = 0.01
"""@ivar: Event loop dispatcher loop delay (seconds)."""
_javascript_files = ["jquery.min.js", "jquery.simulate.js"]
_javascript_directories = [
os.path.join(os.path.dirname(__file__), "../javascript"),
os.path.join(sys.prefix, "share/spynner/javascript"),
]
def __init__(self, qappargs=None, debug_level=None):
"""
Init a Browser instance.
@param qappargs: Arguments for QApplication constructor.
@param debug_level: Debug level logging (L{ERROR} by default)
"""
self.application = QApplication(qappargs or [])
"""PyQt4.QtGui.Qapplication object."""
if debug_level is not None:
self.debug_level = debug_level
self.webpage = QWebPage()
"""PyQt4.QtWebKit.QWebPage object."""
self.webpage.userAgentForUrl = self._user_agent_for_url
self.webframe = self.webpage.mainFrame()
"""PyQt4.QtWebKit.QWebFrame main webframe object."""
self.webview = None
"""PyQt4.QtWebKit.QWebView object."""
self._url_filter = None
self._html_parser = None
# Javascript
directory = _first(self._javascript_directories, os.path.isdir)
if not directory:
raise SpynnerError("Cannot find javascript directory: %s" %
self._javascript_directories)
self.javascript = "".join(
open(os.path.join(directory, fn)).read()
for fn in self._javascript_files)
self.webpage.javaScriptAlert = self._javascript_alert
self.webpage.javaScriptConsoleMessage = self._javascript_console_message
self.webpage.javaScriptConfirm = self._javascript_confirm
self.webpage.javaScriptPrompt = self._javascript_prompt
self._javascript_confirm_callback = None
self._javascript_confirm_prompt = None
# Network Access Manager and cookies
self.manager = QNetworkAccessManager()
"""PyQt4.QtNetwork.QTNetworkAccessManager object."""
self.manager.createRequest = self._manager_create_request
self.webpage.setNetworkAccessManager(self.manager)
self.cookiesjar = _ExtendedNetworkCookieJar()
"""PyQt4.QtNetwork.QNetworkCookieJar object."""
self.manager.setCookieJar(self.cookiesjar)
self.manager.connect(
self.manager,
SIGNAL("sslErrors(QNetworkReply *, const QList<QSslError> &)"),
self._on_manager_ssl_errors)
self.manager.connect(self.manager, SIGNAL('finished(QNetworkReply *)'),
self._on_reply)
self.manager.connect(
self.manager,
SIGNAL(
'authenticationRequired(QNetworkReply *, QAuthenticator *)'),
self._on_authentication_required)
self._operation_names = dict(
(getattr(QNetworkAccessManager, s + "Operation"), s.lower())
for s in ("Get", "Head", "Post", "Put"))
# Webpage slots
self._load_status = None
self._replies = 0
self.webpage.setForwardUnsupportedContent(True)
self.webpage.connect(self.webpage,
SIGNAL('unsupportedContent(QNetworkReply *)'),
self._on_unsupported_content)
self.webpage.connect(self.webpage, SIGNAL('loadFinished(bool)'),
self._on_load_finished)
self.webpage.connect(self.webpage, SIGNAL("loadStarted()"),
self._on_load_started)
def _events_loop(self, wait=None):
if wait is None:
wait = self.event_looptime
self.application.processEvents()
time.sleep(wait)
def _on_load_started(self):
self._load_status = None
self._debug(INFO, "Page load started")
def _on_manager_ssl_errors(self, reply, errors):
url = unicode(reply.url().toString())
if self.ignore_ssl_errors:
self._debug(WARNING, "SSL certificate error ignored: %s" % url)
reply.ignoreSslErrors()
else:
self._debug(WARNING, "SSL certificate error: %s" % url)
def _on_authentication_required(self, reply, authenticator):
url = unicode(reply.url().toString())
realm = unicode(authenticator.realm())
self._debug("HTTP auth required: %s (realm: %s)" % (url, realm))
if not self._http_authentication_callback:
self._debug(WARNING, "HTTP auth required, but no callback defined")
return
credentials = self._http_authentication_callback(url, realm)
if credentials:
user, password = credentials
self._debug(
INFO, "callback returned HTTP credentials: %s/%s" %
(user, "*" * len(password)))
authenticator.setUser(user)
authenticator.setPassword(password)
else:
self._debug(WARNING, "HTTP auth callback returned no credentials")
def _manager_create_request(self, operation, request, data):
url = unicode(request.url().toString())
operation_name = self._operation_names[operation].upper()
self._debug(INFO, "Request: %s %s" % (operation_name, url))
for h in request.rawHeaderList():
self._debug(DEBUG, " %s: %s" % (h, request.rawHeader(h)))
if self._url_filter:
if self._url_filter(self._operation_names[operation],
url) is False:
self._debug(INFO, "URL filtered: %s" % url)
request.setUrl(QUrl("about:blank"))
else:
self._debug(DEBUG, "URL not filtered: %s" % url)
reply = QNetworkAccessManager.createRequest(self.manager, operation,
request, data)
return reply
def _on_reply(self, reply):
self._replies += 1
url = unicode(reply.url().toString())
if reply.error():
self._debug(
WARNING, "Reply error: %s - %d (%s)" %
(url, reply.error(), reply.errorString()))
else:
self._debug(INFO, "Reply successful: %s" % url)
for header in reply.rawHeaderList():
self._debug(DEBUG, " %s: %s" % (header, reply.rawHeader(header)))
def _on_unsupported_content(self, reply, outfd=None):
if not reply.error():
self._start_download(reply, outfd)
else:
self._debug(
ERROR,
"Error on unsupported content: %s" % reply.errorString())
def _javascript_alert(self, webframe, message):
self._debug(INFO, "Javascript alert: %s" % message)
if self.webview:
QWebPage.javaScriptAlert(self.webpage, webframe, message)
def _javascript_console_message(self, message, line, sourceid):
if line:
self._debug(
INFO,
"Javascript console (%s:%d): %s" % (sourceid, line, message))
else:
self._debug(INFO, "Javascript console: %s" % message)
def _javascript_confirm(self, webframe, message):
smessage = unicode(message)
url = webframe.url()
self._debug(
INFO,
"Javascript confirm (webframe url = %s): %s" % (url, smessage))
if self._javascript_confirm_callback:
value = self._javascript_confirm_callback(url, smessage)
self._debug(INFO,
"Javascript confirm callback returned %s" % value)
return value
return QWebPage.javaScriptConfirm(self.webpage, webframe, message)
def _javascript_prompt(self, webframe, message, defaultvalue, result):
url = webframe.url()
smessage = unicode(message)
self._debug(
INFO,
"Javascript prompt (webframe url = %s): %s" % (url, smessage))
if self._javascript_prompt_callback:
value = self._javascript_prompt_callback(url, smessage,
defaultvalue)
self._debug(INFO,
"Javascript prompt callback returned: %s" % value)
if value in (False, None):
return False
result.clear()
result.append(value)
return True
return QWebPage.javaScriptPrompt(self.webpage, webframe, message,
defaultvalue, result)
def _on_webview_destroyed(self, window):
self.webview = None
def _on_load_finished(self, successful):
self._load_status = successful
status = {True: "successful", False: "error"}[successful]
self._debug(
INFO, "Page load finished (%d bytes): %s (%s)" %
(len(self.html), self.url, status))
def _get_filepath_for_url(self, url):
urlinfo = urlparse.urlsplit(url)
path = os.path.join(self.download_directory,
urlinfo.netloc + urlinfo.path)
if not os.path.isdir(os.path.dirname(path)):
os.makedirs(os.path.dirname(path))
return path
def _start_download(self, reply, outfd):
def _on_ready_read():
data = reply.readAll()
if not hasattr(reply, "downloaded_nbytes"):
reply.downloaded_nbytes = 0
reply.downloaded_nbytes += len(data)
outfd.write(data)
self._debug(
DEBUG,
"Read from download stream (%d bytes): %s" % (len(data), url))
def _on_network_error():
self.debug(ERROR, "Network error on download: %s" % url)
def _on_finished():
self._debug(INFO, "Download finished: %s" % url)
url = unicode(reply.url().toString())
if outfd is None:
path = self._get_filepath_for_url(url)
outfd = open(path, "wb")
reply.connect(reply, SIGNAL("readyRead()"), _on_ready_read)
reply.connect(reply, SIGNAL("NetworkError()"), _on_network_error)
reply.connect(reply, SIGNAL("finished()"), _on_finished)
self._debug(INFO, "Start download: %s" % url)
def _wait_load(self, timeout=None):
self._events_loop(0.0)
if self._load_status is not None:
load_status = self._load_status
self._load_status = None
return load_status
itime = time.time()
while self._load_status is None:
if timeout and time.time() - itime > timeout:
raise SpynnerTimeout("Timeout reached: %d seconds" % timeout)
self._events_loop()
self._events_loop(0.0)
if self._load_status:
jscode = "var %s = jQuery.noConflict();" % self.jslib
self.runjs(self.javascript + jscode, debug=False)
self.webpage.setViewportSize(
self.webpage.mainFrame().contentsSize())
load_status = self._load_status
self._load_status = None
return load_status
def _debug(self, level, *args):
if level <= self.debug_level:
kwargs = dict(outfd=self.debug_stream)
_debug(*args, **kwargs)
def _user_agent_for_url(self, url):
if self.user_agent:
return self.user_agent
return QWebPage.userAgentForUrl(self.webpage, url)
def _runjs_on_jquery(self, name, code):
code2 = "result = %s; result.length" % code
if self.runjs(code2).toInt() < 1:
raise SpynnerJavascriptError("error on %s: %s" % (name, code))
def _get_html(self):
return unicode(self.webframe.toHtml())
#return str(self.webframe.toHtml().toAscii())
def _get_soup(self):
if not self._html_parser:
raise SpynnerError("Cannot get soup with no HTML parser defined")
return self._html_parser(self.html)
def _get_url(self):
return unicode(self.webframe.url().toString())
# Properties
url = property(_get_url)
"""Current URL."""
html = property(_get_html)
"""Rendered HTML in current page."""
soup = property(_get_soup)
"""HTML soup (see L{set_html_parser})."""
#{ Basic interaction with browser
def load(self, url):
"""Load a web page and return status (a boolean)."""
self.webframe.load(QUrl(url))
return self._wait_load()
def click(self,
selector,
wait_load=False,
wait_requests=None,
timeout=None):
"""
Click any clickable element in page.
@param selector: jQuery selector.
@param wait_load: If True, it will wait until a new page is loaded.
@param timeout: Seconds to wait for the page to load before
raising an exception.
@param wait_requests: How many requests to wait before returning. Useful
for AJAX requests.
By default this method will not wait for a page to load.
If you are clicking a link or submit button, you must call this
method with C{wait_load=True} or, alternatively, call
L{wait_load} afterwards. However, the recommended way it to use
L{click_link}.
When a non-HTML file is clicked this method will download it. The
file is automatically saved keeping the original structure (as
wget --recursive does). For example, a file with URL
I{http://server.org/dir1/dir2/file.ext} will be saved to
L{download_directory}/I{server.org/dir1/dir2/file.ext}.
"""
jscode = "%s('%s').simulate('click')" % (self.jslib, selector)
self._replies = 0
self._runjs_on_jquery("click", jscode)
if wait_requests:
while self._replies < wait_requests:
self._events_loop()
self._events_loop(0.0)
if wait_load:
return self._wait_load(timeout)
def click_link(self, selector, timeout=None):
"""Click a link and wait for the page to load."""
return self.click(selector, wait_load=True, timeout=timeout)
def click_ajax(self, selector, wait_requests=1, timeout=None):
"""Click a AJAX link and wait for the request to finish."""
return self.click(selector,
wait_requests=wait_requests,
timeout=timeout)
def wait_load(self, timeout=None):
"""
Wait until the page is loaded.
@param timeout: Time to wait (seconds) for the page load to complete.
@return: Boolean state
@raise SpynnerTimeout: If timeout is reached.
"""
return self._wait_load(timeout)
def wait(self, waittime):
"""
Wait some time.
@param waittime: Time to wait (seconds).
This is an active wait, the events loop will be run, so it
may be useful to wait for synchronous Javascript events that
change the DOM.
"""
itime = time.time()
while time.time() - itime < waittime:
self._events_loop()
def close(self):
"""Close Browser instance and release resources."""
if self.webview:
self.destroy_webview()
if self.webpage:
del self.webpage
@classmethod
def configure_proxy(cls,
hostname,
port,
user=None,
password=None,
proxy_type=QNetworkProxy.HttpProxy):
"""
Configure network proxy layer.
@param proxy_type: see QNetworkProxy.ProxyType. Default: HttpProxy.
@param hostname: Proxy hostname.
@param port: Proxy port.
@param username: Proxy username (optional).
@param passwrod: Proxy password (optional).
"""
proxy = QNetworkProxy()
proxy.setType(proxy_type)
proxy.setHostName(hostname)
proxy.setPort(port)
if user and password is not None:
proxy.setUser(user)
proxy.setPassword(password)
QNetworkProxy.setApplicationProxy(proxy)
#}
#{ Webview
def create_webview(self, show=False):
"""Create a QWebView object and insert current QWebPage."""
if self.webview:
raise SpynnerError("Cannot create webview (already initialized)")
self.webview = QWebView()
self.webview.setPage(self.webpage)
window = self.webview.window()
window.setAttribute(Qt.WA_DeleteOnClose)
window.connect(window, SIGNAL('destroyed(QObject *)'),
self._on_webview_destroyed)
if show:
self.show()
def destroy_webview(self):
"""Destroy current QWebView."""
if not self.webview:
raise SpynnerError("Cannot destroy webview (not initialized)")
del self.webview
def show(self):
"""Show webview browser."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.webview.show()
def hide(self):
"""Hide webview browser."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.webview.hide()
def browse(self):
"""Let the user browse the current page (infinite loop)."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.show()
while self.webview:
self._events_loop()
#}
#{ Form manipulation
def fill(self, selector, value):
"""Fill an input text with a string value using a jQuery selector."""
escaped_value = value.replace("'", "\\'")
jscode = "%s('%s').val('%s')" % (self.jslib, selector, escaped_value)
self._runjs_on_jquery("fill", jscode)
def check(self, selector):
"""Check an input checkbox using a jQuery selector."""
jscode = "%s('%s').attr('checked', true)" % (self.jslib, selector)
self._runjs_on_jquery("check", jscode)
def uncheck(self, selector):
"""Uncheck input checkbox using a jQuery selector"""
jscode = "%s('%s').attr('checked', false)" % (self.jslib, selector)
self._runjs_on_jquery("uncheck", jscode)
def choose(self, selector):
"""Choose a radio input using a jQuery selector."""
jscode = "%s('%s').simulate('click')" % (self.jslib, selector)
self._runjs_on_jquery("choose", jscode)
def select(self, selector):
"""Choose a option in a select using a jQuery selector."""
jscode = "%s('%s').attr('selected', 'selected')" % (self.jslib,
selector)
self._runjs_on_jquery("select", jscode)
submit = click_link
#}
#{ Javascript
def runjs(self, jscode, debug=True):
"""
Inject Javascript code into the current context of page.
@param jscode: Javascript code to injected.
@param debug: Set to False to disable debug output for this injection.
You can call Jquery even if the original page does not include it
as Spynner injects the library for every loaded page. You must
use C{_jQuery(...)} instead of of C{jQuery} or the common {$(...)}
shortcut.
@note: You can change the _jQuery alias (see L{jslib}).
"""
if debug:
self._debug(DEBUG, "Run Javascript code: %s" % jscode)
r = self.webpage.mainFrame().evaluateJavaScript(jscode)
if not r.isValid():
r = self.webpage.mainFrame().evaluateJavaScript(jscode)
return r
def set_javascript_confirm_callback(self, callback):
"""
Set function callback for Javascript confirm pop-ups.
By default Javascript confirmations are not answered. If the webpage
you are working pops Javascript confirmations, be sure to set a callback
for them.
Calback signature: C{javascript_confirm_callback(url, message)}
- url: Url where the popup was launched.
- param message: String message.
The callback should return a boolean (True meaning 'yes', False meaning 'no')
"""
self._javascript_confirm_callback = callback
def set_javascript_prompt_callback(self, callback):
"""
Set function callback for Javascript prompt.
By default Javascript prompts are not answered. If the webpage
you are working pops Javascript prompts, be sure to set a callback
for them.
Callback signature: C{javascript_prompt_callback(url, message, defaultvalue)}
- url: Url where the popup prompt was launched.
- message: String message.
- defaultvalue: Default value for prompt answer
The callback should return a string with the answer or None to cancel the prompt.
"""
self._javascript_prompt_callback = callback
#}
#{ Cookies
def get_cookies(self):
"""Return string containing the current cookies in Mozilla format."""
return self.cookiesjar.mozillaCookies()
def set_cookies(self, string_cookies):
"""Set cookies from a string with Mozilla-format cookies."""
return self.cookiesjar.setMozillaCookies(string_cookies)
#}
#{ Download files
def download(self, url, outfd=None):
"""
Download a given URL using current cookies.
@param url: URL or path to download
@param outfd: Output file-like stream. If None, return data string.
@return: Bytes downloaded (None if something went wrong)
@note: If url is a path, the current base URL will be pre-appended.
"""
def _on_reply(reply):
url = unicode(reply.url().toString())
self._download_reply_status = not bool(reply.error())
self._download_reply_status = None
if not urlparse.urlsplit(url).scheme:
url = urlparse.urljoin(self.url, url)
request = QNetworkRequest(QUrl(url))
# Create a new manager to process this download
manager = QNetworkAccessManager()
reply = manager.get(request)
if reply.error():
raise SpynnerError("Download error: %s" % reply.errorString())
reply.downloaded_nbytes = 0
manager.setCookieJar(self.manager.cookieJar())
manager.connect(manager, SIGNAL('finished(QNetworkReply *)'),
_on_reply)
outfd_set = bool(outfd)
if not outfd_set:
outfd = StringIO()
self._start_download(reply, outfd)
while self._download_reply_status is None:
self._events_loop()
if outfd_set:
return (reply.downloaded_nbytes if not reply.error() else None)
else:
return outfd.getvalue()
#}
#{ HTML and tag soup parsing
def set_html_parser(self, parser):
"""
Set HTML parser used to generate the HTML L{soup}.
@param parser: Callback called to generate the soup.
When a HTML parser is set for a Browser, the property L{soup} returns
the parsed HTML.
"""
self._html_parser = parser
def html_contains(self, regexp):
"""Return True if current HTML contains a given regular expression."""
return bool(re.search(regexp, self.html))
#}
#{ HTTP Authentication
def set_http_authentication_callback(self, callback):
"""
Set HTTP authentication request callback.
The callback must have this signature:
C{http_authentication_callback(url, realm)}:
- C{url}: URL where the requested was made.
- C{realm}: Realm requiring authentication.
The callback should return a pair of string containing (user, password)
or None if you don't want to answer.
"""
self._http_authentication_callback = callback
#}
#{ Miscellaneous
def snapshot(self, box=None, format=QImage.Format_ARGB32):
"""
Take an image snapshot of the current frame.
@param box: 4-element tuple containing box to capture (x1, y1, x2, y2).
If None, capture the whole page.
@param format: QImage format (see QImage::Format_*).
@return: A QImage image.
Typical usage:
>>> browser.load(url)
>>> browser.snapshot().save("webpage.png")
"""
if box:
x1, y1, x2, y2 = box
w, h = (x2 - x1), (y2 - y1)
image0 = QImage(QSize(x2, y2), format)
painter = QPainter(image0)
self.webpage.mainFrame().render(painter)
painter.end()
image = image0.copy(x1, y1, w, h)
else:
image = QImage(self.webpage.viewportSize(), format)
painter = QPainter(image)
self.webpage.mainFrame().render(painter)
painter.end()
return image
def get_url_from_path(self, path):
"""Return the URL for a given path using the current URL as base."""
return urlparse.urljoin(self.url, path)
def set_url_filter(self, url_filter):
"""
Set function callback to filter URL.
By default all requested elements of a page are loaded. That includes
stylesheets, images and many other elements that you may not need at all.
Use this method to define the callback that will be called every time
a new request is made. The callback must have this signature:
C{my_url_filter(operation, url)}:
- C{operation}: string with HTTP operation: C{get}, C{head},
C{post} or C{put}.
- C{url}: requested item URL.
It should return C{True} (proceed) or C{False} (reject).
"""
self._url_filter = url_filter
class Browser:
"""
Stateful programmatic web browser class based upon QtWebKit.
>>> browser = Browser()
>>> browser.load("http://www.wordreference.com")
>>> browser.runjs("console.log('I can run Javascript!')")
>>> browser.runjs("_jQuery('div').css('border', 'solid red')") # and jQuery!
>>> browser.select("#esen")
>>> browser.fill("input[name=enit]", "hola")
>>> browser.click("input[name=b]", wait_load=True)
>>> print browser.url, len(browser.html)
>>> browser.close()
"""
ignore_ssl_errors = True
"""@ivar: If True, ignore SSL certificate errors."""
user_agent = None
"""@ivar: User agent for requests (see QWebPage::userAgentForUrl for details)"""
jslib = "jq"
"""@ivar: Library name for jQuery library injected by default to pages."""
download_directory = "."
"""@ivar: Directory where downloaded files will be stored."""
debug_stream = sys.stderr
"""@ivar: File-like stream where debug output will be written."""
debug_level = ERROR
"""@ivar: Debug verbose level (L{ERROR}, L{WARNING}, L{INFO} or L{DEBUG})."""
event_looptime = 0.01
"""@ivar: Event loop dispatcher loop delay (seconds)."""
errorCode = None
errorMessage = None
_javascript_files = ["jquery.min.js", "jquery.simulate.js"]
_javascript_directories = [
os.path.join(os.path.dirname(__file__), "../javascript"),
os.path.join(sys.prefix, "share/spynner/javascript"),
]
def __init__(self, qappargs=None, debug_level=None):
"""
Init a Browser instance.
@param qappargs: Arguments for QApplication constructor.
@param debug_level: Debug level logging (L{ERROR} by default)
"""
self.application = QApplication(qappargs or [])
"""PyQt4.QtGui.Qapplication object."""
if debug_level is not None:
self.debug_level = debug_level
self.webpage = QWebPage()
"""PyQt4.QtWebKit.QWebPage object."""
self.webpage.userAgentForUrl = self._user_agent_for_url
self.webframe = self.webpage.mainFrame()
"""PyQt4.QtWebKit.QWebFrame main webframe object."""
self.webview = None
"""PyQt4.QtWebKit.QWebView object."""
self._url_filter = None
self._html_parser = None
# Javascript
directory = _first(self._javascript_directories, os.path.isdir)
if not directory:
raise SpynnerError("Cannot find javascript directory: %s" %
self._javascript_directories)
self.javascript = "".join(open(os.path.join(directory, fn)).read()
for fn in self._javascript_files)
self.webpage.javaScriptAlert = self._javascript_alert
self.webpage.javaScriptConsoleMessage = self._javascript_console_message
self.webpage.javaScriptConfirm = self._javascript_confirm
self.webpage.javaScriptPrompt = self._javascript_prompt
self._javascript_confirm_callback = None
self._javascript_confirm_prompt = None
# Network Access Manager and cookies
self.manager = QNetworkAccessManager()
"""PyQt4.QtNetwork.QTNetworkAccessManager object."""
self.manager.createRequest = self._manager_create_request
self.webpage.setNetworkAccessManager(self.manager)
self.cookiesjar = _ExtendedNetworkCookieJar()
"""PyQt4.QtNetwork.QNetworkCookieJar object."""
self.manager.setCookieJar(self.cookiesjar)
self.manager.connect(self.manager,
SIGNAL("sslErrors(QNetworkReply *, const QList<QSslError> &)"),
self._on_manager_ssl_errors)
self.manager.connect(self.manager,
SIGNAL('finished(QNetworkReply *)'),
self._on_reply)
self.manager.connect(self.manager,
SIGNAL('authenticationRequired(QNetworkReply *, QAuthenticator *)'),
self._on_authentication_required)
self._operation_names = dict(
(getattr(QNetworkAccessManager, s + "Operation"), s.lower())
for s in ("Get", "Head", "Post", "Put"))
# Webpage slots
self._load_status = None
self._replies = 0
self.webpage.setForwardUnsupportedContent(True)
self.webpage.connect(self.webpage,
SIGNAL('unsupportedContent(QNetworkReply *)'),
self._on_unsupported_content)
self.webpage.connect(self.webpage,
SIGNAL('loadFinished(bool)'),
self._on_load_finished)
self.webpage.connect(self.webpage,
SIGNAL("loadStarted()"),
self._on_load_started)
def _events_loop(self, wait=None):
if wait is None:
wait = self.event_looptime
self.application.processEvents()
time.sleep(wait)
def _on_load_started(self):
self._load_status = None
self._debug(INFO, "Page load started")
def _on_manager_ssl_errors(self, reply, errors):
url = unicode(reply.url().toString())
if self.ignore_ssl_errors:
self._debug(WARNING, "SSL certificate error ignored: %s" % url)
reply.ignoreSslErrors()
else:
self._debug(WARNING, "SSL certificate error: %s" % url)
def _on_authentication_required(self, reply, authenticator):
url = unicode(reply.url().toString())
realm = unicode(authenticator.realm())
self._debug("HTTP auth required: %s (realm: %s)" % (url, realm))
if not self._http_authentication_callback:
self._debug(WARNING, "HTTP auth required, but no callback defined")
return
credentials = self._http_authentication_callback(url, realm)
if credentials:
user, password = credentials
self._debug(INFO, "callback returned HTTP credentials: %s/%s" %
(user, "*"*len(password)))
authenticator.setUser(user)
authenticator.setPassword(password)
else:
self._debug(WARNING, "HTTP auth callback returned no credentials")
def _manager_create_request(self, operation, request, data):
url = unicode(request.url().toString())
operation_name = self._operation_names[operation].upper()
self._debug(INFO, "Request: %s %s" % (operation_name, url))
for h in request.rawHeaderList():
self._debug(DEBUG, " %s: %s" % (h, request.rawHeader(h)))
if self._url_filter:
if self._url_filter(self._operation_names[operation], url) is False:
self._debug(INFO, "URL filtered: %s" % url)
request.setUrl(QUrl("about:blank"))
else:
self._debug(DEBUG, "URL not filtered: %s" % url)
reply = QNetworkAccessManager.createRequest(self.manager,
operation, request, data)
return reply
def _on_reply(self, reply):
self._replies += 1
self._reply_url = unicode(reply.url().toString())
self._reply_status = not bool(reply.error())
if reply.error():
self._debug(WARNING, "Reply error: %s - %d (%s)" %
(self._reply_url, reply.error(), reply.errorString()))
self.errorCode = reply.error()
self.errorMessage = reply.errorString()
else:
self._debug(INFO, "Reply successful: %s" % self._reply_url)
for header in reply.rawHeaderList():
self._debug(DEBUG, " %s: %s" % (header, reply.rawHeader(header)))
def _on_unsupported_content(self, reply, outfd=None):
if not reply.error():
self._start_download(reply, outfd)
else:
self._debug(ERROR, "Error on unsupported content: %s" % reply.errorString())
def _javascript_alert(self, webframe, message):
self._debug(INFO, "Javascript alert: %s" % message)
if self.webview:
QWebPage.javaScriptAlert(self.webpage, webframe, message)
def _javascript_console_message(self, message, line, sourceid):
if line:
self._debug(INFO, "Javascript console (%s:%d): %s" %
(sourceid, line, message))
else:
self._debug(INFO, "Javascript console: %s" % message)
def _javascript_confirm(self, webframe, message):
smessage = unicode(message)
url = webframe.url()
self._debug(INFO, "Javascript confirm (webframe url = %s): %s" %
(url, smessage))
if self._javascript_confirm_callback:
value = self._javascript_confirm_callback(url, smessage)
self._debug(INFO, "Javascript confirm callback returned %s" % value)
return value
return QWebPage.javaScriptConfirm(self.webpage, webframe, message)
def _javascript_prompt(self, webframe, message, defaultvalue, result):
url = webframe.url()
smessage = unicode(message)
self._debug(INFO, "Javascript prompt (webframe url = %s): %s" %
(url, smessage))
if self._javascript_prompt_callback:
value = self._javascript_prompt_callback(url, smessage, defaultvalue)
self._debug(INFO, "Javascript prompt callback returned: %s" % value)
if value in (False, None):
return False
result.clear()
result.append(value)
return True
return QWebPage.javaScriptPrompt(self.webpage, webframe, message,
defaultvalue, result)
def _on_webview_destroyed(self, window):
self.webview = None
def _on_load_finished(self, successful):
self._load_status = successful
status = {True: "successful", False: "error"}[successful]
self._debug(INFO, "Page load finished (%d bytes): %s (%s)" %
(len(self.html), self.url, status))
def _get_filepath_for_url(self, url):
urlinfo = urlparse.urlsplit(url)
path = os.path.join(self.download_directory,
urlinfo.netloc + urlinfo.path)
if not os.path.isdir(os.path.dirname(path)):
os.makedirs(os.path.dirname(path))
return path
def _start_download(self, reply, outfd):
def _on_ready_read():
data = reply.readAll()
reply.downloaded_nbytes += len(data)
outfd.write(data)
self._debug(DEBUG, "Read from download stream (%d bytes): %s"
% (len(data), url))
def _on_network_error():
self.debug(ERROR, "Network error on download: %s" % url)
def _on_finished():
self._debug(INFO, "Download finished: %s" % url)
url = unicode(reply.url().toString())
if outfd is None:
path = self._get_filepath_for_url(url)
outfd = open(path, "wb")
reply.connect(reply, SIGNAL("readyRead()"), _on_ready_read)
reply.connect(reply, SIGNAL("NetworkError()"), _on_network_error)
reply.connect(reply, SIGNAL("finished()"), _on_finished)
self._debug(INFO, "Start download: %s" % url)
def _wait_load(self, timeout=None):
self._events_loop(0.0)
if self._load_status is not None:
load_status = self._load_status
self._load_status = None
return load_status
itime = time.time()
while self._load_status is None:
if timeout and time.time() - itime > timeout:
raise SpynnerTimeout("Timeout reached: %d seconds" % timeout)
self._events_loop()
self._events_loop(0.0)
if self._load_status:
jscode = "var %s = jQuery.noConflict();" % self.jslib
self.runjs(self.javascript + jscode, debug=False)
self.webpage.setViewportSize(self.webpage.mainFrame().contentsSize())
load_status = self._load_status
self._load_status = None
return load_status
def _debug(self, level, *args):
if level <= self.debug_level:
kwargs = dict(outfd=self.debug_stream)
_debug(*args, **kwargs)
def _user_agent_for_url(self, url):
if self.user_agent:
return self.user_agent
return QWebPage.userAgentForUrl(self.webpage, url)
def get_js_obj_length(self, res):
if res.type() != res.Map:
return False
resmap = res.toMap()
lenfield = QString(u'length')
if lenfield not in resmap:
return False
return resmap[lenfield].toInt()[0]
def jslen(self, selector):
res = self.runjs("%s('%s')" % (self.jslib, selector))
return self.get_js_obj_length(res)
def _runjs_on_jquery(self, name, code):
res = self.runjs(code)
if self.get_js_obj_length(res) < 1:
raise SpynnerJavascriptError("error on %s: %s" % (name, code))
def _get_html(self):
return unicode(self.webframe.toHtml())
def _get_soup(self):
if not self._html_parser:
raise SpynnerError("Cannot get soup with no HTML parser defined")
return self._html_parser(self.html)
def _get_url(self):
return unicode(self.webframe.url().toString())
# Properties
url = property(_get_url)
"""Current URL."""
html = property(_get_html)
"""Rendered HTML in current page."""
#soup = property(_get_soup)
soup = None #change to none so that changes are retained through mulitple calls
"""HTML soup (see L{set_html_parser})."""
#{ Basic interaction with browser
def load(self, url):
"""Load a web page and return status (a boolean)."""
self.webframe.load(QUrl(url))
return self._wait_load()
def load_request(self, req):
"""Load a network request and return status (a boolean)."""
self.webframe.load(req)
return self._wait_load()
def wait_requests(self, wait_requests = None, url = None, url_regex = None):
if wait_requests:
while self._replies < wait_requests:
self._events_loop()
self._events_loop(0.0)
if url_regex or url:
last_replies = self._replies
while True:
if last_replies != self._replies:
if url_regex:
if re.search(url_regex, self._reply_url):
break
elif url:
if url == self._reply_url:
break
self._events_loop()
self._events_loop(0.0)
def click(self, selector, wait_load=False, wait_requests=None, timeout=None):
"""
Click any clickable element in page.
@param selector: jQuery selector.
@param wait_load: If True, it will wait until a new page is loaded.
@param timeout: Seconds to wait for the page to load before
raising an exception.
@param wait_requests: How many requests to wait before returning. Useful
for AJAX requests.
By default this method will not wait for a page to load.
If you are clicking a link or submit button, you must call this
method with C{wait_load=True} or, alternatively, call
L{wait_load} afterwards. However, the recommended way it to use
L{click_link}.
When a non-HTML file is clicked this method will download it. The
file is automatically saved keeping the original structure (as
wget --recursive does). For example, a file with URL
I{http://server.org/dir1/dir2/file.ext} will be saved to
L{download_directory}/I{server.org/dir1/dir2/file.ext}.
"""
jscode = "%s('%s').simulate('click')" % (self.jslib, selector)
self._replies = 0
self._runjs_on_jquery("click", jscode)
self.wait_requests(wait_requests)
if wait_load:
return self._wait_load(timeout)
def click_link(self, selector, timeout=None):
"""Click a link and wait for the page to load."""
return self.click(selector, wait_load=True, timeout=timeout)
def click_ajax(self, selector, wait_requests=1, timeout=None):
"""Click a AJAX link and wait for the request to finish."""
return self.click(selector, wait_requests=wait_requests, timeout=timeout)
def wait_load(self, timeout=None):
"""
Wait until the page is loaded.
@param timeout: Time to wait (seconds) for the page load to complete.
@return: Boolean state
@raise SpynnerTimeout: If timeout is reached.
"""
return self._wait_load(timeout)
def wait(self, waittime):
"""
Wait some time.
@param waittime: Time to wait (seconds).
This is an active wait, the events loop will be run, so it
may be useful to wait for synchronous Javascript events that
change the DOM.
"""
itime = time.time()
while time.time() - itime < waittime:
self._events_loop()
def close(self):
"""Close Browser instance and release resources."""
if self.webview:
self.destroy_webview()
if self.webpage:
del self.webpage
#}
#{ Webview
def create_webview(self, show=False):
"""Create a QWebView object and insert current QWebPage."""
if self.webview:
raise SpynnerError("Cannot create webview (already initialized)")
self.webview = QWebView()
self.webview.setPage(self.webpage)
window = self.webview.window()
window.setAttribute(Qt.WA_DeleteOnClose)
window.connect(window, SIGNAL('destroyed(QObject *)'),
self._on_webview_destroyed)
if show:
self.show()
def destroy_webview(self):
"""Destroy current QWebView."""
if not self.webview:
raise SpynnerError("Cannot destroy webview (not initialized)")
del self.webview
def show(self):
"""Show webview browser."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.webview.show()
def hide(self):
"""Hide webview browser."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.webview.hide()
def browse(self):
"""Let the user browse the current page (infinite loop)."""
if not self.webview:
raise SpynnerError("Webview is not initialized")
self.show()
while self.webview:
self._events_loop()
#}
#{ Webframe
def set_webframe_to_default(self):
self.webframe = self.webpage.mainFrame()
def set_webframe(self, framenumber):
cf = self.webframe.childFrames()
try:
self.webframe = cf[int(framenumber)]
except:
raise SpynnerError("childframe does not exist")
"""Inject jquery into frame"""
jscode = "var %s = jQuery.noConflict();" % self.jslib
self.runjs(self.javascript + jscode, debug=False)
#}
#{ Form manipulation
def fill(self, selector, value):
"""Fill an input text with a string value using a jQuery selector."""
escaped_value = value.replace("'", "\\'")
jscode = "%s('%s').val('%s')" % (self.jslib, selector, escaped_value)
self._runjs_on_jquery("fill", jscode)
def check(self, selector):
"""Check an input checkbox using a jQuery selector."""
jscode = "%s('%s').attr('checked', true)" % (self.jslib, selector)
self._runjs_on_jquery("check", jscode)
def uncheck(self, selector):
"""Uncheck input checkbox using a jQuery selector"""
jscode = "%s('%s').attr('checked', false)" % (self.jslib, selector)
self._runjs_on_jquery("uncheck", jscode)
def choose(self, selector, value):
"""Choose a radio input using a jQuery selector."""
escaped_value = value.replace("'", "\\'")
jscode = "%s('%s').filter('[value=%s]').simulate('click')" % (self.jslib, selector, escaped_value)
self._runjs_on_jquery("choose", jscode)
def select(self, selector):
"""Choose a option in a select using a jQuery selector."""
jscode = "%s('%s').attr('selected', 'selected')" % (self.jslib, selector)
self._runjs_on_jquery("select", jscode)
submit = click_link
#}
#{ Javascript
def runjs(self, jscode, debug=True):
"""
Inject Javascript code into the current context of page.
@param jscode: Javascript code to injected.
@param debug: Set to False to disable debug output for this injection.
You can call Jquery even if the original page does not include it
as Spynner injects the library for every loaded page. You must
use C{jq(...)} instead of of C{jQuery} or the common {$(...)}
shortcut.
@note: You can change the jq alias (see L{jslib}).
"""
if debug:
self._debug(DEBUG, "Run Javascript code: %s" % jscode)
#XXX evaluating JS twice must be wrong but finding the bug is proving tricky...
#JavaScriptCore/interpreter/Interpreter.cpp and JavaScriptCore/runtime/Completion.cpp
#JavaScriptCore/runtime/Completion.cpp is catching an exception (sometimes) and
#returning "TypeError: Type error" - BUT it looks like the JS does complete after
#the function has already returned
r = self.webframe.evaluateJavaScript(jscode)
if r.isValid() == False:
r = self.webframe.evaluateJavaScript(jscode)
return r
def set_javascript_confirm_callback(self, callback):
"""
Set function callback for Javascript confirm pop-ups.
By default Javascript confirmations are not answered. If the webpage
you are working pops Javascript confirmations, be sure to set a callback
for them.
Calback signature: C{javascript_confirm_callback(url, message)}
- url: Url where the popup was launched.
- param message: String message.
The callback should return a boolean (True meaning 'yes', False meaning 'no')
"""
self._javascript_confirm_callback = callback
def set_javascript_prompt_callback(self, callback):
"""
Set function callback for Javascript prompt.
By default Javascript prompts are not answered. If the webpage
you are working pops Javascript prompts, be sure to set a callback
for them.
Callback signature: C{javascript_prompt_callback(url, message, defaultvalue)}
- url: Url where the popup prompt was launched.
- message: String message.
- defaultvalue: Default value for prompt answer
The callback should return a string with the answer or None to cancel the prompt.
"""
self._javascript_prompt_callback = callback
#}
#{ Cookies
def get_cookies(self):
"""Return string containing the current cookies in Mozilla format."""
return self.cookiesjar.mozillaCookies()
def set_cookies(self, string_cookies):
"""Set cookies from a string with Mozilla-format cookies."""
return self.cookiesjar.setMozillaCookies(string_cookies)
#}
#{ Proxies
def get_proxy(self):
"""Return string containing the current proxy."""
return self.manager.proxy()
def set_proxy(self, string_proxy):
"""Set proxy [http|socks5]://username:password@hostname:port"""
urlinfo = urlparse.urlparse(string_proxy)
proxy = QNetworkProxy()
if urlinfo.scheme == 'socks5' :
proxy.setType(1)
elif urlinfo.scheme == 'http' :
proxy.setType(3)
else :
proxy.setType(2)
self.manager.setProxy(proxy)
return self.manager.proxy()
proxy.setHostName(urlinfo.hostname)
proxy.setPort(urlinfo.port)
if urlinfo.username != None :
proxy.setUser(urlinfo.username)
else :
proxy.setUser('')
if urlinfo.password != None :
proxy.setPassword(urlinfo.password)
else :
proxy.setPassword('')
self.manager.setProxy(proxy)
return self.manager.proxy()
#}
#{ Download files
def download(self, url, outfd=None):
"""
Download a given URL using current cookies.
@param url: URL or path to download
@param outfd: Output file-like stream. If None, return data string.
@return: Bytes downloaded (None if something went wrong)
@note: If url is a path, the current base URL will be pre-appended.
"""
def _on_reply(reply):
url = unicode(reply.url().toString())
self._download_reply_status = not bool(reply.error())
self._download_reply_status = None
if not urlparse.urlsplit(url).scheme:
url = urlparse.urljoin(self.url, url)
request = QNetworkRequest(QUrl(url))
# Create a new manager to process this download
manager = QNetworkAccessManager()
reply = manager.get(request)
if reply.error():
raise SpynnerError("Download error: %s" % reply.errorString())
reply.downloaded_nbytes = 0
manager.setCookieJar(self.manager.cookieJar())
manager.connect(manager, SIGNAL('finished(QNetworkReply *)'), _on_reply)
outfd_set = bool(outfd)
if not outfd_set:
outfd = StringIO()
self._start_download(reply, outfd)
while self._download_reply_status is None:
self._events_loop()
if outfd_set:
return (reply.downloaded_nbytes if not reply.error() else None)
else:
return outfd.getvalue()
#}
#{ HTML and tag soup parsing
def set_html_parser(self, parser):
"""
Set HTML parser used to generate the HTML L{soup}.
@param parser: Callback called to generate the soup.
When a HTML parser is set for a Browser, the property L{soup} returns
the parsed HTML.
"""
self._html_parser = parser
def html_contains(self, regexp):
"""Return True if current HTML contains a given regular expression."""
return bool(re.search(regexp, self.html))
#}
#{ HTTP Authentication
def set_http_authentication_callback(self, callback):
"""
Set HTTP authentication request callback.
The callback must have this signature:
C{http_authentication_callback(url, realm)}:
- C{url}: URL where the requested was made.
- C{realm}: Realm requiring authentication.
The callback should return a pair of string containing (user, password)
or None if you don't want to answer.
"""
self._http_authentication_callback = callback
#}
#{ Miscellaneous
def snapshot(self, box=None, format=QImage.Format_ARGB32):
"""
Take an image snapshot of the current frame.
@param box: 4-element tuple containing box to capture (x1, y1, x2, y2).
If None, capture the whole page.
@param format: QImage format (see QImage::Format_*).
@return: A QImage image.
Typical usage:
>>> browser.load(url)
>>> browser.snapshot().save("webpage.png")
"""
if box:
x1, y1, x2, y2 = box
w, h = (x2 - x1), (y2 - y1)
image0 = QImage(QSize(x2, y2), format)
painter = QPainter(image0)
self.webpage.mainFrame().render(painter)
painter.end()
image = image0.copy(x1, y1, w, h)
else:
image = QImage(self.webpage.viewportSize(), format)
painter = QPainter(image)
self.webpage.mainFrame().render(painter)
painter.end()
return image
def get_url_from_path(self, path):
"""Return the URL for a given path using the current URL as base."""
return urlparse.urljoin(self.url, path)
def set_url_filter(self, url_filter):
"""
Set function callback to filter URL.
By default all requested elements of a page are loaded. That includes
stylesheets, images and many other elements that you may not need at all.
Use this method to define the callback that will be called every time
a new request is made. The callback must have this signature:
C{my_url_filter(operation, url)}:
- C{operation}: string with HTTP operation: C{get}, C{head},
C{post} or C{put}.
- C{url}: requested item URL.
It should return C{True} (proceed) or C{False} (reject).
"""
self._url_filter = url_filter
class WK2pdf(object):
option_defaults = {
"disable_javascript": False,
"dpi": -1,
"colorMode": True,
"resolution": False,
"margin_bottom": '10mm',
"margin_left": '10mm',
"margin_right": '10mm',
"margin_top": '10mm',
"background": True,
"user_style_sheet": '',
"orientation": 'V',
"pageSize": 'A4',
"proxy": False,
"quiet": True,
"jsredirectwait": 200,
"enable_plugins": False,
"use_x11": False
}
page_sizes = {
'A0': QPrinter.A0, 'B1': QPrinter.B1, 'A5': QPrinter.A5, 'B6': QPrinter.B6,
'A1': QPrinter.A1, 'B2': QPrinter.B2, 'A6': QPrinter.A6, 'B7': QPrinter.B7,
'A2': QPrinter.A2, 'B3': QPrinter.B3, 'A7': QPrinter.A7, 'B8': QPrinter.B8,
'A3': QPrinter.A3, 'B4': QPrinter.B4, 'A8': QPrinter.A8, 'B9': QPrinter.B9,
'A4': QPrinter.A4, 'B5': QPrinter.B5, 'A9': QPrinter.A9, 'B0': QPrinter.B0,
'C5E': QPrinter.C5E, 'COMM10E': QPrinter.Comm10E, 'DLE': QPrinter.C5E,
'Executive': QPrinter.Executive, 'Folio': QPrinter.Folio, 'Ledger': QPrinter.Ledger,
'Legal': QPrinter.Legal, 'Letter': QPrinter.Letter, 'Tabloid': QPrinter.Tabloid
}
def __init__(self, *args, **kwargs):
self.app = QApplication(sys.argv)
self.option_defaults.update(kwargs)
for k, v in self.option_defaults.items():
if isinstance(v, basestring):
if k in ['margin_left', 'margin_right', 'margin_top', 'margin_bottom']:
v = self.parseUnitReal(v)
if k == 'pageSize':
v = self.setPageSize(v)
if k == 'orientation':
v = self.setOrientation(v)
if k == 'colorMode':
v = self.setColorMode(v)
setattr(self, k, v)
if len(args) < 2:
self.app.quit(1)
raise Exception(
'You need to specify atleast one input file, and exactly one output file\nUse - for stdin or stdout\n\n')
args = list(args)
self.out = args.pop(-1)
self.in_ = args
self.pageNum = 0
self.pages = []
self.am = QNetworkAccessManager()
self.am.connect(self.am, SIGNAL("sslErrors(QNetworkReply*, const QList<QSslError>&)"), self.sslErrors)
self.loading = 0
self.quiet = True
def guessUrlFromString(self, url):
url = url.strip()
host_re = re.compile(HOST_PORT_RE)
if host_re.match(url):
url = 'http://%s' % url
url_re = re.compile(URL_RE)
if url_re.match(url):
url = QUrl(url, QUrl.TolerantMode)
if url.isValid():
return url
elif os.path.isfile(url):
return QUrl.fromLocalFile(os.path.abspath(url))
else:
splitted_url = url.split('.')
if len(splitted_url) > 1:
prefix = splitted_url[0]
schema = prefix.lower() == 'ftp' and 'ftp' or 'http'
return QUrl('%s://%s' % (schema, url), QUrl.TolerantMode)
def loadDefaults(self):
self.pageSize = QPrinter.A4
self.orientation = QPrinter.Portrait
def setPageSize(self, size):
self.pageSize = self.page_sizes.get(size.upper())
return self.pageSize
def setOrientation(self, orientation):
if orientation.lower() in ['l', 'landscape']:
return QPrinter.Landscape
return QPrinter.Portrait
def setColorMode(self, colorMode):
if not colorMode:
return QPrinter.GrayScale
return QPrinter.Color
def parseUnitReal(self, string):
scale = 1.0
unit = QPrinter.Millimeter
r = re.compile(UNIT_RE)
parsed = r.match(string)
if not parsed:
return
number = float(parsed.group(1))
unit_text = parsed.group(2).lower()
if unit_text in ['', 'mm', 'millimeter']:
scale = 1.0
elif unit_text in ['cm', 'centimeter']:
scale = 10.0
elif unit_text in ['m', 'meter']:
scale = 1000.0
elif unit_text in ['didot']:
unit = QPrinter.Didot
elif unit_text in ['in', 'inch']:
unit = QPrinter.Inch
elif unit_text in ['pc', 'pica']:
unit = QPrinter.Pica
elif unit_text in ['cicero']:
unit = QPrinter.Cicero
elif unit_text in ['px', 'pixel']:
unit = QPrinter.DevicePixel
elif unit_text in ['m', 'meter']:
unit = QPrinter.Point
elif unit in ['point', 'pt']:
unit = QPrinter.Point
return number * scale, unit
def resetPages(self):
self.pages = []
self.pageStart = []
def hfreplace(self, q):
replaces = []
#replaces.append(("[page]",pageNum))
#replaces.append(("[toPage]",str(self.pageStart[-1])))
#replaces.append(("[fromPage]",str(1)))
#replaces.append(("[section]",sec[0]))
#replaces.append(("[subsection]",sec[1]))
#replaces.append(("[subsubsection]",sec[3]))
#replaces.append(("[webpage]",currentPage ==-1 and "Table Of Content" or index[currentPage]))
for tag, replace in replaces:
q = q.replace(tag, replace)
return q
def newPage(self, printer, f, t, p):
self.pageNum += 1
painter = printer.paintEngine().painter()
painter.save();
painter.resetMatrix();
h = printer.pageRect().height();
w = printer.pageRect().width();
# do stuff like header and footer
painter.restore();
def run(self):
for in_url in self.in_:
webpage = QWebPage()
webpage.setNetworkAccessManager(self.am)
webpage.connect(webpage, SIGNAL('loadProgress(int)'), self.loadProgress)
webpage.connect(webpage, SIGNAL('loadFinished(bool)'), self.loadFinished)
webpage.connect(webpage, SIGNAL('loadStarted()'), self.loadStarted)
webpage.settings().setAttribute(QWebSettings.JavaEnabled, self.enable_plugins)
webpage.settings().setAttribute(QWebSettings.JavascriptEnabled, not self.disable_javascript)
webpage.settings().setAttribute(QWebSettings.JavascriptCanOpenWindows, False)
webpage.settings().setAttribute(QWebSettings.JavascriptCanAccessClipboard, False)
webpage.settings().setAttribute(QWebSettings.PrintElementBackgrounds, self.background)
webpage.settings().setAttribute(QWebSettings.PluginsEnabled, self.enable_plugins)
if self.user_style_sheet:
webpage.settings().setUserStyleSheetUrl(self.guessUrlFromString(self.user_style_sheet))
url = in_url
if url == '-':
tmp = tempfile.NamedTemporaryFile(prefix='tmp', suffix='.html')
tmp.write(sys.stdin.read())
url = tmp.name
url = self.guessUrlFromString(url)
webpage.mainFrame().load(url)
self.pages.append(webpage)
def sslErrors(self, reply, error):
reply.ignoreSslErrors()
def printPage(self):
if self.loading:
return
printer = QPrinter()
painter = QPainter()
printer.setOutputFormat(QPrinter.PdfFormat)
printer.setOutputFileName(self.out)
self.pageNum = 0;
self.currentPage = 0;
if not (self.margin_left[1] == self.margin_right[1] == self.margin_top[1] == self.margin_bottom[1]):
raise Exception("Currently all margin units must be the same!")
printer.setPageMargins(self.margin_left[0], self.margin_top[0], self.margin_right[0], self.margin_bottom[0],
self.margin_left[1])
printer.setPageSize(self.pageSize)
printer.setOrientation(self.orientation)
printer.setColorMode(self.colorMode)
if not printer.isValid():
self.app.quit(1)
raise Exception("Unable to write to output file")
while len(self.pages):
page = self.pages.pop(0)
page.mainFrame().print_(printer)
if self.pages:
printer.newPage()
self.app.quit()
def loadFinished(self, ok):
self.loading -= 1
#if not ok:
# raise Exception("Failed loading page")
if not self.quiet:
print "Waiting for redirect"
if self.loading == 0:
QTimer.singleShot(self.jsredirectwait, self.printPage)
def loadStarted(self):
self.loading += 1
def loadProgress(self, progress):
if not self.quiet:
print "Loading page: %02i\r" % progress
def exec_(self):
return self.app.exec_()