So I have the routes defined for my app inside main.py, something like:
app = webapp2.WSGIApplication([
webapp2.Route('/', handler=HomePage, name="home")
])
Inside the cron job I can't seem to access the routes of the app, for example this doesn't work:
self.uri_for('home')
I found somewhere online a snippet that fixes it, but it's ugly to use:
cls.app.router.add(r)
Where r would be an array of routes.
Is there a way to have acces to the app's routes inside an app engine cron job?
Your example is incorrect, it seems to be a cross between simple routes and extended routes.
To be able to use self.uri_for('home') you need to use named routes, i.e. extended routes:
app = webapp2.WSGIApplication([
webapp2.Route(r'/', handler=HomePage, name='home'),
])
With that in place self.uri_for('home') should work, assuming self is a webapp2.RequestHandler instance.
The workaround just looks ugly, but that is pretty much what uri_for does under the hood as well:
def uri_for(self, _name, *args, **kwargs):
"""Returns a URI for a named :class:`Route`.
.. seealso:: :meth:`Router.build`.
"""
return self.app.router.build(self.request, _name, args, kwargs)
Related
I have an AngularJS app being served via Flask. I am using the HTML5 routing mode and thus need to redirect several URLs to the client app. I'm not sure how to do the wildcard matching needing to do this correctly. Currently I just match multiple levels of path like this:
#app.route('/ui/')
def ui():
return app.send_static_file('index.html')
#app.route('/ui/<path>')
def ui(path):
return app.send_static_file('index.html')
#app.route('/ui/<path>/<path2>')
def ui(path,path2):
return app.send_static_file('index.html')
Obviously I don't like this and would like to just have one route (everything starting with ui/).
The path url converter can do this for you:
#app.route('/ui/<path:p>')
def ui(p):
return app.send_static_file('index.html')
I am building a project on GAE, webapp2, jinja2 and I use engineauth for authorization. I need something like Django's context_processor in order to use session, user and some other variables from webapp2.request in templates. Please, help me to solve this problem.
There are many ways to achieve this.
The simplest way probably looks like this:
def extra_context(handler, context=None):
"""
Adds extra context.
"""
context = context or {}
# You can load and run various template processors from settings like Django does.
# I don't do this in my projects because I'm not building yet another framework
# so I like to keep it simple:
return dict({'request': handler.request}, **context)
# --- somewhere in response handler ---
def get(self):
my_context = {}
template = get_template_somehow()
self.response.out.write(template.render(**extra_context(self, my_context))
I like when my variables are in template globals, then I can access them in my template widgets without having to pass around bunch of vars in template. So I am doing it like this:
def get_template_globals(handler):
return {
'request': handler.request,
'settings': <...>
}
class MyHandlerBase(webapp.RequestHandler):
def render(self, context=None):
context = context or {}
globals_ = get_template_globals(self)
template = jinja_env.get_template(template_name, globals=globals_)
self.response.out.write(template.render(**context))
There are other methods in: Context processor using Werkzeug and Jinja2
I am learning the webapp2 framework with its powerful Route mechanism.
My application is supposed to accept URIs like these:
/poll/abc-123
/poll/abc-123/
/poll/abc-123/vote/ # post new vote
/poll/abc-123/vote/456 # view/update a vote
Polls may optionally be organized into categories, so all the above should work also like this:
/mycategory/poll/abc-123
/mycategory/poll/abc-123/
/mycategory/poll/abc-123/vote/
/mycategory/poll/abc-123/vote/456
My incorrect configuration:
app = webapp2.WSGIApplication([
webapp2.Route('/<category>/poll/<poll_id><:/?>', PollHandler),
webapp2.Route('/<category>/poll/<poll_id>/vote/<vote_id>', VoteHandler),
], debug=True)
Question: How could I fix my configuration?
If possible it should be optimized for GAE CPU-time/hosting fee. For example, it may be faster if I add two lines for each entry: one line with category and another one without category...
webapp2 has a mechanism to reuse common prefixes, but in this case they vary, so you can't avoid duplicating those routes, as in:
app = webapp2.WSGIApplication([
webapp2.Route('/poll/<poll_id><:/?>', PollHandler),
webapp2.Route('/poll/<poll_id>/vote/<vote_id>', VoteHandler),
webapp2.Route('/<category>/poll/<poll_id><:/?>', PollHandler),
webapp2.Route('/<category>/poll/<poll_id>/vote/<vote_id>', VoteHandler),
], debug=True)
You should not worry about adding many routes. They are really cheap to build and match. Unless you have tens of thousands, reducing the number of routes won't matter.
A small note: the first route accepts an optional end slash. You could instead use the RedirectRoute to accept only one and redirect if the other is accessed, using the option strict_slash=True. This is not well documented but has been around for a while. See the explanation in the docstring.
I am going to add my solution to this as a complimentary answer on top of #moraes.
So other people having problems like below can get a more complete answer.
Trailing Slash Problem
Optional Parameter Problem
In addition, I figured out how to route both /entity/create and /entity/edit/{id} in one regex.
Below are my routes that supports the following url patterns.
/
/myentities
/myentities/
/myentities/create
/myentities/create/
/myentities/edit/{entity_id}
SITE_URLS = [
webapp2.Route(r'/', handler=HomePageHandler, name='route-home'),
webapp2.Route(r'/myentities/<:(create/?)|edit/><entity_id:(\d*)>',
handler=MyEntityHandler,
name='route-entity-create-or-edit'),
webapp2.SimpleRoute(r'/myentities/?',
handler=MyEntityListHandler,
name='route-entity-list'),
]
app = webapp2.WSGIApplication(SITE_URLS, debug=True)
Below is my BaseHandler that all my handlers inherit from.
class BaseHandler(webapp2.RequestHandler):
#webapp2.cached_property
def jinja2(self):
# Sets the defaulte templates folder to the './app/templates' instead of 'templates'
jinja2.default_config['template_path'] = s.path.join(
os.path.dirname(__file__),
'app',
'templates'
)
# Returns a Jinja2 renderer cached in the app registry.
return jinja2.get_jinja2(app=self.app)
def render_response(self, _template, **context):
# Renders a template and writes the result to the response.
rv = self.jinja2.render_template(_template, **context)
self.response.write(rv)
Below is my MyEntityHandler python class with the get() method signature for the Google App Engine Datastore API.
class MyEntityHandler(BaseHandler):
def get(self, entity_id, **kwargs):
if entity_id:
entity = MyEntity.get_by_id(int(entity_id))
template_values = {
'field1': entity.field1,
'field2': entity.field2
}
else:
template_values = {
'field1': '',
'field2': ''
}
self.render_response('my_entity_create_edit_view_.html', **template_values)
def post(self, entity_id, **kwargs):
# Code to save to datastore. I am lazy to write this code.
self.redirect('/myentities')
I am having problems getting the logout link work in GAE (Python).
This is the page I am looking at.
In my template, I create a link
<p>Logout</p>
But when I click on it I get "broken link" message from Chrome. The url for the link looks like this:
http://localhost:8085/users.create_logout_url(
My questions:
Can anybody explain how this works in general?
What is the correct url for the dev server?
What is the correct url for the app server?
What is the ("/") in the logout url?
Thanks.
EDIT
This link works; but I don't know why:
<p>Logout</p>
What sort of templates are you using? It's clear from the output that you're not escaping your code correctly.
Seems to me that you want to do this instead:
self.response.out.write("This is the url: %s", users.create_logout_url("/"))
You could also pass it to your template, using GAEs implemented django templates.
from google.appengine.ext.webapp import template
...
...
(inside your request handler)
class Empty: pass
data = Empty()
data.logout = users.create_logout_url("/")
self.response.out.write(template.render(my_tmpl, {'data': data})
A useful approach is to add all sorts of info to a BaseRequestHandler and then use this as base class for all of your other request handler classes.
from google.appengine.ext import webapp
...
class BaseRequestHandler(webapp.RequestHandler):
def __init__(self):
webapp.RequestHandler.__init__(self) # extend the base class
class Empty: pass
data = Empty()
data.foo = "bar"
Then your new classes will have access to all the data you provided in the base class.
class OtherHandler(BaseRequestHandler):
def get(self):
self.response.out.write("This is foo: %s" % self.data.foo) # passes str "bar"
Hope it helps.
A.
Hi following more or less what this article is showing for the user account stuff. In gwt I store server side the logout/login url and I pass them to the client
http://www.dev-articles.com/article/App-Engine-User-Services-in-JSP-3002
Using tipfy, how does one express a catch-all route in urls.py if more specific routes do not match?
Tipfy uses Werkzeug-like routing, so there's this (in urls.py):
def get_rules(app):
rules = [
Rule('/<any>', endpoint='any', handler='apps.main.handlers.MainHandler'),
Rule('/', endpoint='main', handler='apps.main.handlers.MainHandler'),
]
This will match most random entry points into the application (app.example.com/foo, app.example.com/%20 etc) but does not cover the app.example.com/foo/bar case which results in a 404.
Alternatively, is there a graceful way to handle 404 in Tipfy that I'm missing?
I think you want:
Rule('/<path:any>', endpoint='any', handler='apps.main.handlers.MainHandler')
The path matcher also matches slashes.
Maybe you could write custom middle ware:
class CustomErrorPageMiddleware(object):
def handle_exception(self, e):
return Response("custom error page")
To enable it add somewhere to tipfy config:
config['tipfy'] = {
'middleware': [
'apps.utils.CustomErrorPageMiddleware',
]
}
It gives you quite a flexibility - you could for example send mail somewhere to inform that there was a problem. This will intercept all exceptions in your application