FWIW, I've recreated the virtualenv a couple times, before pinning all the
versions, and after pinning with pyramid-1.7. Same behavior. :-(
On Thursday, June 23, 2016 at 8:29:46 PM UTC-4, Bert JW Regeer wrote:
>
> No. He is calling pserve from his bin directory directly, that’s perfectly
> accep
I've been using "pserve --reload myenv.ini" forever and recently the flags
stopped working. Perhaps it's me, or perhaps I didn't pin versions in my
setup.py file, but the "--reload" and other flags which used to work with
pserve are no longer recognized.
My .ini has:
[server:main]
use = egg:w
I know I'm late to the party, but found this post while trying to achieve
the same thing. We've found a simpler way, taking advantage of the fact
that Response subclasses WebOb which (since 2012) has a 'json' kwarg
(aliased to 'json_body'). So we can do this, short-n-sweet:
@view_config(route_n
"Mike Orr" <[EMAIL PROTECTED]> writes:
> We need somebody who has used AuthKit to write the simple HOWTOs that
> people are asking for.
I did and did, but it was a while back and I suspect AuthKit's changed
since I wrote it:
http://pylonshq.com/project/pylonshq/wiki/PylonsWithAuthKitForward
I'
I've got an app that needs to read and parse a data file (a RADIUS
dictionary, not a Python "dictionary" but a flat file). I want to
include this file in my egg so I don't have to depend on the OS having
RADIUS installed.
In the application's top-level setup.py, I add my data/ subdir to the
pack
Graham Dumpleton <[EMAIL PROTECTED]> writes:
>
> PythonInterpreter er
>...
>
>
>
> PythonInterpreter zwerfull
>...
>
Ah, didn't know about these directives. Will read up on them.
> If you need true process separation between the applications then you
> cant use
Can someone explain how Pylons uses SQLAlchemy's connections and pools?
It seems that each of my instances -- with different dburi (same mysql,
different db name) -- are sharing the same connection pool. Sometimes
my new instance gets the old data, sometimes the old instance gets the
new data.
I
I'm running an app on Pylons-0.9.4.1 with SQLAlchemy-0.3.4, part webapp,
part cron-driven python which manipulates the MySQL DB through the same
SQLAlchemy models. I now want to essentially clone the instance and run
another with a different configuration and different database within the
same My
Been away for a few weeks but this appears the root of Owen Davis'
"Session Issue" reported June 20:
http://www.mail-archive.com/pylons-discuss@googlegroups.com/msg04306.html
If my pylons .ini file doesn't specify a setting for 'data_dir' then
beaker gets passed a data_dir variable whose value
I had issues in the past with (for example) umlauted-chars being
entered into my data that then couldn't be rendere by Myghty with the
default ascii codec; changing python's sitecustomize.py to use:
sys.setdefaultencoding("utf-8")
has fixed this -- until now.
My user uploaded a CSV file us
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> The current favorite way to run Pylons apps behind Apache is to proxy
> them and use Paster. It's exactly the sort of problems that you ran
> into that make some of us wary of mod_python.
I did get it working (expat problem, reported here with
alecf <[EMAIL PROTECTED]> writes:
>
> SetHandler python-program
> PythonHandler paste.modpython
> PythonOption wsgi.application client::client_app
>
>
> And this works great.
Yeah, I just found my problem: an incompatibility between the expat
built into Python-2.4.4 and the one I had installed
I used mod_python about 6 months back to get Apache to proxy HTTPS to
my Pylons app; it wasn't easy.
Now I'm trying again with current Pylons, paste and friends. But I'm
again having troubles getting it to work.
The simplest apache config I've seen is this from
http://pythonpaste.org/module-pa
My app shows an overview of items, and lets me drill down into the
details for a single item; pretty common. These item instances get
their data from a (RDF) database and renders with Myghty. Fine so far.
But if I delete one of these individual items, then return to the
overview or specific det
Evert Rol <[EMAIL PROTECTED]> writes:
> I was afraid so. I guess somewhere along the line, things have to be
> decoded anyway.
> I have little say over the ID, so I cannot really prevent a slash or %
> 2f occurring (or simply tell how many there'll be).
> For the moment, I took Ron's idea and
Ian Bicking <[EMAIL PROTECTED]> writes:
> what I'd really like to do instead of the setup-app stuff is have
> paster just do an internal web request, and you'd write the setup as
> a controller in your application (probably a controller class
> written for this particular use case).
This sounds
Ben Bangert <[EMAIL PROTECTED]> writes:
> Or, the other way I've seen it done, is that you click some names in
> a list, hit a "Add" button in the middle, and they pop over to the
> other list. So you have a list of remaining items, and a list of
> selected items. This paradigm also is pret
I've got a list of about 100 servers that I want folks to be able to
select one or more of. A typical list with multiple=True
works but its a bit clunky -- you can't see very well whether you've
selected the right ones unless you display the entire list on the
page.
I don't see anything like t
Using Pylons-0l9.4.1 and Myghty-1.1. I can't seem to set encoding to
utf-8 for Myghty or mod_python so that Pylons gets it.
My app was working well but just started bombing, likely because one
of my techs entered a name into a form which had non-ascii chars, and
then the rendering in the display
Rick Flosi <[EMAIL PROTECTED]> writes:
> You could alternatively have 2 submit buttons with the same name and
> different values and check the values no?
I initially did this but found that the formencode validator got
confused and changed my button names on an non-validating form. So I
gave e
"ben adam" <[EMAIL PROTECTED]> writes:
> This is what i have in my template:
>
>
>
I do the same thing and specifically check for a value indicating the
Submit was clicked, like:
if self.form_result.get('get_merges'):
...
if self.form_result.get('plot_merges'):
...
If a submi
I'm trying to use formencode.validators Set() to check a multi-select
form input. I cannot get it to allow me to select nothing, to return
the empty list if nothing is selected: if I don't select something
from the multi-select it complains "Missing value".
The docs say:
C Set(...) ...
I'm using AuthKit with an SQLite DB and wrote a couple authorization
checking classes like:
class RoleIn(Permission):
"""Does the user have a role in the supplied list of roles (logical OR).
A user can have more than one role.
"""
def __init__(self, roles):
sel
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> This multiple variable values vs. single has been a bugger for Python
> developers since the early days of the CGI module. I think the
> thinking is that by saying self.form_result['fieldname'], you're
> implicitly saying you're expecting one t
> On Mar 28, 2007, at 8:39 AM, laurent wrote:
>>
>> 1
>> 2
>> 3
>> 4
>>
>>
>> In my controller, h.log(request.params) outputs:
>> => MultiDict([('bob', u'4')])
>>
>> Shouldn't I receive (1,2,3,4) instead ?
Philip Jenvey <[EMAIL PROTECTED]> writes:
> You should have entries for all four values
Sorry to self.followup but I've got AuthKit working with
authentication checks and two new Permission checks, one for Roles and
one for Groups.
My implementation may not be the cleanest and I'm still uncertain
about doing auth in my account.py controller versus doing something
with valid() in a
James Gardner <[EMAIL PROTECTED]> writes:
> I extracted all the lifeproject code into the AuthDemo code here.
> http://authkit.org/svn/AuthKit/trunk/examples/pylons/AuthDemo/
>
> Well, the AuthDemo code is a full example of how to store user data in
> an SQLAlchemy database but other than that I
Using AuthKit-0.3.0pre5.
The code gets the file pointer but then reads from the filename string:
def __init__(self, filename):
string = None
try:
fp = open(filename, 'r')
string = filename.read()
finally:
fp.close()
should be:
I need to be able to create/modify/delete my AuthKit users through my
webapp. The documented built-in users can come from the .ini file
config vars or their own file. While I could conceivably change user
info and write it back out to a stand-alone file, I'm a bit concerned
about race conditions
I'm just starting to play with OpenID and tried to follow the example
in the AuthKit 0.3.0pre5 docs, section 4.8:
http://authkit.org/docs/manual.html
from paste.httpserver import serve
from authkit.authenticate import middleware, test_app
app = middleware(
test_app,
Ben Bangert <[EMAIL PROTECTED]> writes:
> You likely have stale objects in the session, are you making sure to
> clear() it or create a new one every request?
No, didn't realize I needed to. I've just done this and it seems to
have fixed the spontaneously changing data I was seeing. Many tha
I've got pylons app that uses SQLAlchemy with a bunch of tables. I've
got one relations table between two others: system, contact, and
system_contact.
Problem is: when I run the controller method over and over (by simple
page reload) I get *changing* data some times. The controller data is
out
Ian Bicking <[EMAIL PROTECTED]> writes:
> FWIW, there will be no way for Routes to distinguish an incoming request
> for FastEnet-10/100 from FastEnet-10%2f100 -- WSGI/CGI dictates that the
> path is decoded before it is sent to Routes, so the two end up the same.
I didn't realize that. So ev
I've got a route defined in routing.py like:
m.connect('kb/:fac/:sev/:expires/:checksum/*text', controller='kb',
action='search')
In a command line tool, I import routes and routing to get the map
then use generate to construct a URL:
kb_url = conf.URL_BASE + m.generate('kb','search',
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> I think you're on the right path. It sounds like you're having a hard
> time figuring out what you should do than figuring out how to do it.
> If I had to implement different access controls based on different
> URLs, I'd probably just do it in
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> On 2/5/07, Philip Jenvey <[EMAIL PROTECTED]> wrote:
>> Your project's setup.py can call the setup function with a 'scripts'
>> argument. These scripts will be installed the the PREFIX/bin/
>> directory. scripts is a standard distutils argument.
I'm trying to debug a problem I'm having with routing.py. Most of
mine work but this one matches 5 dynamic parts and one tends to have
long strings, error messages typically.
m.connect('kb_search', 'kb/:fac/:sev/:expires/:checksum/*text',
controller='kb', action='search')
I use urllib.quote
My application has two parts, a webapp, and a bunch of commandline
programs that do a bunch of analysis; they both talk to the same
SQLAlchemy-fronted DB, and both reference files from each other.
The commandline tools all live under .../lib/* so they could be made
part of the egg.
I've built and
I've got a pylons app that started out mostly as an internal
customer/sytem/contact management system. Until now it's been
internal only so we've had no auth whatsoever.
I've been adding a way to send email to customers with status and
give them a URL back to the webapp where they can search a
I've got a bunch of tables in a MySQL DB and accessing them with
SQLAlchemy. Now I want to add a FULLTEXT index and use its MATCH()
queries. Where should I be putting the more raw, low-level SQL
statements to create the FULLTEXT index, then how should I be
accessing it to do the non-SQLAlchemy MA
I want to use MySQL's FULLTEXT search so I need to specify the right
index for a table I've defined to SQLAlchemy. Then I'll want to do a
"SELECT ... MATCH()" query on it.
Where is the best place to specify this index that fits with the
Pylons and SQLAlchemy way of doing things. I presume I'll
Nice overview of the various pieces and a good comparison with other
approaches and frameworks. I wish I had your document when I started
working with Pylons to get the big picture.
Thanks!
--~--~-~--~~~---~--~~
You received this message because you are subscrib
Ian Bicking <[EMAIL PROTECTED]> writes:
Chris> Is there a good way to get access to the development.ini
Chris> configuration variables from those files?
Ian> Try using paste.deploy.appconfig('/full/path/to/development.ini')
No joy using PasteDeploy-1.1-py24.egg, Paste-1.1.1-py2.4.egg. Tried
i
I'm working on a pylons app that has a bunch of non-webapp code run
from cron. Is there a good way to get access to the development.ini
configuration variables from those files? I want to make sure they
cron-driven scripts use all the same DB connection parameters and
such.
Thanks.
--~--~--
I've got an app which is using Pylons for the web portion, but have a
bunch of cron jobs that do data gathering and reporting by email. I
put this code in the pylons projectname/lib/ subdir so I can get at
the app's database code and such.
I didn't find any simple doc on myghty.org about crea
I render a form in Myghty with a web helper populate from
SQLAlchemy data like:
<% h.select('systems',
h.options_for_select_from_objects(c.systems,
selected=[s.system_id for s in c.contact.systems],
name_attr='name', value_attr='system_id'), multiple=True) %>
It
I have a bunch of tables I'm storing in MySQL with SQLAlchemy; using
automatically set timestamps with func.now() works but
datetime.datetime.now() appears to corrupt other entries.
My System table looks like the following; constructor and creation
code are at the end to avoid clutter:
system_
I have controllers for objects Client and System. In the Client 'show'
action I list the systems the client has and have a form inline at the
bottom to add a new system. It invokes an 'add' action for the
*System* object, in the system controller. If successful, it
re-invokes the Client index ac
I do a bunch of U.S. government work and they require all web apps to
be Section 508 compliant for accessibility and I want to make sure my
pylons apps comply.
Web form labels and input boxes are one area that's problemmatic for
screen readers. The 508 recommendation is to use tags pointing
at
[oops, meant to send this to the list rather than James Gardner, sorry]
I've been reading and rereading the docs and AuthKit code and can't
figure out how to get AuthKit to work with 'forward' authentication.
(I got the 'form' style working fine but want the form to be wrapped
in my Myghty autoha
I've got an app I'm writing which will need to present a different
look-n-feel for different users. The site admins might see one look,
but each of their customers might see a differently "branded" skin.
The code (model, controller) will be the same and I'd like to reuse as
much of the templates
I started playing with AuthKit and the form-based authentication looks
exactly like what I need. The example code worked fine and generated
a bare-bones username/password login page. The doc mentions you can
specify your own templates to use in python, but I'm don't know how to
specify these.
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> This question sounds familiar, so I wonder if I've already talked to
> you.
Yeah, you pointed out that HTTP Auth won't do what I need -- hence my
interest in choosing between Barrel and AuthKit. (or perhaps something
else).
> * As for a st
We're currently using Apache's HTTP auth but need to provide a Logout
button and have an idle timeout.We'll eventually want to have
groups/roles so I'm looking for modules to do this, using form based
authentication.
AuthKit looks promising but mentions using version 0.3 for
pylons-0.9.3, but
I forgot to mention that the import does work from a python shell:
Python 2.4.3 (#1, Sep 1 2006, 11:54:34)
>>> from paste.deploy import loadapp
>>>
So I didn't understand why mod_python/wsgi.py couldn't do the same import.
Finally I looked at my .../site-packages
We just upgraded 0.8.2 to 0.9.3 and the app is now working in
development, so I made an egg to run under apache mod_python. I'm
using a modified mod_python/wsgi.py that Ben provided me a while
back. When I try to connect I get this:
Mod_python error: "PythonHandler mod_python.wsgi"
Traceba
We're finally migrating our customer's app from Pylons 0.8.2 to 0.9.2.
We followed Ben's email instructions from June 22:
2) Update your project templates to 0.9, and immediately remove all
references to the 'm' object.[...]
paster create --template=pylons yourproj
It will pr
"Shannon -jj Behrens" <[EMAIL PROTECTED]> writes:
> Either you're confused, or I'm confused.
I admit it, I am.
> If the browser is presenting a login popup, you're using HTTP-based
> authentication. Such a system does not "time out". The browser
> actually sends the credentials with every re
We're running Pylons (0.8.2, upgrading this week) behind Apache2 with
mod_python. We use Apache for HTTPS and Digest authentication, works
well.
Except our customer says we need two features to satisfy our Security
group?
* A "logout" button
Is there some Apache session variable accessible
59 matches
Mail list logo
