From andrewm at object-craft.com.au  Wed Mar 14 11:35:51 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 14 Mar 2007 11:35:51 +1100
Subject: [albatross-users] Auto list bug/issue... 
In-Reply-To: <C0FC22C08B82074A88B500617641577787A84D@johnson.mediapulse.net> 
References: <C0FC22C08B82074A88B500617641577787A84D@johnson.mediapulse.net>
Message-ID: <20070314003551.32E035CCB04@longblack.object-craft.com.au>

>I've come across a strang which is in 1.11+, maybe prior versions as
>well but I didin't check them.
>
>Normally, if you try to have several form fields defined with the same
>name, you either need to use the "list" argument on the tags, or make
>the list in the page object/module.  This is very nice, because
>albatross doesn't pull an perl like bs of thinking for you, and letting
>some bug creep into production because it never caused a problem when
>testing.
>
>However, if you name several fields the same name, but they are
>different types, you can get a field that will sometimes be a list,
>sometime a string.  To show this, here is a simple template that will
>show the problem:

Thanks for the report, Michael. I know it's been outstanding for a while!

The problem is due to inputs that can appear multiple times, but return
a single value (such as radio buttons and submit buttons). This simple
example exhibits the problem also (it should raise an exception on load,
but doesn't):

    <al-form>
     <al-input type="text" name="name">
     <al-input type="radio" name="name">
    </al-form>

I've checked a change in (r8871) that addresses this issue, and I hope
to be able to put out a new release later in the week.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/


From andrewm at object-craft.com.au  Wed Mar 14 11:54:46 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 14 Mar 2007 11:54:46 +1100
Subject: [albatross-users] validate_request in Albatross docstring 
In-Reply-To: <45671B9B.4040002@sauvages.be> 
References: <45671B9B.4040002@sauvages.be>
Message-ID: <20070314005446.E270B5CCB04@longblack.object-craft.com.au>

>However, I just lost some precious minutes because of the wrong
>docstring for method
>
>validate_request in Albatross latest online doc
>(http://object-craft.com.au/projects/albatross/albatross/pack-app.html#l2h-272)
>
>Could it be updated, so that it explains what it means to return 0,
>and what it means to return 1?

I've changed this to read:

    \begin{methoddesc}[Application]{validate_request}{ctx}
    Returns \code{TRUE}.

    You should override this method in your application object if you need
    to validate browser requests before processing them. Returning False
    will prevent the browser request being merged into the local namespace
    and the page_process logic will not be called (see also the description
    of the Albatross Application Model in \ref{app-model}).
    \end{methoddesc}

The updated description will appear after the next Albatross release
(hopefully this week).

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/


From andrewm at object-craft.com.au  Wed Mar 14 12:08:43 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Wed, 14 Mar 2007 12:08:43 +1100
Subject: [albatross-users] Documentation test case fails on amd64 
In-Reply-To: <1160494844.5061.13.camel@localhost> 
References: <1160494844.5061.13.camel@localhost>
Message-ID: <20070314010843.E50C25CC90E@longblack.object-craft.com.au>

>I got a Debian bug report about the albatross build failing on amd64.
>The build fails on a documentation test because the Python built-in
>hash() function doesn't return the same value on all platforms:
>
>Testing "doctest/tags-value1" failed, Line 6: output does not match
>example
>Expected:
>hash('spam') is 1626740519
>
>Got:
>hash('spam') is -1840390907728881881

Thanks for the report, Fabian. I've changed the example to the simpler:

    >>> import albatross
    >>> ctx = albatross.SimpleContext('.')
    >>> ctx.locals.items = ['pencil', 'eraser', 'lunchbox']
    >>> albatross.Template(ctx, '<magic>', '''
    ... There are <al-value expr="len(items)" whitespace> items
    ... ''').to_html(ctx)
    >>> ctx.flush_content()
    There are 3 items

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/


From unclezzzen at gmail.com  Thu Mar 15 14:53:33 2007
From: unclezzzen at gmail.com (Uncle Zzzen)
Date: Thu, 15 Mar 2007 05:53:33 +0200
Subject: [albatross-users] Problem with command-line parsing. Affects OpenID
	auth
Message-ID: <24132b510703142053y1fbc0d0bvd96f5a844e683215@mail.gmail.com>

Hi.
I'm trying to write an Albatross wrapper for OpenID, and one of the
phases of authentication is when the OpenID server redirects to
something like:
http://myserver/login_continue?openid.assoc_handle=%7BHMAC-SHA1%...
Nothing I can do about it. OpenID is not MY standard

openid.assoc_handle gets parsed and this raises an error:

Traceback (most recent call last):
  File "/usr/lib/python2.4/site-packages/albatross/app.py", line 282, in run
    self.merge_request(ctx)
  File "/usr/lib/python2.4/site-packages/albatross/app.py", line 366,
in merge_request
    ctx.merge_request()
  File "/usr/lib/python2.4/site-packages/albatross/context.py", line
332, in merge_request
    self.set_value(name, value)
  File "/usr/lib/python2.4/site-packages/albatross/context.py", line
418, in set_value
    parent = getattr(parent, elem)
AttributeError: Vars instance has no attribute 'openid'
    in al-input field "openid.assoc_handle"

I'm gonna patch the lib or something to make it work, but there should
be a sane way to cope with OpenID

Any help with tricks around this problem will be appreciated.

Cheers,
The Dod


From unclezzzen at gmail.com  Fri Mar 16 02:08:53 2007
From: unclezzzen at gmail.com (Uncle Zzzen)
Date: Thu, 15 Mar 2007 17:08:53 +0200
Subject: [albatross-users] Easy to recreate the OpenID problem on
	object-craft site itself
Message-ID: <24132b510703150808t478039b3u6c786b1d7a01d603@mail.gmail.com>

try http://www.object-craft.com.au/cgi-bin/alsamp/form4/form.py?funny.field=2
and see the error you get.

Maybe there's even a way to inject stuff into object this way. Have no
time to check. Would rather write my app at the moment, but there are
many kids out there ;)

Cheers,
The dod

-- 
"The road tried to cross the chicken, but the chicken was faster" -- Abu-Zeresh

gpg: http://zzzen.com/zzzen.asc
fingerprint(0x7F1BCD50) = 0FEE 7BA9 6F1C 23F9 BACD 0F72 6434 F9A7 7F1B CD50


From andrewm at object-craft.com.au  Fri Mar 16 10:04:08 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Fri, 16 Mar 2007 10:04:08 +1100
Subject: [albatross-users] Problem with command-line parsing. Affects
	OpenID auth 
In-Reply-To: <24132b510703142053y1fbc0d0bvd96f5a844e683215@mail.gmail.com> 
References: <24132b510703142053y1fbc0d0bvd96f5a844e683215@mail.gmail.com>
Message-ID: <20070315230408.E49E76007B0@longblack.object-craft.com.au>

>I'm trying to write an Albatross wrapper for OpenID, and one of the
>phases of authentication is when the OpenID server redirects to
>something like:
>http://myserver/login_continue?openid.assoc_handle=%7BHMAC-SHA1%...
>Nothing I can do about it. OpenID is not MY standard
>
>openid.assoc_handle gets parsed and this raises an error:
>
>Traceback (most recent call last):
>  File "/usr/lib/python2.4/site-packages/albatross/app.py", line 282, in run
>    self.merge_request(ctx)
>  File "/usr/lib/python2.4/site-packages/albatross/app.py", line 366,
>in merge_request
>    ctx.merge_request()
>  File "/usr/lib/python2.4/site-packages/albatross/context.py", line
>332, in merge_request
>    self.set_value(name, value)
>  File "/usr/lib/python2.4/site-packages/albatross/context.py", line
>418, in set_value
>    parent = getattr(parent, elem)
>AttributeError: Vars instance has no attribute 'openid'
>    in al-input field "openid.assoc_handle"
>
>I'm gonna patch the lib or something to make it work, but there should
>be a sane way to cope with OpenID
>
>Any help with tricks around this problem will be appreciated.

The simplest answer is to put an object called openid into ctx.locals.
After the page loads, the dummy object will have an attribute
"assoc_handle", and you can do with this what you please.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/


From andrewm at object-craft.com.au  Fri Mar 16 11:47:24 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Fri, 16 Mar 2007 11:47:24 +1100
Subject: [albatross-users] Easy to recreate the OpenID problem on
	object-craft site itself 
In-Reply-To: <24132b510703150808t478039b3u6c786b1d7a01d603@mail.gmail.com> 
References: <24132b510703150808t478039b3u6c786b1d7a01d603@mail.gmail.com>
Message-ID: <20070316004724.3E4D26007B0@longblack.object-craft.com.au>

>try http://www.object-craft.com.au/cgi-bin/alsamp/form4/form.py?funny.field=2
>and see the error you get.
>
>Maybe there's even a way to inject stuff into object this way. Have no
>time to check. Would rather write my app at the moment, but there are
>many kids out there ;)

This is how it's meant to work. I was hoping to point you at the section
of the manual that talks about this, but it doesn't seem to exist,
which is a bit unfortunate.

The code that implements this is in context.py, NamespaceMixin.set_value.
You will see there is a fairly complex state machine that emulates a
subset of python expression parsing (. and []).

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/


From unclezzzen at gmail.com  Fri Mar 16 10:39:33 2007
From: unclezzzen at gmail.com (Uncle Zzzen)
Date: Fri, 16 Mar 2007 01:39:33 +0200
Subject: [albatross-users] Re: Problem with command-line parsing. Affects
	OpenID auth
In-Reply-To: <24132b510703151638k5b0f352csc9ab0ca29f2fc733@mail.gmail.com>
References: <24132b510703142053y1fbc0d0bvd96f5a844e683215@mail.gmail.com>
	<20070315230408.E49E76007B0@longblack.object-craft.com.au>
	<24132b510703151638k5b0f352csc9ab0ca29f2fc733@mail.gmail.com>
Message-ID: <24132b510703151639t71630816y109c8d4ffb223bc4@mail.gmail.com>

On 3/16/07, Andrew McNamara <andrewm at object-craft.com.au> wrote:
> The simplest answer is to put an object called openid into ctx.locals.
> After the page loads, the dummy object will have an attribute
> "assoc_handle", and you can do with this what you please.

It worked.
I've subclassed the context and added
self.locals.openid=albatros.context.Vars()
at the end of __init__
Thanks.

--
"What do you give a man who has everything? A condom, unless you also
want to have what he's got" -- Um Zeresh

gpg: http://zzzen.com/zzzen.asc
fingerprint(0x7F1BCD50) = 0FEE 7BA9 6F1C 23F9 BACD 0F72 6434 F9A7 7F1B CD50


-- 
"What do you give a man who has everything? A condom, unless you also
want to have what he's got" -- Um Zeresh

gpg: http://zzzen.com/zzzen.asc
fingerprint(0x7F1BCD50) = 0FEE 7BA9 6F1C 23F9 BACD 0F72 6434 F9A7 7F1B CD50


From andrewm at object-craft.com.au  Mon Mar 19 17:01:08 2007
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Mon, 19 Mar 2007 17:01:08 +1100
Subject: [albatross-users] Albatross 1.36 released
Message-ID: <20070319060108.8D3581D404C@longblack.object-craft.com.au>

OVERVIEW

Albatross is a small toolkit for developing highly stateful web
applications.

The toolkit has been designed to take a lot of the pain out of
constructing intranet applications although you can also use Albatross
for deploying publicly accessed web applications.

In slightly more than 5300 lines of Python (according to pycount) you
get the following:

  * An extensible HTML templating system similar to DTML including tags for:
    - Conditional processing.
    - Macro definition and expansion.
    - Sequence iteration and pagination.
    - Tree browsing.
    - Lookup tables to translate Python values to arbitrary template
      text.

  * Application classes which offer the following features:
    - Optional server side or browser side sessions.
    - The ability to place Python code for each page in a dynamically
      loaded module, or to place all page processing code in a single
      mainline.

  * The ability to deploy applications as CGI, FastCGI, mod_python or
    a pure python HTTP server by changing less than 10 lines of code.

The toolkit application functionality is defined by a collection of
fine grained mixin classes.  Nine different application types and six
different execution contexts are prepackaged, you are able to define
your own drop in replacements for any of the mixins to alter any
aspect of the toolkit semantics.

Application deployment is controlled by your choice of either cgi,
FastCGI, mod_python, or BaseHTTPServer Request class.  It should be
possible to develop a Request class for Medusa or Twisted to allow
applications to be deployed on those platforms with minimal changes.

Albatross comes with over 180 pages of documentation.  HTML and PDF
formatted documentation is available from the toolkit homepage.

The toolkit homepage:

        http://www.object-craft.com.au/projects/albatross/

The Albatross mailing list subscription and archives:

	http://object-craft.com.au/cgi-bin/mailman/listinfo/albatross-users

CHANGES SINCE 1.35:
==================-

New Features
------------

* A "vars" attribute has been added to <al-for> - this inserts the
  iterator value into the local namespace. The iterator value was
  previously only accessible via the iterator value() method.

* <al-for> now loops over sequences using the iterator protocol by
  default. The previous behaviour of indexing the sequence is retained
  where pagination or columns are requested.

* New <al-setdefault> tag which allows macros to define default values
  for their arguments (suggested by Greg Bond).

* Macro expansion can now specify macro arguments via attributes on the
  <al-expand> tag (suggested by Greg Bond).

* Added get_param() method to Request classes, which deployment method
  agnostic access to http "environment" variables.

Functional Changes
------------------

* NameRecorderMixin no longer falls back to merging all request
  fields if no __albform__ field is present. Applications using the
  NameRecorderMixin and GET-style requests will need to explicitly merge
  relevent fields with the ctx.merge_vars(...) method.

* Pickles embedded in hidden fields are now signed using the HMAC-SHA1
  algorithm (was MD5).

* The experimental FastCGI driver included in version 1.35 now becomes
  the default. The old driver has been renamed fcgiappold. This driver
  implements the FastCGI protocol itself, rather than depending on an
  external module.

Bug Fixes
---------

* If the <al-option> tag was not immediately enclosed within an
  <al-select> tag, it would be silently ignored. <al-option> now works
  within containing flow control tags such as <al-for> or <al-if>,
  and supports attributes for dynamically setting value and label.

* Communication with the session server has been made more robust. EINTR
  is handled, as are partial reads and writes, and requests are restarted
  if the socket closes.

* Mixing radio inputs with other inputs of the same name did not raise
  an exception (reported by Michael Neel).

* fixed FCGI_MAX_DATA, which was incorrect due to an operator precedence
  mistake. Found by Tom Limoncelli.

Miscellaneous Changes
---------------------

* Clean up and replace many pre-python 2.2 constructs.

* Many new tests have been added, and existing tests restructured.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/