Mailman 3 June 2005 - Twisted-web

issue with session and LivePage
by Carl Shimer 12 Jun '05

12 Jun '05

Hi, I have a guarded site using livepage. If the user is currently viewing a livepage site and lets the session expire (i.e they do nothing), livepage will generate a client side alert that is basically the contents of the stylesheet in failure.py. Not such a good user experience :) After some digging I noticed that Livepage doesn't have a notifyOnExpire handler (liveevil does however). The solution is to add a handler in the goingLive implementation: def goingLive(self,ctx,handle): self.handle = handle inevow.ISession(ctx).notifyOnExpire(self.OnExpireCallBack) The problem is in the implementation of OnExpireCallback. The current code: def OnExpireCallBack(self): # send the user back to the main page self.handle.sendScript('window.location.href = "/"') self.handle._closeComplete() A couple of issues here: 1) it doesn't seem right to call _closeComplete() but it seems to work. Is there a better solution? 2) I would like to do something a bit fancier here that involves displaying a DIV for 1 sec and then redirecting to the main page. Sample client side JS (you could do the same thing by using the client handle). function handleLogout(strval) { newval = "<div id='logoutpopup'>" + strval + "</div>"; nevow_appendNode('maincontent',newval) self.setTimeout("window.location.href = \"/\"",500) } However, calling handleLogout before calling _closeComplete() will result in the same error mentioned at the top. Perhaps a timing issue? Perhaps some of these issues will be resolved with donovan's new livepage, although I noticed on his branch that there still isn't a notifyOnExpire handler. Regards, Carl

1 0

Using lineReceived with HTTP requests
by Andrea Soong 11 Jun '05

11 Jun '05

Hi all, First off, I'd like to say thanks to Ken Whitesell for giving me some much needed help. Here's my problem: I am using an HTTPClient instance to connect to a webpage, and am trying to retrieve its contents and display to the screen line by line, using the lineReceived method. This is all being done in an .rpy file. For some reason, I am only receiving the header information: --------------------------------------------- [start line received] HTTP/1.0 200 OK [end line received] [start line received] Cache-Control: private [end line received] [start line received] Content-Type: text/html [end line received] [start line received] Set-Cookie: PREF=ID=aacd1eed2e435ca5:TM=1118474192:LM=1118474192:S=QyMnwhj46cXbhy_1; expires=Sun, 17-Jan-2038 19:14:07 GMT; path=/; domain=.google.com [end line received] [start line received] Server: GWS/2.1 [end line received] [start line received] Date: Sat, 11 Jun 2005 07:16:32 GMT [end line received] [start line received] Connection: Close [end line received] [start line received] [end line received] --------------------------------------------- Does anyone know if there is some incompatibility in using lineReceived with the format of a webpage (i.e. maybe they are missing the standard EOL delimiters)? dataReceived works fine and outputs the whole webpage at once. Also, I am using handleRequestEnd to detect the end of the transmission and call request.finish(), but it does not seem to be getting called at all - the browser just keeps waiting for more data. I've put a simple request.write("handleRequestEnd called") call on the first line of the method and it does not get written to the screen. I can't find any explanation for how to use most of the methods in the HTTPClient class in the API - is there a better way to do this (i.e. one that works!) ? Thanks, Andrea --------------------------------- Discover Yahoo! Have fun online with music videos, cool games, IM & more. Check it out!

1 0

Re: [Twisted-Python] Difficulty with twisted.web.client.getPage function
by Dave Gray 10 Jun '05

10 Jun '05

Alan Jern wrote: > Hey, I'm new to Twisted and I've run into a problem that I can't solve. > > I'm trying to write a simple Resource class that fetches a page from a > specified URL and returns it to the browser. I created a class called > MyResource in a .rpy file with the following two member functions: > > def print_page (self, html): > self.rq.write(html) > self.rq.finish() > > def render_GET (self, request): > self.rq = request > target = escape(request.args['target'][0]) > getPage(target).addCallback(self.print_page) > return server.NOT_DONE_YET > > When I run this script and send a request, the print_page function is > invoked but the variable 'html' has nothing in it. According to the > API and the examples I saw on the HOWTO section, this looks like it > ought to work. Clearly I'm missing something. You will probably get more enthusiastic responses to questions like this on twisted-web (cc'd), FYI That said, this works for me: ---fetchpage.tac--- # run me with twistd -noy fetchpage.tac from twisted.web import server from twisted.web.resource import Resource from twisted.web.client import getPage from cgi import escape class Foo(Resource): isLeaf=True def print_page (self, html): self.rq.write(html) self.rq.finish() def render_GET (self, request): self.rq = request target = escape(request.args['target'][0]) getPage(target).addCallback(self.print_page) return server.NOT_DONE_YET resource = Foo() site = server.Site(resource) from twisted.application import service, internet application = service.Application("PageFetcher") internet.TCPServer(9000, site).setServiceParent(application) # and then from your command prompt: > telnet localhost 9000 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. GET /?target=http://google.com/ HTTP/1.0 Host: localhost HTTP/1.0 200 OK Date: Fri, 10 Jun 2005 21:20:49 GMT Content-type: text/html Server: TwistedWeb/2.0.0 <html><head><meta http-equiv="content-type" content="text/html; charset=ISO-8859-1"><title>Google</title><style><!-- etc etc

1 0

Where is freeform, renderer, and formless
by Alberto Trujillo 09 Jun '05

09 Jun '05

Hello: I have problems with all the examples that I have found about Webform, even with the example that is in the tutorial. This is the most complete example that I've found: -------------------------------------------------------------------------------------------- from twisted.application import service from twisted.application import internet from nevow import appserver from nevow import renderer from nevow import tags from nevow import formless from nevow import freeform class IMyForm(formless.TypedInterface): foo = formless.Integer() class FormPage(renderer.Renderer): __implements__ = IMyForm, renderer.Renderer.__implements__ foo = 5 document = tags.html[ tags.body[ "Hello! Here is a form:", freeform.configure ] ] application = service.Application("formpost") internet.TCPServer( 8080, appserver.NevowSite( FormPage() ) ).setServiceParent(application) -------------------------------------------------------------------------------------------------------------- But still I see the same problems each time that I try to execute the program: - cannot import name renderer (why he is not using rend?) - cannot import name freeform - cannot import name formless I running Nevow 0.3.0, because when I try to install Nevow 0.4.1 I received the next message: - error: invalid Python installation: unable to open /usr/lib/python2.3/config/Makefile (No such file or directory) Could anybody give me the solutions to these problems. Thank you very much

3 2

Need help getting started...
by Andrea Soong 08 Jun '05

08 Jun '05

Hello, I'm taking a class in which we're supposed to be implementing a pseudo-proxy server that does some text parsing on web pages before passing them back to the client. The proxy is supposed to be accessed via the URL and port that the server is running on, using an .rpy file. I've looked through all of the available Twisted documentation, but just can't seem to get a handle on how to get started. It's mostly setup questions - what code should go in which files? How do I run it? How do I test it? There are a lot of code examples available, but very little information on how they interact and call each other. It doesn't help that I am not too familiar with the OO-style API documentation. I know that I need to be creating a subclass of HTTPClient, and creating a resource in the .rpy file. I'm guessing that the .rpy file should create an instance of the HTTPClient subclass and use it to retrieve the requested URL (we have to get it line by line), but I don't understand how to create the instance. A single good example that creates a web proxy server which simply retrieves the data line by line and displays it to the screen, plus the shell commands needed to run/test would be extremely helpful. I am completely confused as to how the web server and .rpy file would interact. I hope someone has the time and know-how to help me out. THANKS --------------------------------- Discover Yahoo! Stay in touch with email, IM, photo sharing & more. Check it out!

2 1

problem with livepage-completion-notification-3 branche
by Siegmund Fuehringer 07 Jun '05

07 Jun '05

hi! livepage doesn't work, when i return a LivePage instance from a childFactory. eg: class Test( livepage.LivePage ): ... class Root( rend.Page ): def childFactory( self, ctx, name ): return Test( ) <---- doesn't work child_muh = Test( ) <---- works it works in nevow trunk though. bye - sifu

2 1

Livepage, ClientHandle and context
by Thomas HERVE 06 Jun '05

06 Jun '05

I've just started to use nevow i18n and I'm pretty happy with it, it works very well and the automatic detection of browser language is invisible. But my app uses livepage a lot, and I began to face problems when trying to translate strings in "dynamic renderers". After some investigations, I found my problem : i18n uses the i18nconfig saved in context to translate strings, but livepage.ClientHandle makes a new "fake" context for each calls via the livepage.flt function. I resolved my problem by creating my ClientHandle and ClientHandleFactory implementations, the ClientHandle saving the context at its creation (pretty easy, but I would have found it cleaner if clientHandleFactory was a class variable of LivePage, Input and Output, and not a module global). Thus, the i18nconfig remains in context. My question is : is this choice good, or is there a good reason for ClientHandle not to keep track of the context ? If not, wouldn't it be better to save context in ClientHandle (it would useful for others things, getting session for example) ? -- Thomas

6 8

ReverseProxy subpath
by Siegmund Fuehringer 01 Jun '05

01 Jun '05

hi! i would like to setup a reverse proxy, which serves the root of an internal server as '/subpath' of the external server: http://external.org/subpath -> http://internal.org/ but i can't get it to work. my current setup requires that the internal server also serves the content as '/subpath': root.putChild( 'subpath', ReverseProxyResource( '127.0.0.1', 6080, '/vhost/http/external.org:80/subpath' ) ) is there a way to achieve, what i would like to have? thank you, sifu

2 1