[web2py] Re: Include another file, but do not process it

[Daniel Gonzalez]

Thanks Anthony, looks like a good solution.

On Sunday, February 15, 2015 at 2:35:23 AM UTC+1, Anthony wrote:
>
> Maybe something like this:
>
> In a model file or module:
>
> def include_hbs(view_path):
> import os
> response.write(open(os.path.join(request.folder, 'views', view_path), 
> 'rb').read(),
>escape=False)
>
> In the view:
>
> {{include_hbs('dashboard.hbs')}}
>
> You might also add an option to cache in RAM for some period of time.
>
> Anthony
>
> On Saturday, February 14, 2015 at 6:42:38 AM UTC-5, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> This is my use case: I have some web2py templates which are serving some 
>> views with embedded handlebar templates, which are intended for processing 
>> on the client side: they must arrive untouched to the browser.
>> The problem is that handlebars and web2py have by default the same 
>> delimiters: {{ }}
>>
>> I can change these default settings for either web2py or handlebars, but 
>> this has some unintended consequences. For example, I need to go around 
>> configuring tools to understand these changes, like code highlighters 
>> (web-mode in emacs, etc) which is not always obvious / possible to do.
>>
>> So, I would prefer to stick to the default delimiters. What I would like 
>> is something like this:
>>
>> 
>>
>> ...
>>
>> {{include 'dashboard.hbs'}}
>>
>>
>> 
>>
>>
>> {{sdfsdf}}
>>
>> But I would like the web2py engine not to process dashboard.hbs, and 
>> serve it as-is.
>> Web2py could make this decision based on the extension of the file ("do 
>> not process .hbs files") or with a flag to disable processing for a 
>> specific include: {{include 'dashboard.hbs' dont-process}}
>> Is this implemented? I have found no documentation regarding the web2py 
>> include statement (there is documentation scattered around the guide, but 
>> nothing specific as far as I can see).
>>
>> Thanks!
>> DanG
>>
>>

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] Include another file, but do not process it

[Daniel Gonzalez]

Hi,

This is my use case: I have some web2py templates which are serving some 
views with embedded handlebar templates, which are intended for processing 
on the client side: they must arrive untouched to the browser.
The problem is that handlebars and web2py have by default the same 
delimiters: {{ }}

I can change these default settings for either web2py or handlebars, but 
this has some unintended consequences. For example, I need to go around 
configuring tools to understand these changes, like code highlighters 
(web-mode in emacs, etc) which is not always obvious / possible to do.

So, I would prefer to stick to the default delimiters. What I would like is 
something like this:



...

{{include 'dashboard.hbs'}}





{{sdfsdf}}

But I would like the web2py engine not to process dashboard.hbs, and serve 
it as-is.
Web2py could make this decision based on the extension of the file ("do not 
process .hbs files") or with a flag to disable processing for a specific 
include: {{include 'dashboard.hbs' dont-process}}
Is this implemented? I have found no documentation regarding the web2py 
include statement (there is documentation scattered around the guide, but 
nothing specific as far as I can see).

Thanks!
DanG

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] Re: define_table does not create table in database

[Daniel Gonzalez]

I have solved my problem. I was not properly reusing the db (DAL object) in 
my library. Instead, a new DAL connection was being created in some cases. 
Therefore, even though the tables were defined at the pyhton level, auth 
was not really aware of it, and was trying to redefine them.

Facit: make sure that a single db object is used in the models and in any 
external modules.

Thanks for your assistance!
Daniel Gonzalez

On Thursday, October 30, 2014 4:29:26 PM UTC+1, Daniel Gonzalez wrote:
>
> Getting closer to identifying my problem. It seems I am having several 
> problems at once.
>
> First, it seems that migrate_enabled=False prevents not only migrations, 
> but also table creation (am I right here?)
>
> Second: while trying to guess the problems with my application, I am using 
>  the  test script provided. Running manually that script, and at the same 
> time working with pgadmin (that is, manually creating / droping 
> databases/tables), causes serious conflicts with web2py. It seems web2py 
> has its own caching mechanism on the file system for the table definitions 
> (pickled objects?)
>
> I have now clarified that, and to mitigate those problems I am doing: 
> 1. drop database, remove pickled files, restart web2py
> 2. Do not use lazy_tables and enable migrations (this ensures that the 
> tables will be created exactly on the call to define_tables)
> 3. Call .commit(). I have the code to define_tables outside in a module, 
> and according to the manual, no autocommit is performed for modules. The 
> reason I am doing this is that I want to be able to reuse the same code 
> from outside my web2py application, for maintenance/reporting purposes: I 
> have a library with all web2py related functionality, which I can call from 
> the web2py application itself, or simply from an external script.
>
> But still, I see a problem: I can see clearly in the sql.log file that my 
> define_table creates the table as I want, but immediately after, 
> auth.define_tables() modifies that table: drops my fields, adds the 
> standard auth fields.
>
> According to the manual (
> http://web2py.com/books/default/chapter/29/9#Customizing-Auth):
>
> If a table is declared beforeauth.define_tables() it is used instead of 
> the default one
>
> I would expect that Auth sees that auth_user is already defined, skip the 
> definition of that table, and define the other auth tables. But in my case 
> it seems auth is overwrriting the table defintion. Why can this be?
>
> BR,
> Daniel
>
>
> On Thursday, October 30, 2014 1:04:27 PM UTC+1, Anthony wrote:
>>
>> On Thursday, October 30, 2014 7:41:07 AM UTC-4, Leonel Câmara wrote:
>>>
>>> Well lazy_tables=True is the default.
>>>
>>
>> Actually, it does default to False, but that's not the issue here.
>>
>> Anthony 
>>
>

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] Re: define_table does not create table in database

[Daniel Gonzalez]

Getting closer to identifying my problem. It seems I am having several 
problems at once.

First, it seems that migrate_enabled=False prevents not only migrations, 
but also table creation (am I right here?)

Second: while trying to guess the problems with my application, I am using 
 the  test script provided. Running manually that script, and at the same 
time working with pgadmin (that is, manually creating / droping 
databases/tables), causes serious conflicts with web2py. It seems web2py 
has its own caching mechanism on the file system for the table definitions 
(pickled objects?)

I have now clarified that, and to mitigate those problems I am doing: 
1. drop database, remove pickled files, restart web2py
2. Do not use lazy_tables and enable migrations (this ensures that the 
tables will be created exactly on the call to define_tables)
3. Call .commit(). I have the code to define_tables outside in a module, 
and according to the manual, no autocommit is performed for modules. The 
reason I am doing this is that I want to be able to reuse the same code 
from outside my web2py application, for maintenance/reporting purposes: I 
have a library with all web2py related functionality, which I can call from 
the web2py application itself, or simply from an external script.

But still, I see a problem: I can see clearly in the sql.log file that my 
define_table creates the table as I want, but immediately after, 
auth.define_tables() modifies that table: drops my fields, adds the 
standard auth fields.

According to the manual 
(http://web2py.com/books/default/chapter/29/9#Customizing-Auth):

If a table is declared beforeauth.define_tables() it is used instead of the 
default one

I would expect that Auth sees that auth_user is already defined, skip the 
definition of that table, and define the other auth tables. But in my case 
it seems auth is overwrriting the table defintion. Why can this be?

BR,
Daniel


On Thursday, October 30, 2014 1:04:27 PM UTC+1, Anthony wrote:
>
> On Thursday, October 30, 2014 7:41:07 AM UTC-4, Leonel Câmara wrote:
>>
>> Well lazy_tables=True is the default.
>>
>
> Actually, it does default to False, but that's not the issue here.
>
> Anthony 
>

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] Re: define_table does not create table in database

[Daniel Gonzalez]

One more thing: you mention tables are lazy, but the documentation says 
they are lazy only whenever DAL(...,lazy_tables=True) is set (DAL 
documentation 

)
Which one is right?


-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] Re: define_table does not create table in database

[Daniel Gonzalez]

Thanks for your reply,

Migrate is not the problem: it does not matter if I do 
migrate_enabled=False or migrate_enabled=True.
Besides, there is no table to migrate. That is the whole problem: 
define_table does not create the table!

I am doing my own define_table because I have need to trigger some other 
code whenever new users are created (not shown here)

Why does db.tables show that the table is created, when it is not? 
(verified with pgadmin)

BR,
Daniel

On Thursday, October 30, 2014 11:49:25 AM UTC+1, Leonel Câmara wrote:
>
> Well the tables are lazy, but that shouldn't cause your problem because 
> once you use them they really get defined. May I ask why you are defining 
> your own user table instead of using auth.define_tables()?
>
> Anyway, your problem is caused by migrate_enabled=False.
>

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[web2py] define_table does not create table in database

[Daniel Gonzalez]

I am running `define_tables` in the recommended way:

db = DAL('postgres://user:@localhost:5432/mydb', 
migrate_enabled=False, auto_import=False, lazy_tables=True)
db.define_table('auth_user',
Field('email', unique=True),
Field('password', length=512, type='password', readable=False, 
label='Password'),
...)

This gets executed without errors, but no table is created in the database. 
Whenever I try to insert a new user:

relation "auth_user" does not exist

What can be going on? Once the tables are created (manually, for example), 
the application works fine. I am using a postgres backend. This happens no 
matter what value I give to `lazy_tables`

EDIT


This is the full test script:

from gluon import DAL
from gluon import Field

db = DAL('postgres://user:pass@localhost:5432/mydb', 
migrate_enabled=False)

db.define_table(
'auth_user',
Field('email',  type='string', unique=True),
Field('password',   type='password'),
Field('registration_key',   type='string',   length=512, 
writable=False, readable=False, default=''),
Field('reset_password_key', type='string',   length=512, 
writable=False, readable=False, default=''),
Field('registration_id',type='string',   length=512, 
writable=False, readable=False, default=''),
)
db.commit()

print db.tables

db.auth_user.insert(email='g@b.c')

And I get the following output:

['auth_user']
Traceback (most recent call last):
  File "xxx.py", line 19, in 
db.auth_user.insert(email='g@b.c')
  File "/tmp/web2py/gluon/dal.py", line 9293, in insert
ret =  self._db._adapter.insert(self, self._listify(fields))
  File "/tmp/web2py/gluon/dal.py", line 1361, in insert
raise e
psycopg2.ProgrammingError: relation "auth_user" does not exist
LINE 1: INSERT INTO auth_user(reset_password_key,registration_id,reg...

The table is somehow "created" (in memory?), but it is not really in the 
postgres database. What does this mean?

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [web2py] Re: Proposal: create "namespaces" for translation system

[Daniel Gonzalez Zaballos]

The idea is exactly for this. To have separated localized components. I
hope that in a few weeks/months we can release something that we are
working on components. I cannot say to much now :)


2013/9/23 Massimo Di Pierro 

> I was skeptical but I have been convinced this is an excellent idea. Your
> patch is going to trunk now. Thank you!
>
>
> On Monday, 23 September 2013 02:13:10 UTC-5, jamarcer wrote:
>
>> Hello:
>>
>> I'm Demetrio's mate, and he asked me to upgrade the T with namespaces
>> functionality into trunk (Version 2.6.3-stable+timestamp.2013.**
>> 09.18.14.36.14).
>>
>> I have added a comment into 1196 issue with a code patch and a
>> explanation.
>>
>> https://code.google.com/p/**web2py/issues/detail?id=1196
>>
>> I hope it will be usefull.
>>
>> Best regards.
>>
>  --
> Resources:
> - http://web2py.com
> - http://web2py.com/book (Documentation)
> - http://github.com/web2py/web2py (Source code)
> - https://code.google.com/p/web2py/issues/list (Report Issues)
> ---
> You received this message because you are subscribed to the Google Groups
> "web2py-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to web2py+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>

-- 
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- http://github.com/web2py/web2py (Source code)
- https://code.google.com/p/web2py/issues/list (Report Issues)
--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [web2py] Re: Proposal: create "namespaces" for translation system

[Daniel Gonzalez Zaballos]

Hello Massimo,

We'll try to have it at the end of the month. We had applied it to
production and is a 1.99.7 version and the trunk version has changed a lot.

One question, what do you mean with use "." instead of "app_name"?

Regards


2013/9/1 Massimo Di Pierro 

> Hello Demetrio,
>
> sorry for the late response. Could you please resend your patch, make sure
> it applies to trunk, and use "." instead of "app_name"?
>
>
> On Thursday, 29 November 2012 06:17:22 UTC-6, demetrio wrote:
>
>> Hi everyone I post a reply in the web3py thread in this mailing list
>> asking for a site where make a wishlist, but thinking about what I would
>> want in web3py, I realized that I needed in web2py already so I have
>> make a patch for the actual translation system in the gluon/languages.py
>> file.
>>
>> One of the lacks that I see in the translation system is that is a
>> little difficult to maintain the language file in large applications
>> like the one I am developing at this moment. Think about modules that
>> you can plug or unplug of your application, and each have their own
>> strings, or something more common, the pluggin system.
>>
>> Actually if you want to install a plugin like plugin_wiki you have to do
>> your own translation because you cannot insert translation files (if
>> actually web2py supports this... do it I don't know how to make it xD).
>> It would be great to have a community where you can make plugins and add
>> the language files inside.
>>
>> This is what makes the patch:
>>
>> Allows to use a "namespace" parameter in the T() like this
>>
>> T("Hello World", namespace="plugin_wiki")
>>
>> So in the languages/ folder will appear this:
>>
>> - languages/
>> - en_en.py
>> - plugin_wiki/
>> - en_en.py
>>
>> so if you use T("Hello World", namespace="plugin_wiki") it will go to
>> find the string into languates/plugin_wiki/*your_**lang*.py and if you
>> use
>> the common way T("Hello World") it will go as always to
>> languages/*your_lang*.py
>>
>> I haven't tested it enough but it works for me at now, and I haven't
>> written the part for the admin interface where you can push the button
>> for refreshing the lang files (I don't get on wiht the RegExp at all)
>>
>> My 2 cents
>> I hope that I can send the patch to the list
>>
>  --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "web2py-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to web2py+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Catch-all function in controller

[Daniel Gonzalez]

Hi,

I have an api.py module which is the interface to my restful api. Currently 
I have to do the following to extend my api:

def common(obj):
response.view = 'generic.json'

response.headers['Content-Type'] = CONTENT_TYPE_JSON
def GET(*args,**vars):

return api_router(API_GET, obj, *args, **vars)
def POST(*args,**vars):
return api_router(API_POST, obj, *args, **vars)
def PUT(*args,**vars):
return api_router(API_PUT, obj, *args, **vars)
def DELETE(*args, **vars):
return api_router(API_DELETE, obj, *args, **vars)
return dict(GET=GET, POST=POST, PUT=PUT, DELETE=DELETE)


@request.restful()
def index():
return common('root')


@request.restful()
def nodes():
return common('nodes')


@request.restful()
def agents():
return common('agents')
...


For any new object that I want to offer via my api, I need to list it as 
@request.restful funcion, and link to common (the common api 
implementation). I would like to define this common function as the 
"catch-all" restful function for my api. My api_router already knows which 
objects are supported, and replies with 404 if a requested object is not 
supported, so it makes no sense for me to be forced to do this again on the 
controller.

How can I define this "catch-all" function in my api controller?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: URL for static assets produces wrong value?

[Daniel Gonzalez]

Oos. Sorry, sorry, sorry. My bad. My router rules had an slash too 
many. Here are the correct ones:

routes_in = (
('/admin(?P.*)',   '/admin\g'),
('/app1(?P.*)','/app1\g'),
('/(?P.*)','/app2/\g'),
)
routes_out = (
('/admin(?P.*)',  '\g'),
('/app1(?P.*)',   '\g'),
('/app2/(?P.*)',  '/\g'),
)

And that works. Sorry again.

On Thursday, May 30, 2013 8:12:26 PM UTC+2, Daniel Gonzalez wrote:
>
> Hi,
>
> I have a strange use case, which is maybe not covered by the router / URL 
> implementation.
>
> I have just started using a second application, and I have modified my 
> routes.py:
>
> routes_in = (
> ('/admin(?P.*)',   '/admin\g'),
> ('/app1(?P.*)','/app1\g'),
> ('/(?P.*)','/app2/\g'),
> )
> routes_out = (
> ('/admin(?P.*)',  '/\g'),
> ('/app1(?P.*)',   '/\g'),
> ('/app2/(?P.*)',  '/\g'),
> )
>
> (app2 is my default app, which should be accessible at the root of the url)
>
> In my app1 I am doing something like this:
>
> URL('static','images/image01.png')
>
> And I was expecting a url like: /app1/static/images/image01.png
>
> Instead, I am getting: //static/images/image01.png
>
> So the application part is skipped. Is this normal? Can I solve this 
> problem somehow?
>
> Thanks,
> Daniel
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] URL for static assets produces wrong value?

[Daniel Gonzalez]

Hi,

I have a strange use case, which is maybe not covered by the router / URL 
implementation.

I have just started using a second application, and I have modified my 
routes.py:

routes_in = (
('/admin(?P.*)',   '/admin\g'),
('/app1(?P.*)','/app1\g'),
('/(?P.*)','/app2/\g'),
)
routes_out = (
('/admin(?P.*)',  '/\g'),
('/app1(?P.*)',   '/\g'),
('/app2/(?P.*)',  '/\g'),
)

(app2 is my default app, which should be accessible at the root of the url)

In my app1 I am doing something like this:

URL('static','images/image01.png')

And I was expecting a url like: /app1/static/images/image01.png

Instead, I am getting: //static/images/image01.png

So the application part is skipped. Is this normal? Can I solve this 
problem somehow?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Missing ASIDE tag in web2py gluon

[Daniel Gonzalez]

The ASIDE  helper (HTML5) is 
missing in gluon (using web2py 2.7.4). Useful to do something like this: 
http://purecss.io/layouts/gallery/

I have defined it in terms of DIV: 

class ASIDE(DIV):

tag = 'aside'

And works for me. I am not sure if gluon is supposed to implement all 
possible html tags, or if only a selected few are provided.

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Strategies to select the preferred language

[Daniel Gonzalez]

And a final question: since I will only be supporting a handful of 
languages (english, german and italian) I would like to set the current 
languages. I have found T.set_current_languages. Is this the right function 
to set the supported languages? Is there an example somewhere on how to use 
this function?

I would go with something like: T.set_current_languages(['en', 'it', 'de']) 
in my model, but I am not sure that is the right way to use it.
I have tried (in models/0.py):

T.set_current_languages(['it', 'de', 'en'])

But I get:

 [Errno 2] No such file or directory: 
/xxx/applications/example/languages/en-us.py'

I also get that if I remove 'en': T.set_current_languages(['it', 'de'])

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Strategies to select the preferred language

[Daniel Gonzalez]

You are right. I guess I am having trouble myself because I *am* trying to 
force the language (which is something a user will not do. since they are 
happy with the default browser setting). And I have just found this piece 
of code in admin:

{{if hasattr(T,'get_possible_languages_info'):}}
- {{=T('Admin language')}}

{{for langinfo in sorted([(code,info[1]) for code,info 
in T.get_possible_languages_info().iteritems() if code != 'default']):}}
{{=langinfo[1]}}
{{pass}}

{{else:}}
{{pass}}

And the model:

# set the language
if 'adminLanguage' in request.cookies and not (request.cookies[
'adminLanguage'] is None):
T.force(request.cookies['adminLanguage'].value)

Let me get this straight: a cookie adminLanguage is set with 100 days 
expiration date. This cookie is then accessible via request. Sounds simple 
enough, I guess I can reuse this. I do not get the  part, but I 
guess is just boilerplate to create the options and force a reload of the 
page when a selection is performed.

On Wednesday, May 29, 2013 9:06:56 AM UTC+2, Niphlod wrote:
>
> ? never had issues with users  usually they download the browser with 
> the language they're in. 
> BTW, forcing a language using a session variable is not hard at all in 
> web2py.
>
> Il giorno mercoledì 29 maggio 2013 09:00:05 UTC+2, Daniel Gonzalez ha 
> scritto:
>>
>> Hi,
>>
>> According to the 
>> documentation<http://web2py.com/books/default/chapter/29/04#Internationalization,-and-Pluralization-with-T>,
>>  
>> the T operator works based on the Accept-Language setting. In my 
>> experience, this is quite cumbersome for a user to set-up (I have not even 
>> succeeded in setting this up in Chrome), so I would prefer to rely on 
>> offering the user a limited set of available languages for my web 
>> application, in the form of a menu, or maybe a profile setting.
>>
>> For users without profile (i.e. not logged in), I would like to persist 
>> this setting across requests. How could I do this?
>>
>> What are the strategies being used by others offering pages in multiple 
>> languages?
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Strategies to select the preferred language

[Daniel Gonzalez]

Hi,

According to the 
documentation,
 
the T operator works based on the Accept-Language setting. In my 
experience, this is quite cumbersome for a user to set-up (I have not even 
succeeded in setting this up in Chrome), so I would prefer to rely on 
offering the user a limited set of available languages for my web 
application, in the form of a menu, or maybe a profile setting.

For users without profile (i.e. not logged in), I would like to persist 
this setting across requests. How could I do this?

What are the strategies being used by others offering pages in multiple 
languages?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Support for internationalization in web2py wiki

[Daniel Gonzalez]

In the same way as internationalization is handled in the T operator? That 
means that an "accept-language: it- it" would tell the wiki engine to look 
for pages corresponding to the URL which are marked in the database as 
being in italian. That means:

   1. a new field is needed in the database to mark the language the page 
   is written in
   2. the wiki engine would select that page
   3. a fall back mechanism would return a page in a default language 
   (english?)
   4. when creating new wiki entries, the support views would offer a 
   drop-down to specifiy the language in which the page is being created 
   (which would default to the accept-language setting)
   

On Tuesday, May 28, 2013 10:29:27 PM UTC+2, Massimo Di Pierro wrote:
>
> There is not. The problem is that normal internationalization libraries 
> are not appropriate for wiki pages. I would recommended using different 
> pages for different languages. In your opinion, how should 
> internationalization for wiki pages work?
>
> On Tuesday, 28 May 2013 14:35:21 UTC-5, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I have started to use the web2py wiki module, but can not find any 
>> support for internationalization of the wiki pages. What I would like is to 
>> be able to select the language based on some query parameters or some 
>> configuration parameters. Of course, the pages must be already be saved in 
>> the desired language.
>>
>> Is there any support for this built-in?
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Support for internationalization in web2py wiki

[Daniel Gonzalez]

Hi,

I have started to use the web2py wiki module, but can not find any support 
for internationalization of the wiki pages. What I would like is to be able 
to select the language based on some query parameters or some configuration 
parameters. Of course, the pages must be already be saved in the desired 
language.

Is there any support for this built-in?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Create URL with embedded json object

[Daniel Gonzalez]

Hi,

In one of my views I am creating a table. During table creation time I have 
some dictionaries that I want to return as json objects. So in column 
"data" I want to provide a link to download this json objects. These 
objects are not stored anywhere (they are dinamically created), and are 
thus only known during the table instantiation time.

Is it possible to do something like this:

my_link = A(URL(data))

So that clicking the link will not perform an access to the web2py (no 
controller needs to be accessed, the data is already in the URL), but just 
show the json data?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Handling OPTIONS request in a web2py RESTful API

[Daniel Gonzalez]

Hi,

I have implemented a RESTful api in web2py which handle the usual GET, 
POST, PUT, DELETE. (using @request.restful())
Is there a standard mechanism in web2py to implement OPTIONS requests?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [web2py] Re: track_changes bug or feature?

[Daniel Gonzalez Zaballos]

Thanks! I said it because restarting the server with every little change in
a development environment, is a little weird. I'll change some imports to
make them inside methods.


2013/4/6 Massimo Di Pierro 

> This is intended. classone is importend once and it imports classtwo. This
> happens only once before http requests arrive. Python caches modules.
>
> track changes only affect import done explicitly at runtime (when http
> requests are processed).
>
> Changing this would require that web2py keep track of all dependencies
> (who import who). This would slow down everything.
>
> Massimo
>
>
> On Monday, 25 March 2013 09:55:19 UTC-5, demetrio wrote:
>>
>> Hi everyone,
>>
>> I have notice an strange behaviour of the track_changes feature (working
>> on web2py 2.3.2 and python 2.7.3).
>>
>> Imagine the following structure:
>>
>> modules/
>> ├── classone.py
>> ├── __init__.py
>> └── mymodule
>> ├── classtwo.py
>> └── __init__.py
>>
>> (assuming that there is a models/0.py file with the track_changes(True)
>> statement)
>>
>> In the file classone.py I have the following:
>> #=**=
>> #!/usr/bin/env python
>> # coding: utf8
>> from gluon import *
>> *from mymodule.classtwo import ClassTwo*
>>
>> class ClassOne(object):
>> def say_something(self):
>> classtwo = ClassTwo()
>> return classtwo.say_something()
>> #=**=
>>
>> And in classtwo.py this:
>> #=**=
>> #!/usr/bin/env python
>> # coding: utf8
>> from gluon import *
>>
>> class ClassTwo(object):
>> def say_something(self):
>> return "Hi!"
>> #=**=
>>
>> In this case, the changes inside ClassTwo are not tracked, but if I
>> change class one to:
>>
>> #=**=
>> #!/usr/bin/env python
>> # coding: utf8
>> from gluon import *
>>
>> class ClassOne(object):
>> def say_something(self):
>> *from mymodule.classtwo import ClassTwo*
>> classtwo = ClassTwo()
>> return classtwo.say_something()
>> #=**=
>>
>> Now it works correctly and all the changes are tracked, notice that now
>> the import statement is inside of a method.
>>
>> In the first case, the custom_importer() function is only called at the
>> first execution to import ClassTwo, but in the second case,
>> custom_importer its called in all executions.
>>
>> I have made an example application with this issue attached to this mail.
>>
>> Is this intended or it is a bug?
>>
>> greetings, Daniel.
>>
>>  --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "web2py-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to web2py+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Access data in PUT request for @request.restful()

[Daniel Gonzalez]

I see. So I assume I will have to parse the json myself.
I wonder if after @request.restful() has done its job, I can still access 
the body.
I need to parse the json data, but maybe the body is already consumed ...

On Wednesday, April 3, 2013 9:24:47 AM UTC+2, Niphlod wrote:
>
> the restful part parses the data as web2py does, so accepts form/* encoded 
> parameters... it doesn't parse json in the body.
>
> On Wednesday, April 3, 2013 8:19:04 AM UTC+2, Daniel Gonzalez wrote:
>>
>> Sorry to come back to this after two weeks (was busy with other stuff). I 
>> am trying to access the post_vars, but they are emtpy (Storage {})
>> My PUT request body is coming from an emberjs client, and looks like this 
>> (this is just test data):
>>
>>
>> {"node":{"type":"voicemenu","name":"sdfsfd","outputs":{"jumpIfBusy":null,"startnode":null,"jumpIfNoAnswer":null,"exten":null},"properties":{"record":null,"multilocator":null,"locator":null}}}
>>
>> But this data is not in post_vars. I do not understand very well what you 
>> mean by:
>>
>> "basically request.post_vars are the variables in body whether it is PUT 
>> or POST or other."
>>
>> I do not have "variables in the body", I have json data. This should be 
>> parsed as json. Should I do it manually, or is @request.restful() doing 
>> that automatically? (I can not see it in the code).
>>
>> Thanks,
>> Daniel
>>
>> On Friday, March 22, 2013 6:02:32 AM UTC+1, Massimo Di Pierro wrote:
>>>
>>> if request.env.request_method=='POST':
>>> put_vars = request.post_vars
>>>
>>> basically request.post_vars are the variables in body whether it is PUT 
>>> or POST or other.
>>>
>>>
>>>
>>> On Thursday, 21 March 2013 19:10:27 UTC-5, Daniel Gonzalez wrote:
>>>>
>>>> Hi,
>>>>
>>>> How can I access the data in the PUT request? (in the body of the 
>>>> request)
>>>> Can this be automatically parsed (my data is JSON)
>>>>
>>>> Thanks,
>>>> Daniel
>>>>
>>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Access data in PUT request for @request.restful()

[Daniel Gonzalez]

Sorry to come back to this after two weeks (was busy with other stuff). I 
am trying to access the post_vars, but they are emtpy (Storage {})
My PUT request body is coming from an emberjs client, and looks like this 
(this is just test data):

{"node":{"type":"voicemenu","name":"sdfsfd","outputs":{"jumpIfBusy":null,"startnode":null,"jumpIfNoAnswer":null,"exten":null},"properties":{"record":null,"multilocator":null,"locator":null}}}

But this data is not in post_vars. I do not understand very well what you 
mean by:

"basically request.post_vars are the variables in body whether it is PUT or 
POST or other."

I do not have "variables in the body", I have json data. This should be 
parsed as json. Should I do it manually, or is @request.restful() doing 
that automatically? (I can not see it in the code).

Thanks,
Daniel

On Friday, March 22, 2013 6:02:32 AM UTC+1, Massimo Di Pierro wrote:
>
> if request.env.request_method=='POST':
>     put_vars = request.post_vars
>
> basically request.post_vars are the variables in body whether it is PUT 
> or POST or other.
>
>
>
> On Thursday, 21 March 2013 19:10:27 UTC-5, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> How can I access the data in the PUT request? (in the body of the request)
>> Can this be automatically parsed (my data is JSON)
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Access data in PUT request for @request.restful()

[Daniel Gonzalez]

Thanks Massimo,

I realize now that those variables are also in **vars (request.restful 
interface<http://www.web2pyslices.com/slice/show/1533/restful-api-with-web2py>
)
I wonder how conflicts between query parameters and body parameters are 
resolved.

BR,
Daniel

On Friday, March 22, 2013 6:02:32 AM UTC+1, Massimo Di Pierro wrote:
>
> if request.env.request_method=='POST':
> put_vars = request.post_vars
>
> basically request.post_vars are the variables in body whether it is PUT 
> or POST or other.
>
>
>
> On Thursday, 21 March 2013 19:10:27 UTC-5, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> How can I access the data in the PUT request? (in the body of the request)
>> Can this be automatically parsed (my data is JSON)
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Access data in PUT request for @request.restful()

[Daniel Gonzalez]

Hi,

How can I access the data in the PUT request? (in the body of the request)
Can this be automatically parsed (my data is JSON)

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Slow DAL initialization

[Daniel Gonzalez]

Great, thanks Niphlod!

Now the setup times have been reduced to the <3ms (with some cases of >3ms, 
and even less over 10ms). I would say, discarding the initial pool setup, 
the average is around 4 ms, with some spikes around 10ms (these are not 
accurate stats)

Anyway, looks like an improvement.

On Thursday, March 14, 2013 2:13:51 PM UTC+1, Niphlod wrote:
>
> pool ?
>
> On Thursday, March 14, 2013 1:16:25 PM UTC+1, Daniel Gonzalez wrote:
>>
>> This codes in the model takes between 5ms and 24ms to complete:
>>
>> WEB2PY_DAL_STORAGE="postgres://myuser:mypass@localhost:5432/mydb"
>> db = DAL(WEB2PY_DAL_STORAGE, migrate_enabled=False, lazy_tables=True)
>>
>> I am in a laptop with:
>>
>>- 2 cores, 3300 bogomips/core
>>- 4 GB
>>- linux 2.6.32-5-686
>>
>> Is this expected?
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

Thanks Jonathan, that makes sense.

On Thursday, March 14, 2013 1:54:35 PM UTC+1, Jonathan Lundell wrote:
>
> On 14 Mar 2013, at 4:45 AM, Daniel Gonzalez > 
> wrote: 
> > Thanks Niphlod, I am making progress thanks to your suggestions. 
> > 
> > I have a strange effect: if I do 
> > 
> > def finish_request(xxx): # It receives None as parameter, I do not know 
> why 
> > print "yadayada" 
> > 
> > response.custom_commit = finish_request 
> > 
> > My function gets called, at the end of the request, *twice* in rapid 
> succession. Is this expected? Maybe a web2py bug? (using commit 10c67e5, 
> from 4 days ago) 
> > 
>
> It's normal, IIRC. It's called with different arguments, once for each db 
> thread and once with None. When I use it (to do some cache flushing after 
> the db commits), my function looks like this: 
>
> def oncommit(_db): 
> "custom_commit callback: commit, then flush the cache" 
> if _db is None: # after all commits 
> flush_active_channels() 
> else: 
> _db.commit()# not called for Google Datastore 
>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Slow DAL initialization

[Daniel Gonzalez]

This codes in the model takes between 5ms and 24ms to complete:

WEB2PY_DAL_STORAGE="postgres://myuser:mypass@localhost:5432/mydb"
db = DAL(WEB2PY_DAL_STORAGE, migrate_enabled=False, lazy_tables=True)

I am in a laptop with:

   - 2 cores, 3300 bogomips/core
   - 4 GB
   - linux 2.6.32-5-686

Is this expected?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

(Strange, your reply arrives per email but is not visible in the google 
groups thread. I quote)

it **should** not happen if you see the code.
> https://github.com/web2py/web2py/blob/master/gluon/main.py#L566
> https://github.com/web2py/web2py/blob/master/gluon/dal.py#L555
> are you sure that the "double execution" is not some generated by a 
> decorator lying around in your code ?


Yes, I have verified the web2py code and it should not happen, that is why 
I am surprised.
And no, I have no old code lying around (that I can see ...). It really 
gets called twice by web2py. I'll have to investigate further.

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

Thanks Niphlod, I am making progress thanks to your suggestions.

I have a strange effect: if I do

def finish_request(xxx): # It receives None as parameter, I do not know why
print "yadayada"

response.custom_commit = finish_request

My function gets called, at the end of the request, *twice* in rapid 
succession. Is this expected? Maybe a web2py bug? (using commit 10c67e5, 
from 4 days ago)

On Thursday, March 14, 2013 12:17:36 PM UTC+1, Niphlod wrote:
>
> as said before, for 1 there's response.custom_commit and for 2 you must 
> render the template inside your function.
>
> On Thursday, March 14, 2013 11:59:26 AM UTC+1, Daniel Gonzalez wrote:
>>
>> But this has two drawbacks:
>>
>>- the decorator must be added to every controller function
>>- it gets executed before rendering the template.
>>
>> On Thursday, March 14, 2013 11:56:53 AM UTC+1, Daniel Gonzalez wrote:
>>>
>>> Are you suggesting something like this (which is what I am currently 
>>> doing):
>>>
>>> In my model I define a decorator:
>>>
>>> def finish_request(fn, *args, **kwargs):
>>> def wrapped():
>>> res = fn(*args, **kwargs)
>>> do_whatever_after_the_request_has_been_processed()
>>> return res
>>> return wrapped
>>>
>>> And then in my controller:
>>>
>>> @finish_request
>>> @auth.requires_login()
>>> def index():
>>> return "Hello"
>>>
>>>
>>>
>>> On Thursday, March 14, 2013 11:48:48 AM UTC+1, Niphlod wrote:
>>>>
>>>> let's dive it into  I don't think there's an integrated way to do 
>>>> it, rather than decorating your function with something else.
>>>> Please note that as far as I know (but I may be totally wrong) you 
>>>> can't actually return from a function (so it gets rendered and transferred 
>>>> to the client) and continue your computationi.e. if you have return 
>>>> 'abcd' to ship 'abcd' to the client, execution stops there.
>>>> So, assuming that you are fine with, e.g.
>>>> def func():
>>>>  result = 'abcd'
>>>>  .some lengthy "after callback"
>>>>  return result
>>>>
>>>> I'd say it's definitely "doable".
>>>> If I remember correctly, there's only one handy shortcut available 
>>>> without starting toying with decorators.... response.custom_commit that 
>>>> can 
>>>> be a callable and gets called at the end of every request (before 
>>>> returning 
>>>> the results) 
>>>>
>>>> PS: if the thing to do "after callback" is fine if executed 
>>>> asynchronously, you can set up a queue of things to do to be processed 
>>>> afterwards.
>>>>
>>>> PS2: to kick in after the rendering you must return the compiled 
>>>> template, e.g.
>>>> result = response.render(...)
>>>> ...some lengthy "after callback"
>>>> return result
>>>>
>>>> On Thursday, March 14, 2013 11:35:20 AM UTC+1, Daniel Gonzalez wrote:
>>>>>
>>>>> And (I am discovering some effects during testing):
>>>>>
>>>>> 4. Is it possible to call this function *after* the templates have 
>>>>> been rendered?
>>>>>
>>>>> On Thursday, March 14, 2013 11:09:51 AM UTC+1, Daniel Gonzalez wrote:
>>>>>>
>>>>>> And:
>>>>>>
>>>>>> 3. Is it possible to call this function also for jsonrpc requests?
>>>>>>
>>>>>> On Thursday, March 14, 2013 11:00:13 AM UTC+1, Daniel Gonzalez wrote:
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> For some tests that I am performing, I would like to call a function 
>>>>>>> "request_processed()" (defined in a model, for example), at the end of 
>>>>>>> every request.
>>>>>>>
>>>>>>> 1. Is this possible? How?
>>>>>>> 2. Is it possible to also call this function in case no controller 
>>>>>>> was found to serve the request?
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Daniel
>>>>>>>
>>>>>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

But this has two drawbacks:

   - the decorator must be added to every controller function
   - it gets executed before rendering the template.

On Thursday, March 14, 2013 11:56:53 AM UTC+1, Daniel Gonzalez wrote:
>
> Are you suggesting something like this (which is what I am currently 
> doing):
>
> In my model I define a decorator:
>
> def finish_request(fn, *args, **kwargs):
> def wrapped():
> res = fn(*args, **kwargs)
> do_whatever_after_the_request_has_been_processed()
> return res
> return wrapped
>
> And then in my controller:
>
> @finish_request
> @auth.requires_login()
> def index():
> return "Hello"
>
>
>
> On Thursday, March 14, 2013 11:48:48 AM UTC+1, Niphlod wrote:
>>
>> let's dive it into  I don't think there's an integrated way to do it, 
>> rather than decorating your function with something else.
>> Please note that as far as I know (but I may be totally wrong) you can't 
>> actually return from a function (so it gets rendered and transferred to the 
>> client) and continue your computationi.e. if you have return 'abcd' to 
>> ship 'abcd' to the client, execution stops there.
>> So, assuming that you are fine with, e.g.
>> def func():
>>  result = 'abcd'
>>  .some lengthy "after callback"
>>  return result
>>
>> I'd say it's definitely "doable".
>> If I remember correctly, there's only one handy shortcut available 
>> without starting toying with decorators response.custom_commit that can 
>> be a callable and gets called at the end of every request (before returning 
>> the results) 
>>
>> PS: if the thing to do "after callback" is fine if executed 
>> asynchronously, you can set up a queue of things to do to be processed 
>> afterwards.
>>
>> PS2: to kick in after the rendering you must return the compiled 
>> template, e.g.
>> result = response.render(...)
>> ...some lengthy "after callback"
>> return result
>>
>> On Thursday, March 14, 2013 11:35:20 AM UTC+1, Daniel Gonzalez wrote:
>>>
>>> And (I am discovering some effects during testing):
>>>
>>> 4. Is it possible to call this function *after* the templates have been 
>>> rendered?
>>>
>>> On Thursday, March 14, 2013 11:09:51 AM UTC+1, Daniel Gonzalez wrote:
>>>>
>>>> And:
>>>>
>>>> 3. Is it possible to call this function also for jsonrpc requests?
>>>>
>>>> On Thursday, March 14, 2013 11:00:13 AM UTC+1, Daniel Gonzalez wrote:
>>>>>
>>>>> Hi,
>>>>>
>>>>> For some tests that I am performing, I would like to call a function 
>>>>> "request_processed()" (defined in a model, for example), at the end of 
>>>>> every request.
>>>>>
>>>>> 1. Is this possible? How?
>>>>> 2. Is it possible to also call this function in case no controller was 
>>>>> found to serve the request?
>>>>>
>>>>> Thanks,
>>>>> Daniel
>>>>>
>>>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

Are you suggesting something like this (which is what I am currently doing):

In my model I define a decorator:

def finish_request(fn, *args, **kwargs):
def wrapped():
res = fn(*args, **kwargs)
do_whatever_after_the_request_has_been_processed()
return res
return wrapped

And then in my controller:

@finish_request
@auth.requires_login()
def index():
return "Hello"



On Thursday, March 14, 2013 11:48:48 AM UTC+1, Niphlod wrote:
>
> let's dive it into  I don't think there's an integrated way to do it, 
> rather than decorating your function with something else.
> Please note that as far as I know (but I may be totally wrong) you can't 
> actually return from a function (so it gets rendered and transferred to the 
> client) and continue your computationi.e. if you have return 'abcd' to 
> ship 'abcd' to the client, execution stops there.
> So, assuming that you are fine with, e.g.
> def func():
>  result = 'abcd'
>  .some lengthy "after callback"
>  return result
>
> I'd say it's definitely "doable".
> If I remember correctly, there's only one handy shortcut available without 
> starting toying with decorators response.custom_commit that can be a 
> callable and gets called at the end of every request (before returning the 
> results) 
>
> PS: if the thing to do "after callback" is fine if executed 
> asynchronously, you can set up a queue of things to do to be processed 
> afterwards.
>
> PS2: to kick in after the rendering you must return the compiled template, 
> e.g.
> result = response.render(...)
> ...some lengthy "after callback"
> return result
>
> On Thursday, March 14, 2013 11:35:20 AM UTC+1, Daniel Gonzalez wrote:
>>
>> And (I am discovering some effects during testing):
>>
>> 4. Is it possible to call this function *after* the templates have been 
>> rendered?
>>
>> On Thursday, March 14, 2013 11:09:51 AM UTC+1, Daniel Gonzalez wrote:
>>>
>>> And:
>>>
>>> 3. Is it possible to call this function also for jsonrpc requests?
>>>
>>> On Thursday, March 14, 2013 11:00:13 AM UTC+1, Daniel Gonzalez wrote:
>>>>
>>>> Hi,
>>>>
>>>> For some tests that I am performing, I would like to call a function 
>>>> "request_processed()" (defined in a model, for example), at the end of 
>>>> every request.
>>>>
>>>> 1. Is this possible? How?
>>>> 2. Is it possible to also call this function in case no controller was 
>>>> found to serve the request?
>>>>
>>>> Thanks,
>>>> Daniel
>>>>
>>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

And (I am discovering some effects during testing):

4. Is it possible to call this function *after* the templates have been 
rendered?

On Thursday, March 14, 2013 11:09:51 AM UTC+1, Daniel Gonzalez wrote:
>
> And:
>
> 3. Is it possible to call this function also for jsonrpc requests?
>
> On Thursday, March 14, 2013 11:00:13 AM UTC+1, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> For some tests that I am performing, I would like to call a function 
>> "request_processed()" (defined in a model, for example), at the end of 
>> every request.
>>
>> 1. Is this possible? How?
>> 2. Is it possible to also call this function in case no controller was 
>> found to serve the request?
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Call a function at the end of every request

[Daniel Gonzalez]

And:

3. Is it possible to call this function also for jsonrpc requests?

On Thursday, March 14, 2013 11:00:13 AM UTC+1, Daniel Gonzalez wrote:
>
> Hi,
>
> For some tests that I am performing, I would like to call a function 
> "request_processed()" (defined in a model, for example), at the end of 
> every request.
>
> 1. Is this possible? How?
> 2. Is it possible to also call this function in case no controller was 
> found to serve the request?
>
> Thanks,
> Daniel
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Call a function at the end of every request

[Daniel Gonzalez]

Hi,

For some tests that I am performing, I would like to call a function 
"request_processed()" (defined in a model, for example), at the end of 
every request.

1. Is this possible? How?
2. Is it possible to also call this function in case no controller was 
found to serve the request?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Strange problem importing a module

[Daniel Gonzalez]

Moving to 2.4.2 did not solve the issue. What did solve it was a 
reorganizing of my libraries: previously I was importing libraries from a 
top-level directory, and now I have put all my libraries inside a new 
namespace, so that instead of doing:

from my_module import my_function

Now I am doing:

from top_namespace.my_module import my_funcion

And the issues are gone. I suspect one of my libraries was conflicting with 
one the modules provided by web2py, but have no prove of that.

On Wednesday, March 6, 2013 7:53:26 PM UTC+1, Daniel Gonzalez wrote:
>
> Currently running Version 2.4.1-alpha.2+timestamp.2013.01.13.13.14.47
>
> And I remember that I had a similar issue, which was indeed related to the 
> custom importer:
> https://groups.google.com/d/topic/web2py/IzhnixOivic/discussion
>
> It seems I am run into a similar problem again. If I find any information 
> I'll post it. Otherwise I'll move to 2.4.2 and see if that solves the 
> problems.
>
> On Wednesday, March 6, 2013 7:41:23 PM UTC+1, Massimo Di Pierro wrote:
>>
>> Which web2py version? Some changes were made to 2.4.2 to the custom 
>> importer.
>>
>> On Wednesday, 6 March 2013 12:19:32 UTC-6, Daniel Gonzalez wrote:
>>>
>>> Hi,
>>>
>>> I have this very strange issue: I am importing a module in the web2py 
>>> models, but the module can not be found.
>>> This is only happening in one of my production hosts: the rest are ok 
>>> (very similar machines, similar setup).
>>> If I import the module from a python shell, it works, but importing from 
>>> the web2py shell fails.
>>> I have verified that os.environ.get('PYTHONPATH') are exactly the same 
>>> in both web2py shell and normal python shell.
>>>
>>> sys.path differ slightly: basically, in web2py shell it has the web2py 
>>> installation directory, and in the normal python shell it has the current 
>>> directory.
>>> But this should not affect finding the module I want to import. So, I do 
>>> not understand why this one module is not being found. 
>>>
>>> I have experienced in the past that web2py complains about a module not 
>>> found, when actually what happens is that importing that module causes an 
>>> error (for whatever reason: because an import there does not work, or 
>>> syntax error, or whatever). But I can not tell in this specific case if 
>>> that is my problem. How can I further debug this annoying issue?
>>>
>>> Thanks,
>>> Daniel
>>>
>>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Re: Strange problem importing a module

[Daniel Gonzalez]

Currently running Version 2.4.1-alpha.2+timestamp.2013.01.13.13.14.47

And I remember that I had a similar issue, which was indeed related to the 
custom importer:
https://groups.google.com/d/topic/web2py/IzhnixOivic/discussion

It seems I am run into a similar problem again. If I find any information 
I'll post it. Otherwise I'll move to 2.4.2 and see if that solves the 
problems.

On Wednesday, March 6, 2013 7:41:23 PM UTC+1, Massimo Di Pierro wrote:
>
> Which web2py version? Some changes were made to 2.4.2 to the custom 
> importer.
>
> On Wednesday, 6 March 2013 12:19:32 UTC-6, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I have this very strange issue: I am importing a module in the web2py 
>> models, but the module can not be found.
>> This is only happening in one of my production hosts: the rest are ok 
>> (very similar machines, similar setup).
>> If I import the module from a python shell, it works, but importing from 
>> the web2py shell fails.
>> I have verified that os.environ.get('PYTHONPATH') are exactly the same in 
>> both web2py shell and normal python shell.
>>
>> sys.path differ slightly: basically, in web2py shell it has the web2py 
>> installation directory, and in the normal python shell it has the current 
>> directory.
>> But this should not affect finding the module I want to import. So, I do 
>> not understand why this one module is not being found. 
>>
>> I have experienced in the past that web2py complains about a module not 
>> found, when actually what happens is that importing that module causes an 
>> error (for whatever reason: because an import there does not work, or 
>> syntax error, or whatever). But I can not tell in this specific case if 
>> that is my problem. How can I further debug this annoying issue?
>>
>> Thanks,
>> Daniel
>>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Strange problem importing a module

[Daniel Gonzalez]

Hi,

I have this very strange issue: I am importing a module in the web2py 
models, but the module can not be found.
This is only happening in one of my production hosts: the rest are ok (very 
similar machines, similar setup).
If I import the module from a python shell, it works, but importing from 
the web2py shell fails.
I have verified that os.environ.get('PYTHONPATH') are exactly the same in 
both web2py shell and normal python shell.

sys.path differ slightly: basically, in web2py shell it has the web2py 
installation directory, and in the normal python shell it has the current 
directory.
But this should not affect finding the module I want to import. So, I do 
not understand why this one module is not being found. 

I have experienced in the past that web2py complains about a module not 
found, when actually what happens is that importing that module causes an 
error (for whatever reason: because an import there does not work, or 
syntax error, or whatever). But I can not tell in this specific case if 
that is my problem. How can I further debug this annoying issue?

Thanks,
Daniel

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"web2py-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to web2py+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[web2py] Querying the DAL by regex

[Daniel Gonzalez]

Hi,

I have this function to delete a user with a certain email address:

def delete_user(self, email):
my_query = self.db[self.web2py_user_table].email == email
my_set   = self.db(my_query)
my_set.delete()

Now I would like to implement a function to delete several users at once, 
based on a regex parameter:

def delete_users(self, regex):
my_query = 
my_set   = self.db(my_query)
my_set.delete()

Is it possible to do this?

Thanks,

Daniel

-- 

[web2py] Specify my own controller for @auth.requires_login()

[Daniel Gonzalez]

Hi,

Whenever my login expires, web2py redirects to 
application/default/user/login?_next=.
Since I have a customized login controller, I would like to redirect to it 
whenever the login expires.
Is it possible to "configure" @auth.requires_login() to point to my own 
controller?

Thanks,
Daniel

-- 

[web2py] Re: Triggering imports during web2py startup

[Daniel Gonzalez]

That is true ... per-process. If my apache server has several processes, I 
must make sure that all processes are triggered. This is difficult to 
guarantee.

On Thursday, January 17, 2013 8:50:53 PM UTC+1, Niphlod wrote:
>
> I'm not very well versed in all the intricacies, but if the "layer" calls 
> os.fork() (if you're on linux, that's probably what the underlying server 
> does) your import will happen only on the first request. 
> All the subsequent should not import anything because they will "acquire" 
> the current status of the process when they are forked. 
> There are some options in most of the "layers" (being apache, uwsgi, and 
> so on) to tell the interpreter to completely restart the process after a 
> while, but if you have those issues you'll be better off never forcing a 
> restart of the interpreter.
>
> On Wednesday, January 16, 2013 5:47:18 PM UTC+1, Daniel Gonzalez wrote:
>>
>> That would work, and is what I am planning to do, but: how can I 
>> guarantee that all processes (threads?) have been targetted?.
>> I guess this is not possible, unless I start at least N parallel 
>> requests, N being the number of processes ...
>>
>> On Wednesday, January 16, 2013 5:37:24 PM UTC+1, Niphlod wrote:
>>>
>>> do an "outside" request as soon as you reload apache with curl ?
>>>
>>> Il giorno mercoledì 16 gennaio 2013 12:06:14 UTC+1, Daniel Gonzalez ha 
>>> scritto:
>>>>
>>>> Hi,
>>>>
>>>> In my models I am importing third party libraries. Importing those 
>>>> libraries takes time (3s) which means the first requests (one per process) 
>>>> take long time to be served.
>>>>
>>>> Do you have any suggestion on how to trigger the importing before the 
>>>> first requests arrive? I am running web2py via apache / mod_wsgi.
>>>>
>>>> Thanks,
>>>> Daniel
>>>>
>>>

-- 

[web2py] Re: Problems with auth.login_bare when using postgres

[Daniel Gonzalez]

M, not really: I still need to create the database. Maybe web2py can do 
that for me if I start with a blank postgres database, but I am not 
familiar with that, so this was an easier path for me.

By the way, I found out another problem: we need "serial PRIMARY KEY", so 
the sed script must be changed to include:

s/INTEGER PRIMARY KEY AUTOINCREMENT/serial PRIMARY KEY/

And the postgres sequences must be initialized after import:

select setval(pg_get_serial_sequence('auth_user'   , 'id'), (select max(
id) from auth_user) );
select setval(pg_get_serial_sequence('fulluser', 'id'), (select max(
id) from fulluser) );
select setval(pg_get_serial_sequence('auth_group'  , 'id'), (select max(
id) from auth_group) );
select setval(pg_get_serial_sequence('auth_membership' , 'id'), (select max(
id) from auth_membership) );
select setval(pg_get_serial_sequence('auth_permission' , 'id'), (select max(
id) from auth_permission) );
select setval(pg_get_serial_sequence('auth_event'  , 'id'), (select max(
id) from auth_event) );
select setval(pg_get_serial_sequence('auth_cas', 'id'), (select max(
id) from auth_cas) );



On Thursday, January 17, 2013 8:52:03 PM UTC+1, Niphlod wrote:
>
> yep, probably exporting all to csv and reimporting would have you saved a 
> bit of hassle.
>
> On Wednesday, January 16, 2013 5:43:09 PM UTC+1, Daniel Gonzalez wrote:
>>
>> I found the problem: CHAR fields in postgres (the default that the .dump 
>> command in sqlite is generating) will be extended with spaces to fill the 
>> available length in postgres, so that username/password do not match 
>> anymore.
>>
>> I have modified my import script like this:
>>
>> import_from_sqlite_ ( ) {
>> # Set the following:
>> #  WEB2PY_PG_DB : the name of the database in postgres to import to
>> #  SQLITE_FILE: the location of the sqlite file
>> #  PG_USER  : the postgres user
>> #  PG_PWD   : the potgres password
>> #
>> sql_dump="/tmp/from_sqlite.sql"
>> # Dump the data:
>> #   - PRAGMA and AUTOINCREMENT are not supported in postgres
>> #   - sqlite_sequence is specific to sqlite
>> #   - CHAR will have fixed size in postgres, so it will be filled 
>> with empty spaces. Use VARCHAR instead.
>> sqlite3 $SQLITE_FILE .dump | sed -e 
>> '/^PRAGMA/d;/sqlite_sequence/d;s/AUTOINCREMENT//;s/ 
>> CHAR/ VARCHAR/' > $sql_dump
>> # Drop / recreate database
>> sudo -u postgres dropdb $WEB2PY_PG_DB
>> sudo -u postgres createdb -O $PG_USER -E UTF8 $WEB2PY_PG_DB
>> # Import the data
>> sudo -u postgres PGPASSWORD=$PG_PWD psql -h localhost -U $PG_USER -d 
>> $WEB2PY_PG_DB < $sql_dump
>> # Show some data, to see if the import worked fine
>> sudo -u postgres psql -l
>> echo "\dt" | sudo -u postgres psql -d $WEB2PY_PG_DB
>> echo "select * from auth_user;" | sudo -u postgres psql -d 
>> $WEB2PY_PG_DB
>> rm -f $sql_dump
>> }
>>
>>
>> Be careful with this, since it will drop and recreate the database. Also, 
>> the changes to the sqlite dump are very specific to my situation. YMMV.
>>
>> On Wednesday, January 16, 2013 4:06:23 PM UTC+1, Daniel Gonzalez wrote:
>>>
>>> Hi,
>>>
>>> The following code was working when using SQLite, and now that I have 
>>> moved to postgres it is failing:
>>>
>>> auth.login_bare(request.vars.email, request.vars.password)
>>>
>>> I have the passwords for the time being in plaintext, both in postgres 
>>> and SQLite. I have imported the data to postgres like this:
>>>
>>> # PRAGMA and AUTOINCREMENT are not supported in postgres
>>> # sqlite_sequence is specific to sqlite
>>> sqlite3 storage.sqlite .dump | sed -e 
>>> '/^PRAGMA/d;/sqlite_sequence/d;s/AUTOINCREMENT//' > /tmp/from_sqlite.sql
>>> sudo -u postgres PGPASSWORD=mypass psql -h localhost -U myuser -d 
>>> web2py_wavportal < /tmp/from_sqlite.sql
>>>
>>> I have verified that the data in postgres looks fine:
>>>
>>> web2py_wavportal=# \dt
>>>   List of relations
>>>  Schema |  Name   | Type  |   Owner   
>>> +-+---+---
>>>  public | auth_cas| table | wavportal
>>>  public | auth_event  | table | wavportal
>>>  public | auth_group  | table | wavportal
>>>  public | auth_membership | table | wavportal
>>>  public | auth_permission | table | wavportal
>>>  public | auth_user   | table | wavportal
>>>  public | fulluser| table | wavportal
>>> (7 rows)
>>>
>>> The credentials in auth_user are as expected, but auth.login_bare is 
>>> failing. Why could that be?
>>>
>>> Thanks,
>>> Daniel
>>>
>>

-- 

[web2py] Re: Why does the login expire when doing periodic jsonrpc requests?

[Daniel Gonzalez]

Here: http://code.google.com/p/web2py/issues/detail?id=1287

On Thursday, January 17, 2013 4:50:08 AM UTC+1, Massimo Di Pierro wrote:
>
> Please open a ticket about this... :-)
>
> On Wednesday, 16 January 2013 17:28:59 UTC-6, Daniel Gonzalez wrote:
>>
>> This is the relevant section in Auth.__init__ for the login expiration:
>>
>> if auth and auth.last_visit and auth.last_visit + \
>> datetime.timedelta(days=0, seconds=auth.expiration) >request
>> .now:
>> self.user = auth.user
>> # this is a trick to speed up sessions
>> if (request.now - auth.last_visit).seconds > (auth.expiration 
>> / 10):
>> auth.last_visit = request.now
>> else:
>> self.user = None
>> if session.auth:
>> del session.auth
>>
>> I have verified that jsonrpc requests will update the last_visit. BUT the 
>> login will expire anyway (after "expiration", since the original login), 
>> which I can not understand from the code above.
>>
>> This does not happen when doing normal accesses: doing periodic requests 
>> will keep an active login forever, which is expected I think.
>>
>> Am I missing something here?
>>
>> Thanks,
>> Daniel
>>
>

-- 

[web2py] Re: Possible bug? session.auth is None in some requests

[Daniel Gonzalez]

Yes Anthony, pretty sure. The little hash that you see in the logs *is* the 
cookie. Otherwise I can not grep for the relevant log messages (lots of 
activity going on with 10 parallel requests). This is the code that I have 
at the very top of my db.py:

try:
cookie_value = request.cookies[response.session_id_name].value
except:
cookie_value = None

from gluon.tools import tools_tracer
tools_tracer.setid(cookie_value)

(tools_tracer is part of my instrumentation) And that setid call with 
compute a cookie hash for my logs.

On Thursday, January 17, 2013 5:32:22 AM UTC+1, Anthony wrote:
>
> You're sure the client is sending the session cookie each time?
>
> On Wednesday, January 16, 2013 10:19:01 PM UTC-5, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I am doing stress testing of my web2py application. The way I am doing 
>> this is:
>>
>>- 10 parallel clients
>>- the clients log-in automatically, and obtain the session cookie
>>- the clients log-in only *once* at the start of testing.
>>- I use the session cookie to tag log messages in order to look for 
>>mý problem: this way I can relate the log messages to a specific client.
>>- I am looking for problems in the web2py code, specifically in 
>>tools.py, Auth.__init__, where the auth.user is defined
>>
>> This is my instrumented code (in tools.py, Auth.__init__):
>>
>> if auth is None:
>> tools_tracer.show('session.auth is None')
>> if auth and auth.last_visit and auth.last_visit + \
>> datetime.timedelta(days=0, seconds=auth.expiration) >request
>> .now:
>> tools_tracer.show('session.auth active, last_visit=%s', auth.
>> last_visit)
>> self.user = auth.user
>> # this is a trick to speed up sessions
>> if (request.now - auth.last_visit).seconds > (auth.expiration 
>> / 10):
>> auth.last_visit = request.now
>> else:
>> self.user = None
>> if session.auth:
>> del session.auth
>> tools_tracer.show('session.auth expired')
>>
>> And (in tools.py, Auth.login_user):
>>
>> current.session.auth = Storage(
>> user = user,
>> last_visit=current.request.now,
>> expiration=self.settings.expiration,
>> hmac_key=web2py_uuid())
>> tools_tracer.show('session.auth created')
>>
>> This is my current log:
>>
>> 2013-01-17 04:00:41,816 - Thread-5  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:00:41,861 - Thread-5  - c2d769 session.auth 
>> created
>> 2013-01-17 04:00:41,984 - Thread-6  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:00:41.791030
>> 2013-01-17 04:00:48,923 - Thread-7  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:01:51,131 - Thread-10 - c2d769 session.auth 
>> is None
>> 2013-01-17 04:01:51,214 - Thread-9  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:02:42,506 - Thread-6  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:42,791 - Thread-9  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:04:07.717197
>> 2013-01-17 04:04:42,816 - Thread-4  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:48,862 - Thread-10 - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:49,251 - Thread-6  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:05:35,713 - Thread-7  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:06:39,764 - Thread-8  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:06:10.853567
>> 2013-01-17 04:06:39,803 - Thread-2  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:07:35,154 - Thread-8  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:07:00.261709
>> 2013-01-17 04:07:40,860 - Thread-9  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:07:46,176 - Thread-8  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:07:46,266 - Thread-5  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:08:27,416 - Thread-9  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:08:03.492030
>> 2013-01-17 04:08:27,436 - Thread-2  - c2d769 session.auth 
>> is None
>> 2

[web2py] Re: Possible bug? session.auth is None in some requests

[Daniel Gonzalez]

I was having a problem with session.auth.user, and my investigation has 
narrowed down the problem to web2py code. Instrumenting the code has 
allowed me to see the problem directly in the logs, but the problem is 
there no matter whether I modify Auth code or not.

Sure, for me Sessions and session.auth work well ... except when I am doing 
stress testing. With 10 parallel clients, I start to have problems. Strange 
problems as you see in the logs: around 30% of the requests overall do not 
have a properly initialized auth. Some clients have nearly no problems, 
other clients have lots of problems.

My question is: is nobody else seeing this? Not even when running several 
clients in parallel? I am fairly confident on my client code ...

I am running:
Version 2.4.1-alpha.2+timestamp.2013.01.13.13.14.47
commit 4bbe6ed, 4 days ago, "minor fix in parse_as_rest"

On Thursday, January 17, 2013 4:56:00 AM UTC+1, Massimo Di Pierro wrote:
>
> You are modifying Auth code. What problem are you trying to solve? 
> Sessions and session.auth work well with both stable and trunk version, do 
> they not?
>
> On Wednesday, 16 January 2013 21:19:01 UTC-6, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I am doing stress testing of my web2py application. The way I am doing 
>> this is:
>>
>>- 10 parallel clients
>>- the clients log-in automatically, and obtain the session cookie
>>- the clients log-in only *once* at the start of testing.
>>- I use the session cookie to tag log messages in order to look for 
>>mý problem: this way I can relate the log messages to a specific client.
>>- I am looking for problems in the web2py code, specifically in 
>>tools.py, Auth.__init__, where the auth.user is defined
>>
>> This is my instrumented code (in tools.py, Auth.__init__):
>>
>> if auth is None:
>> tools_tracer.show('session.auth is None')
>> if auth and auth.last_visit and auth.last_visit + \
>> datetime.timedelta(days=0, seconds=auth.expiration) >request
>> .now:
>> tools_tracer.show('session.auth active, last_visit=%s', auth.
>> last_visit)
>> self.user = auth.user
>> # this is a trick to speed up sessions
>> if (request.now - auth.last_visit).seconds > (auth.expiration 
>> / 10):
>> auth.last_visit = request.now
>> else:
>> self.user = None
>> if session.auth:
>> del session.auth
>> tools_tracer.show('session.auth expired')
>>
>> And (in tools.py, Auth.login_user):
>>
>> current.session.auth = Storage(
>> user = user,
>> last_visit=current.request.now,
>> expiration=self.settings.expiration,
>> hmac_key=web2py_uuid())
>> tools_tracer.show('session.auth created')
>>
>> This is my current log:
>>
>> 2013-01-17 04:00:41,816 - Thread-5  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:00:41,861 - Thread-5  - c2d769 session.auth 
>> created
>> 2013-01-17 04:00:41,984 - Thread-6  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:00:41.791030
>> 2013-01-17 04:00:48,923 - Thread-7  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:01:51,131 - Thread-10 - c2d769 session.auth 
>> is None
>> 2013-01-17 04:01:51,214 - Thread-9  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:02:42,506 - Thread-6  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:42,791 - Thread-9  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:04:07.717197
>> 2013-01-17 04:04:42,816 - Thread-4  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:48,862 - Thread-10 - c2d769 session.auth 
>> is None
>> 2013-01-17 04:04:49,251 - Thread-6  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:05:35,713 - Thread-7  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:06:39,764 - Thread-8  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:06:10.853567
>> 2013-01-17 04:06:39,803 - Thread-2  - c2d769 session.auth 
>> is None
>> 2013-01-17 04:07:35,154 - Thread-8  - c2d769 session.auth 
>> active, last_visit=2013-01-17 04:07:00.261709
>> 2013-01-17 04:07:40,860 - Thread-9  - c2d769 session.auth 
&

[web2py] Why does the login expire when doing periodic jsonrpc requests?

[Daniel Gonzalez]

This is the relevant section in Auth.__init__ for the login expiration:

if auth and auth.last_visit and auth.last_visit + \
datetime.timedelta(days=0, seconds=auth.expiration) >request
.now:
self.user = auth.user
# this is a trick to speed up sessions
if (request.now - auth.last_visit).seconds > (auth.expiration / 
10):
auth.last_visit = request.now
else:
self.user = None
if session.auth:
del session.auth

I have verified that jsonrpc requests will update the last_visit. BUT the 
login will expire anyway (after "expiration", since the original login), 
which I can not understand from the code above.

This does not happen when doing normal accesses: doing periodic requests 
will keep an active login forever, which is expected I think.

Am I missing something here?

Thanks,
Daniel

-- 

[web2py] Re: Moving away from the "default" controller

[Daniel Gonzalez]

Have you modified the routes.py file?

On Wednesday, January 16, 2013 11:39:55 PM UTC+1, GeeksRule wrote:
>
> I think I'm doing something silly. For the first time now I'm moving away 
> from the "default" controller.
> I've created a new controller called "fromuser.py", there is a single 
> function in it : "def submitAStory():" which returns a form.
>
> When I hit : http://site.com/myapp/fromuser/submitAStory
> It gives me an error : "invalid function (default/fromuser)"
>
> I think it is still trying to go into the "default" controller.
>
> Note : I have created a view called* fromuser/submitAStory.html*
> *
> *
> Any pointers ?*
> *
>

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

Added: http://code.google.com/p/web2py/issues/detail?id=1285

On Wednesday, January 16, 2013 9:24:19 PM UTC+1, Anthony wrote:
>
> The open issue is still expiration of the data in memcache, in combination 
>> with the auth.settings.expiration. I think this is not well integrated: if 
>> the memcache data expires, that's it. No matter how 
>> big auth.settings.expiration is, expiration of the data in memcache will 
>> force the login to expire.
>
>
> I'm not really familiar with that implementation. Maybe open an 
> issueabout it.
>
> Anthony 
>

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

I see. You are differentiating between login and session, and I guess I 
understand the distinction.
I have verified the service.jsonrpc calls, and they indeed update 
last_visit.

The open issue is still expiration of the data in memcache, in combination 
with the auth.settings.expiration. I think this is not well integrated: if 
the memcache data expires, that's it. No matter how 
big auth.settings.expiration is, expiration of the data in memcache will 
force the login to expire.

On Tuesday, January 15, 2013 3:08:38 PM UTC+1, Anthony wrote:
>
> First, note that sessions do not expire -- web2py will keep the session as 
> long as the client keeps returning the session cookie, so it's up to the 
> client when to terminate the session (typically when the browser is 
> closed). auth.settings.expiration does not determine when the session 
> expires but when the login expires (i.e., after a period of inactivity, the 
> user will have to log in again). Note, in order to avoid excessive session 
> writing, session.auth.last_visit does not actually get updated on every 
> single request -- it only gets updated if the time between the last visit 
> and the current request exceeds 10% of the auth.settings.expiration time 
> (so, about every 6 minutes with a one hour expiration).
>
> As for the @service.jsonrpc requests, are those requests accessing an 
> action that requires authentication, and if so, are you authenticating via 
> a session cookie? If not, then those calls would not affect the session or 
> the Auth login expiration.
>
> Anthony
>
> On Tuesday, January 15, 2013 8:27:04 AM UTC-5, Daniel Gonzalez wrote:
>>
>> FYI:
>>
>> Regarding my last two questions: the expiration is set in auth.settings, 
>> but since I was using the session in memcache and the default_time_expire 
>> in memche is 300 seconds, what I think is happening is that the session in 
>> memcache is expiring, even though the session in auth.settings is not.
>>
>> That means no data can be recovered from memcache for the session, and 
>> thus the session expires.
>>
>> I see several problems here:
>>
>>- the memcache session does not get updated, even though I am doing 
>>accesses (in this case @service.jsonrpc accesses). I mean, this key 
>> should 
>>be refreshed in memcache since I am accessing it (its ttl must be reset). 
>>Just reading the session is not enough: it must be updated too, otherwise 
>>it will expire!
>>- auth.settings.expiration is not useful if the memcache time_expire 
>>is smaller
>>- session.auth.last_visit does not get updated in @service.jsonrpc 
>>accesses
>>
>> Or maybe I am misunderstanding the expiration concept: it could be that 
>> only login actions are supposed to refresh the session.
>>
>> But this looks a bit strange to me: as long as the client is sending 
>> requests, the session should remain active! Whenever an *inactivity* 
>> (meaning, no requests at all) period longer than auth.settings.expiration 
>> has passed, the session should expire. Is this how things should work with 
>> the current implementation?
>>
>>
>> On Tuesday, January 15, 2013 1:26:48 PM UTC+1, Daniel Gonzalez wrote:
>>>
>>> And even worse: I am setting, for testing:
>>>
>>> auth.settings.expiration = 3600 * 24 * 30  # one month
>>>
>>> But the session is expiring after 5 minutes anyway! What is happening 
>>> here?
>>>
>>> On Tuesday, January 15, 2013 12:14:59 PM UTC+1, Daniel Gonzalez wrote:
>>>>
>>>> And a related question: my session is expiring after 5 minutes (300s). 
>>>> I am not setting the expiration time, and the default in 
>>>> Auth.default_settings.expiration is 3600. 
>>>>
>>>> Where is the 300 coming from!?!?
>>>>
>>>> On Tuesday, January 15, 2013 11:57:03 AM UTC+1, Daniel Gonzalez wrote:
>>>>>
>>>>> Thanks Niphlod,
>>>>>
>>>>> I am now using:
>>>>>
>>>>>- session.auth.expiration
>>>>>- session.auth.last_visit
>>>>>
>>>>> And I have realized about one (at least for me) unexpected thing: 
>>>>> accessing @service.jsonrpc controllers does not 
>>>>> reset session.auth.last_visit. Is this intended?
>>>>>
>>>>> On Tuesday, January 15, 2013 11:43:11 AM UTC+1, Niphlod wrote:
>>>>>>
>>>>>> session expiration is managed by auth.settings.expiration ... 
>>>>>>
>>>>>

-- 

[web2py] Re: Triggering imports during web2py startup

[Daniel Gonzalez]

That would work, and is what I am planning to do, but: how can I guarantee 
that all processes (threads?) have been targetted?.
I guess this is not possible, unless I start at least N parallel requests, 
N being the number of processes ...

On Wednesday, January 16, 2013 5:37:24 PM UTC+1, Niphlod wrote:
>
> do an "outside" request as soon as you reload apache with curl ?
>
> Il giorno mercoledì 16 gennaio 2013 12:06:14 UTC+1, Daniel Gonzalez ha 
> scritto:
>>
>> Hi,
>>
>> In my models I am importing third party libraries. Importing those 
>> libraries takes time (3s) which means the first requests (one per process) 
>> take long time to be served.
>>
>> Do you have any suggestion on how to trigger the importing before the 
>> first requests arrive? I am running web2py via apache / mod_wsgi.
>>
>> Thanks,
>> Daniel
>>
>

-- 

[web2py] Re: Problems with auth.login_bare when using postgres

[Daniel Gonzalez]

I found the problem: CHAR fields in postgres (the default that the .dump 
command in sqlite is generating) will be extended with spaces to fill the 
available length in postgres, so that username/password do not match 
anymore.

I have modified my import script like this:

import_from_sqlite_ ( ) {
# Set the following:
#  WEB2PY_PG_DB : the name of the database in postgres to import to
#  SQLITE_FILE: the location of the sqlite file
#  PG_USER  : the postgres user
#  PG_PWD   : the potgres password
#
sql_dump="/tmp/from_sqlite.sql"
# Dump the data:
#   - PRAGMA and AUTOINCREMENT are not supported in postgres
#   - sqlite_sequence is specific to sqlite
#   - CHAR will have fixed size in postgres, so it will be filled with 
empty spaces. Use VARCHAR instead.
sqlite3 $SQLITE_FILE .dump | sed -e 
'/^PRAGMA/d;/sqlite_sequence/d;s/AUTOINCREMENT//;s/ 
CHAR/ VARCHAR/' > $sql_dump
# Drop / recreate database
sudo -u postgres dropdb $WEB2PY_PG_DB
sudo -u postgres createdb -O $PG_USER -E UTF8 $WEB2PY_PG_DB
# Import the data
sudo -u postgres PGPASSWORD=$PG_PWD psql -h localhost -U $PG_USER -d 
$WEB2PY_PG_DB < $sql_dump
# Show some data, to see if the import worked fine
sudo -u postgres psql -l
echo "\dt" | sudo -u postgres psql -d $WEB2PY_PG_DB
echo "select * from auth_user;" | sudo -u postgres psql -d $WEB2PY_PG_DB
rm -f $sql_dump
}


Be careful with this, since it will drop and recreate the database. Also, 
the changes to the sqlite dump are very specific to my situation. YMMV.

On Wednesday, January 16, 2013 4:06:23 PM UTC+1, Daniel Gonzalez wrote:
>
> Hi,
>
> The following code was working when using SQLite, and now that I have 
> moved to postgres it is failing:
>
> auth.login_bare(request.vars.email, request.vars.password)
>
> I have the passwords for the time being in plaintext, both in postgres and 
> SQLite. I have imported the data to postgres like this:
>
> # PRAGMA and AUTOINCREMENT are not supported in postgres
> # sqlite_sequence is specific to sqlite
> sqlite3 storage.sqlite .dump | sed -e 
> '/^PRAGMA/d;/sqlite_sequence/d;s/AUTOINCREMENT//' > /tmp/from_sqlite.sql
> sudo -u postgres PGPASSWORD=mypass psql -h localhost -U myuser -d 
> web2py_wavportal < /tmp/from_sqlite.sql
>
> I have verified that the data in postgres looks fine:
>
> web2py_wavportal=# \dt
>   List of relations
>  Schema |  Name   | Type  |   Owner   
> +-+---+---
>  public | auth_cas| table | wavportal
>  public | auth_event  | table | wavportal
>  public | auth_group  | table | wavportal
>  public | auth_membership | table | wavportal
>  public | auth_permission | table | wavportal
>  public | auth_user   | table | wavportal
>  public | fulluser| table | wavportal
> (7 rows)
>
> The credentials in auth_user are as expected, but auth.login_bare is 
> failing. Why could that be?
>
> Thanks,
> Daniel
>

-- 

[web2py] Problems with auth.login_bare when using postgres

[Daniel Gonzalez]

Hi,

The following code was working when using SQLite, and now that I have moved 
to postgres it is failing:

auth.login_bare(request.vars.email, request.vars.password)

I have the passwords for the time being in plaintext, both in postgres and 
SQLite. I have imported the data to postgres like this:

# PRAGMA and AUTOINCREMENT are not supported in postgres
# sqlite_sequence is specific to sqlite
sqlite3 storage.sqlite .dump | sed -e 
'/^PRAGMA/d;/sqlite_sequence/d;s/AUTOINCREMENT//' > /tmp/from_sqlite.sql
sudo -u postgres PGPASSWORD=mypass psql -h localhost -U myuser -d 
web2py_wavportal < /tmp/from_sqlite.sql

I have verified that the data in postgres looks fine:

web2py_wavportal=# \dt
  List of relations
 Schema |  Name   | Type  |   Owner   
+-+---+---
 public | auth_cas| table | wavportal
 public | auth_event  | table | wavportal
 public | auth_group  | table | wavportal
 public | auth_membership | table | wavportal
 public | auth_permission | table | wavportal
 public | auth_user   | table | wavportal
 public | fulluser| table | wavportal
(7 rows)

The credentials in auth_user are as expected, but auth.login_bare is 
failing. Why could that be?

Thanks,
Daniel

-- 

[web2py] Triggering imports during web2py startup

[Daniel Gonzalez]

Hi,

In my models I am importing third party libraries. Importing those 
libraries takes time (3s) which means the first requests (one per process) 
take long time to be served.

Do you have any suggestion on how to trigger the importing before the first 
requests arrive? I am running web2py via apache / mod_wsgi.

Thanks,
Daniel

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

FYI:

Regarding my last two questions: the expiration is set in auth.settings, 
but since I was using the session in memcache and the default_time_expire 
in memche is 300 seconds, what I think is happening is that the session in 
memcache is expiring, even though the session in auth.settings is not.

That means no data can be recovered from memcache for the session, and thus 
the session expires.

I see several problems here:

   - the memcache session does not get updated, even though I am doing 
   accesses (in this case @service.jsonrpc accesses). I mean, this key should 
   be refreshed in memcache since I am accessing it (its ttl must be reset). 
   Just reading the session is not enough: it must be updated too, otherwise 
   it will expire!
   - auth.settings.expiration is not useful if the memcache time_expire is 
   smaller
   - session.auth.last_visit does not get updated in @service.jsonrpc 
   accesses
   
Or maybe I am misunderstanding the expiration concept: it could be that 
only login actions are supposed to refresh the session.

But this looks a bit strange to me: as long as the client is sending 
requests, the session should remain active! Whenever an *inactivity* 
(meaning, no requests at all) period longer than auth.settings.expiration 
has passed, the session should expire. Is this how things should work with 
the current implementation?


On Tuesday, January 15, 2013 1:26:48 PM UTC+1, Daniel Gonzalez wrote:
>
> And even worse: I am setting, for testing:
>
> auth.settings.expiration = 3600 * 24 * 30  # one month
>
> But the session is expiring after 5 minutes anyway! What is happening here?
>
> On Tuesday, January 15, 2013 12:14:59 PM UTC+1, Daniel Gonzalez wrote:
>>
>> And a related question: my session is expiring after 5 minutes (300s). I 
>> am not setting the expiration time, and the default in 
>> Auth.default_settings.expiration is 3600. 
>>
>> Where is the 300 coming from!?!?
>>
>> On Tuesday, January 15, 2013 11:57:03 AM UTC+1, Daniel Gonzalez wrote:
>>>
>>> Thanks Niphlod,
>>>
>>> I am now using:
>>>
>>>- session.auth.expiration
>>>- session.auth.last_visit
>>>
>>> And I have realized about one (at least for me) unexpected thing: 
>>> accessing @service.jsonrpc controllers does not 
>>> reset session.auth.last_visit. Is this intended?
>>>
>>> On Tuesday, January 15, 2013 11:43:11 AM UTC+1, Niphlod wrote:
>>>>
>>>> session expiration is managed by auth.settings.expiration ... 
>>>>
>>>

-- 

Re: [web2py] Re: Fatal Python error: Couldn't create autoTLSkey mapping

[Daniel Gonzalez]

I have not yet upgraded :(

On Tuesday, January 15, 2013 1:25:06 PM UTC+1, Richard Penman wrote:
>
> Already 2.7.3. 
>
> Did you find the problem fixed after upgrading? 
>
>
> On Tue, Jan 15, 2013 at 11:22 PM, Daniel Gonzalez 
> > 
> wrote: 
> > The links I sent are related to mod_wsgi / apache / python 2.7.2, but it 
> is 
> > a generic python 2.7.2 problem. The problems do not happen always. 
> > 
> > What python version are you using? 
> > 
> > 
> > On Tuesday, January 15, 2013 1:20:16 PM UTC+1, Richard Penman wrote: 
> >> 
> >> I used subprocesses on my previous server with uwsgi and had no 
> problem. 
> >> 
> >> 
> >> On Tue, Jan 15, 2013 at 9:51 PM, Niphlod  wrote: 
> >> > I may be out of line here, but the problem may lie in the fact that 
> if 
> >> > you 
> >> > want to start a subprocess usually subprocess.call uses a thread to 
> >> > communicate with it. uwsgi doesn't allow to spawn your own 
> threads 
> >> > as 
> >> > long as you don't supply the --enable-threads parameter. 
> >> > 
> >> > PS: spawning processes from a web-app poses always a lot of problems 
> >> > (concurrency, pipes, stderrs, inherited socket, and so on). Be sure 
> of 
> >> > what 
> >> > you're doing. 
> >> > 
> >> > Il giorno martedì 15 gennaio 2013 06:30:14 UTC+1, Richard Penman ha 
> >> > scritto: 
> >> >> 
> >> >> hello, 
> >> >> 
> >> >> I set up a uwsgi / nginx server using the 
> >> >> setup-web2py-nginx-uwsgi-ubuntu.sh script. 
> >> >> However I find when the app tries to start a sub-process I get this 
> >> >> error: 
> >> >> 
> >> >> Fatal Python error: Couldn't create autoTLSkey mapping 
> >> >> 
> >> >> I read around and apparently this can happen when lack of memory. 
> >> >> 
> >> >> Do you see any problems with these uwsgi settings: 
> >> >>  
> >> >> /tmp/web2py.socket 
> >> >> /var/www/web2py/ 
> >> >> /=wsgihandler:application 
> >> >>  
> >> >> 4 
> >> >> 60 
> >> >> 8 
> >> >> 1 
> >> >> /tmp/stats.socket 
> >> >> 2000 
> >> >> 512 
> >> >> 256 
> >> >> 192 
> >> >> www-data 
> >> >> www-data 
> >> >> 0 0 -1 -1 -1 python /var/www/web2py/web2py.py -Q -S 
> welcome 
> >> >> -M 
> >> >> -R scripts/sessions2trash.py -A -o 
> >> >>  
> >> >>  
> >> > 
> >> > -- 
> >> > 
> >> > 
> >> > 
> > 
> > -- 
> > 
> > 
> > 
>

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

And even worse: I am setting, for testing:

auth.settings.expiration = 3600 * 24 * 30  # one month

But the session is expiring after 5 minutes anyway! What is happening here?

On Tuesday, January 15, 2013 12:14:59 PM UTC+1, Daniel Gonzalez wrote:
>
> And a related question: my session is expiring after 5 minutes (300s). I 
> am not setting the expiration time, and the default in 
> Auth.default_settings.expiration is 3600. 
>
> Where is the 300 coming from!?!?
>
> On Tuesday, January 15, 2013 11:57:03 AM UTC+1, Daniel Gonzalez wrote:
>>
>> Thanks Niphlod,
>>
>> I am now using:
>>
>>- session.auth.expiration
>>- session.auth.last_visit
>>
>> And I have realized about one (at least for me) unexpected thing: 
>> accessing @service.jsonrpc controllers does not 
>> reset session.auth.last_visit. Is this intended?
>>
>> On Tuesday, January 15, 2013 11:43:11 AM UTC+1, Niphlod wrote:
>>>
>>> session expiration is managed by auth.settings.expiration ... 
>>>
>>

-- 

Re: [web2py] Re: Fatal Python error: Couldn't create autoTLSkey mapping

[Daniel Gonzalez]

The links I sent are related to mod_wsgi / apache / python 2.7.2, but it is 
a generic python 2.7.2 problem. The problems do not happen always.

What python version are you using?

On Tuesday, January 15, 2013 1:20:16 PM UTC+1, Richard Penman wrote:
>
> I used subprocesses on my previous server with uwsgi and had no problem. 
>
>
> On Tue, Jan 15, 2013 at 9:51 PM, Niphlod > 
> wrote: 
> > I may be out of line here, but the problem may lie in the fact that if 
> you 
> > want to start a subprocess usually subprocess.call uses a thread to 
> > communicate with it. uwsgi doesn't allow to spawn your own threads 
> as 
> > long as you don't supply the --enable-threads parameter. 
> > 
> > PS: spawning processes from a web-app poses always a lot of problems 
> > (concurrency, pipes, stderrs, inherited socket, and so on). Be sure of 
> what 
> > you're doing. 
> > 
> > Il giorno martedì 15 gennaio 2013 06:30:14 UTC+1, Richard Penman ha 
> scritto: 
> >> 
> >> hello, 
> >> 
> >> I set up a uwsgi / nginx server using the 
> >> setup-web2py-nginx-uwsgi-ubuntu.sh script. 
> >> However I find when the app tries to start a sub-process I get this 
> error: 
> >> 
> >> Fatal Python error: Couldn't create autoTLSkey mapping 
> >> 
> >> I read around and apparently this can happen when lack of memory. 
> >> 
> >> Do you see any problems with these uwsgi settings: 
> >>  
> >> /tmp/web2py.socket 
> >> /var/www/web2py/ 
> >> /=wsgihandler:application 
> >>  
> >> 4 
> >> 60 
> >> 8 
> >> 1 
> >> /tmp/stats.socket 
> >> 2000 
> >> 512 
> >> 256 
> >> 192 
> >> www-data 
> >> www-data 
> >> 0 0 -1 -1 -1 python /var/www/web2py/web2py.py -Q -S welcome 
> -M 
> >> -R scripts/sessions2trash.py -A -o 
> >>  
> >>  
> > 
> > -- 
> > 
> > 
> > 
>

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

And a related question: my session is expiring after 5 minutes (300s). I am 
not setting the expiration time, and the default in 
Auth.default_settings.expiration is 3600. 

Where is the 300 coming from!?!?

On Tuesday, January 15, 2013 11:57:03 AM UTC+1, Daniel Gonzalez wrote:
>
> Thanks Niphlod,
>
> I am now using:
>
>- session.auth.expiration
>- session.auth.last_visit
>
> And I have realized about one (at least for me) unexpected thing: 
> accessing @service.jsonrpc controllers does not 
> reset session.auth.last_visit. Is this intended?
>
> On Tuesday, January 15, 2013 11:43:11 AM UTC+1, Niphlod wrote:
>>
>> session expiration is managed by auth.settings.expiration ... 
>>
>

-- 

[web2py] Re: Fatal Python error: Couldn't create autoTLSkey mapping

[Daniel Gonzalez]

I am actually researching the same problem. Please take a look here:

https://groups.google.com/d/topic/modwsgi/DW-SlIb07rE/discussion
http://bugs.python.org/issue13156

This is a bug in python 2.7.2, and has been solved in 2.7.3. I am also 
affected, and will have to upgrade all my servers :(

On Tuesday, January 15, 2013 11:51:44 AM UTC+1, Niphlod wrote:
>
> I may be out of line here, but the problem may lie in the fact that if you 
> want to start a subprocess usually subprocess.call uses a thread to 
> communicate with it. uwsgi doesn't allow to spawn your own threads as 
> long as you don't supply the --enable-threads parameter.
>
> PS: spawning processes from a web-app poses always a lot of problems 
> (concurrency, pipes, stderrs, inherited socket, and so on). Be sure of what 
> you're doing.
>
> Il giorno martedì 15 gennaio 2013 06:30:14 UTC+1, Richard Penman ha 
> scritto:
>>
>> hello,
>>
>> I set up a uwsgi / nginx server using 
>> the setup-web2py-nginx-uwsgi-ubuntu.sh script.
>> However I find when the app tries to start a sub-process I get this error:
>>
>> Fatal Python error: Couldn't create autoTLSkey mapping
>>
>> I read around and apparently this can happen when lack of memory.
>>
>> Do you see any problems with these uwsgi settings:
>> 
>> /tmp/web2py.socket
>> /var/www/web2py/
>> /=wsgihandler:application
>> 
>> 4
>> 60
>> 8
>> 1
>> /tmp/stats.socket
>> 2000
>> 512
>> 256
>> 192
>> www-data
>> www-data
>> 0 0 -1 -1 -1 python /var/www/web2py/web2py.py -Q -S welcome -M 
>> -R scripts/sessions2trash.py -A -o
>> 
>> 
>>
>

-- 

[web2py] Re: What is the timeout for the session?

[Daniel Gonzalez]

Thanks Niphlod,

I am now using:

   - session.auth.expiration
   - session.auth.last_visit
   
And I have realized about one (at least for me) unexpected thing: 
accessing @service.jsonrpc controllers does not 
reset session.auth.last_visit. Is this intended?

On Tuesday, January 15, 2013 11:43:11 AM UTC+1, Niphlod wrote:
>
> session expiration is managed by auth.settings.expiration ... 
>

-- 

[web2py] What is the timeout for the session?

[Daniel Gonzalez]

Hi,

In one of my controllers I would to display:

   - what is the timeout for this session
   - how long has the current session still to stay active
   - a list of requests / timestamps (I guess this is not tracked in web2py)
   - when was the last request performed (unfortunately this will be the 
   current request, so it will be "now"?)
   
I have taken a look at the code and I found:

   - Session._last_timestamp

But I am unable to see where the timeout parameter is set for a 
session. Definitely not in Session.connect() ...

Thanks,
Daniel

-- 

Re: [web2py] Set up of db.py takes too long time

[Daniel Gonzalez]

No imports in the code-area profiled. Also, I am testing some requests, but 
the time I am measuring is the time spent in the code-area that I have 
provided. That is, in the define tables code. The controllers are of course 
making use of the table definitions later, but that is not what I am 
measuring.

I have put a "print time.time()" before and after the "define tables" code. 
So no controller is involved in my measurements.

On Tuesday, January 15, 2013 9:56:57 AM UTC+1, rochacbruno wrote:
>
> It will run only when called!
>
> But if he is testing a form, the problem maybe there.
>
> There are some imports?
>

-- 

Re: [web2py] Set up of db.py takes too long time

[Daniel Gonzalez]

Very few records (around 10). I am testing still.

On Tuesday, January 15, 2013 9:23:43 AM UTC+1, rochacbruno wrote:
>
>
> On Tue, Jan 15, 2013 at 6:20 AM, Daniel Gonzalez 
> 
> > wrote:
>
>> IS_NOT_IN_DB(db, '%s.email' % (web2py_user_table))
>
>
> If you too many records on user table, the above code will take a long 
> time, because on every request it will "select email from auth_user"
>
>
>
>

-- 

[web2py] Set up of db.py takes too long time

[Daniel Gonzalez]

Hi,

My db.py takes a long time to run (around 40 ms). This means that just the 
overhead of db.py limits my throughput to 25 req/s (in a single process / 
single thread configuration). This seems to me quite low.

I have tried to take a look at where the time is spent. I have seen the 
following runtimes. These are average values:

 8 ms -> session.connect(request,response,db=MEMDB(cache.memcache))
 2 ms -> db = DAL('sqlite://storage.sqlite', migrate=False)
 5 ms -> auth = Auth(db)
20 ms -> define tables

My define tables is like this:

# Use the authorization table for users
web2py_user_table = auth.settings.table_user_name
# web2py_user_table = 'web2py_user'
db.define_table(
web2py_user_table,
Field('org_id',   type='integer', compute=set_org_id),
Field('email',unique=True),
Field('user_doc_id',  length=128, compute=create_new_user),
Field('password', length=512, compute=automatic_password
, type='password', readable=False, label='Password'), # TODO: use CRYPT
# These fields are needed for authorization
Field('registration_key', length=512, writable=False, readable=
False, default=''),
Field('reset_password_key',   length=512, writable=False, readable=
False, default=''),
Field('registration_id',  length=512, writable=False, readable=
False, default=''),
format = '%(email)s')

# Add constraints to the web2py_user table
web2py_user = db[web2py_user_table]
web2py_user.email.requires= [IS_EMAIL(error_message=auth.
messages.invalid_email), IS_NOT_IN_DB(db, '%s.email' % (web2py_user_table))]
# web2py_user.user_doc_id.requires = 
IS_NOT_EMPTY(error_message=auth.messages.is_empty)
# web2py_user.password.requires = [IS_STRONG(min=5, special=0, 
upper=0), CRYPT()]

auth.define_tables()

Those run times look to me quite high. Maybe some of you can give me hints:

   - Is there anything I can do to speed this up?
   - Why is the define tables taking such a long time, even with 
   migrate=False?
   - Would moving to postgres improve things a lot?
   - Is sqlite the bottleneck here?
   
It is difficult for me to believe that sqlite is the cultprit here: I am 
using a single client in my tests, sending requests sequentially (no 
parallel requests).

Thanks,
Daniel

-- 

[web2py] Handling unicode keys in jsonrpc calls

[Daniel Gonzalez]

Recently web2py changed to parse the jsonrpc parameters in 
tools.py/Service.serve_jsonrpc. Now I have a frawework in my client 
(ember.js) which is sending unicode keys in the jsonrpc parameters. This 
causes the code to fail, like this:

{"version": "1.1", "id": "ID1", "error": {"message": "TypeError: 
access_view() argument after ** must be a mapping, not unicode", "code": 
100, "data": ["  File \"/xxx/web2py/gluon/tools.py\", line 4266, in 
serve_jsonrpc\ns = methods[method](**params)\n"], "name": 
"JSONRPCError"}}

Previously I was "sanitizing" those keys in the controller like this:

def process_json_from_client(**pars):
# I do not know why, but sometimes the keys and values come in unicode
#   I will convert all keys to strings, and some values too
#   (not all, since sometimes unicode is expected)
sanitized = { }
keys_to_convert = ['cdr_doc_id']
for key, value in pars.iteritems():
key = str(key)
if key in keys_to_convert : value = str(value)
sanitized[str(key)] = value
if __debug__ and SHOW_SANITIZED_PARS: log.warning('sanitized -> %s', gd.
pp.pformat(sanitized))
return sanitized

But since now this is done in tools.py/Service.serve_jsonrpc, I can not 
sanitize before calling the jsonrpc controller.

How can unicode keys be accepted in controllers? I think python does not 
allow for unicode parameters, right?

-- 

[web2py] Changes in jsonrpc calls?

[Daniel Gonzalez]

I am moving to a recent web2py version, and I realize that now the json 
parameters are parsed inside tools.serve_jsonrpc. Previously I was parsing 
them in my jsonrpc controller, so that means I have to adapt my 
controllers. Unexpected, but I can manage.

My question is:

Should I do simplejson.dumps in my controller for the reply, or is this now 
also done automatically in tools.serve_jsonrpc? I do not see that.

I am asking because it looks inconsistent that the parameters are 
automatically parsed from json, but the reply is not automatically 
converted to json.

Maybe the reply must not always be json, I am not sure about that.

-- 

[web2py] Re: Strange import problem: is sys.path not used in web2py?

[Daniel Gonzalez]

I have found a workaround (hack) to my problem. I had to modify:

INVALID_MODULES = set(('', 'gluon', 'applications', 'custom_import'))

to

INVALID_MODULES = set(('', 'gluon', 'applications', 'custom_import', 'nltk', 
'collocations', 'numpy', 'testing'))

(there were more modules affected by the same problem). This is ugly, so I 
hope we can find a real fix for this.

It seems the custom_import_install:

def custom_import_install():
if __builtin__.__import__ == NATIVE_IMPORTER:
INVALID_MODULES.update(sys.modules.keys())
__builtin__.__import__ = custom_importer

Is not working as expected?

-- 

[web2py] Re: Strange import problem: is sys.path not used in web2py?

[Daniel Gonzalez]

I have logged the import activity, modifying custom_import like this:

FORMAT="%s -> %-40s %-10s %-10s %-10s %-10s"
# if not relative and not from applications:
if hasattr(current, 'request') \
and level <= 0 \
and not name.split('.')[0] in INVALID_MODULES \
and isinstance(globals, dict):
print FORMAT % ("custom", name, hasattr(current, 'request'), level,name
.split('.')[0] in INVALID_MODULES, isinstance(globals, dict))
...
finally:
if import_tb:
import_tb = None
print FORMAT % ("native", name, hasattr(current, 'request'), level, name
.split('.')[0] in INVALID_MODULES, isinstance(globals, dict))
return NATIVE_IMPORTER(name, globals, locals, fromlist, level)

Here you see some of the data:

native -> random   True   -1 
True   True  
custom -> nltk.dataTrue   -1 
False  True  
native -> __future__   True   -1 
True   True  
native -> os   True   -1 
True   True  
custom -> internalsTrue   -1 
False  True  
native -> subprocess   True   -1 
True   True  
native -> os   True   -1 
True   True  
native -> os.path  True   -1 
True   True  
native -> re   True   -1 
True   True  
native -> warnings True   -1 
True   True  
native -> textwrap True   -1 
True   True  
native -> typesTrue   -1 
True   True  
native -> sys  True   -1 
True   True  
native -> stat True   -1 
True   True   

For nltk the test name.split('.')[0] in INVALID_MODULES is False, but 
should be True. Why could that be?

- I had to type this 7 times! the google groups message composer is not 
very stable :(

On Monday, January 14, 2013 12:38:43 PM UTC+1, Alan Etkin wrote:
>
> The problem is this code in custom_import.py
>
>
> With a very similar configuration but the trunk version I could not 
> reproduce the problem (altough the nltk library has taken about 20 seconds 
> to import in the model)
>
> My system config:
> Python 2.7.3 Ubuntu 12.04 Rocket server
>
> This has something to do with virtualenv surely. BTW, there's a recent 
> issue posted about module headers that might be related.
>
>

-- 

[web2py] Re: Strange import problem: is sys.path not used in web2py?

[Daniel Gonzalez]

The problem is this code in custom_import.py

# if not relative and not from applications:
if hasattr(current, 'request') \
and level <= 0 \
and not name.split('.')[0] in INVALID_MODULES \
and isinstance(globals, dict):

"import nltk.data" goes through, but it shouldn't. Why could that be? (I do 
not understand the code very well)

On Monday, January 14, 2013 11:52:26 AM UTC+1, Daniel Gonzalez wrote:
>
> Version 2.3.2 (2012-12-17 08:59:58) stable, commit 9557c46, in Ubuntu 
> 12.04.1 LTS. I am running from a virtualenv with python 2.7.2.
>
> I have no idea how to force nltk to be found: the sys.path is correct, as 
> I have verified by setting manually in the interpreter and running the 
> import statement. That is why I mean this is a "strange" error: everything 
> seems to be set up correctly, but running from web2py I get an import 
> error, and doing the same thing directly in the interpreter does not.
>
> I am taking a look at the custom_importer: I do no know why this import is 
> falling to line 77 (if I understand correctly, that part is doing the 
> custom web2py modules/models/controllers importing). It should go to line 
> 96:
>
> return NATIVE_IMPORTER(name, globals, locals, fromlist, level)
>
> since nltk.data is a system import. It has nothing to do with web2py.
>
> On Monday, January 14, 2013 11:42:10 AM UTC+1, Alan Etkin wrote:
>>
>> I have a problem with an import in my model:
>>>
>>
>> What version are you running?, in what OS?. As a temporary workaround and 
>> If it's a path error you can try fixing sys.path so nltk is found. 
>>
>

-- 

[web2py] Re: Strange import problem: is sys.path not used in web2py?

[Daniel Gonzalez]

Version 2.3.2 (2012-12-17 08:59:58) stable, commit 9557c46, in Ubuntu 
12.04.1 LTS. I am running from a virtualenv with python 2.7.2.

I have no idea how to force nltk to be found: the sys.path is correct, as I 
have verified by setting manually in the interpreter and running the import 
statement. That is why I mean this is a "strange" error: everything seems 
to be set up correctly, but running from web2py I get an import error, and 
doing the same thing directly in the interpreter does not.

I am taking a look at the custom_importer: I do no know why this import is 
falling to line 77 (if I understand correctly, that part is doing the 
custom web2py modules/models/controllers importing). It should go to line 
96:

return NATIVE_IMPORTER(name, globals, locals, fromlist, level)

since nltk.data is a system import. It has nothing to do with web2py.

On Monday, January 14, 2013 11:42:10 AM UTC+1, Alan Etkin wrote:
>
> I have a problem with an import in my model:
>>
>
> What version are you running?, in what OS?. As a temporary workaround and 
> If it's a path error you can try fixing sys.path so nltk is found. 
>

-- 

[web2py] Strange import problem: is sys.path not used in web2py?

[Daniel Gonzalez]

Hi,

I have a problem with an import in my model:

Traceback (most recent call last):
  File "/installdir/web2py/gluon/restricted.py", line 212, in restricted
exec ccode in environment
  File "/installdir/web2py/applications/wavilon_portal/models/db.py", line 
10, in 
import nltk.data
  File "/installdir/web2py/gluon/custom_import.py", line 77, incustom_importer
raise ImportError, 'Cannot import module %s' % str(e)
ImportError: Cannot import module 'nltk'

I thought this would be a problem with sys.path, so I am displaying it in 
the top of the controller ()
Then I am doing, manually in the python interpreter:

import sys
sys.path=
import nltk.data

But this is giving no error!

What is happening here? Is the import mechanism in web2py using something 
different than sys,path?

Thanks,
Daniel

-- 

[web2py] Re: Putting the session in memcache does not work

[Daniel Gonzalez]

Thanks Niphlod, you hit it right again!

On Sunday, January 13, 2013 1:26:32 PM UTC+1, Niphlod wrote:
>
> I think that you need to restart web2py completely because for 
> performances reasons the "location" of the session (meaning filesystem, 
> cookie, memcache, redis) is evaluated one-time-only  can you try 
> restarting the web2py process (or apache, or uwsgi, depending on your 
> setup...) ?
>
> On Sunday, January 13, 2013 12:31:24 PM UTC+1, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I have 0_memcache.py in my models, with following content:
>>
>> from gluon.contrib.memcache import MemcacheClient
>> memcache_servers = ['127.0.0.1:11211']
>> cache.memcache = MemcacheClient(request, memcache_servers)
>> cache.ram = cache.disk = cache.memcache
>>
>> Now in my db.py model, I have defined (close to the top of the file, 
>> before session is used):
>>
>> from gluon.contrib.memdb import MEMDB
>> session.connect(request,response,db=MEMDB(cache.memcache))
>>
>> After clearing the sessions directory, and logging-in to my application, 
>> I see that the sessions are still being created on the file system. Why is 
>> this, and what can I do to activate memcache for sessions?
>>
>> (Memcache is installed and running fine on the localhost)
>>
>> Thanks,
>> Daniel
>>
>>

-- 

[web2py] Putting the session in memcache does not work

[Daniel Gonzalez]

Hi,

I have 0_memcache.py in my models, with following content:

from gluon.contrib.memcache import MemcacheClient
memcache_servers = ['127.0.0.1:11211']
cache.memcache = MemcacheClient(request, memcache_servers)
cache.ram = cache.disk = cache.memcache

Now in my db.py model, I have defined (close to the top of the file, before 
session is used):

from gluon.contrib.memdb import MEMDB
session.connect(request,response,db=MEMDB(cache.memcache))

After clearing the sessions directory, and logging-in to my application, I 
see that the sessions are still being created on the file system. Why is 
this, and what can I do to activate memcache for sessions?

(Memcache is installed and running fine on the localhost)

Thanks,
Daniel

-- 

[web2py] Make gluon components accessible in mod_wsgi

[Daniel Gonzalez]

Hi,

As part of my library, I have a module which I use to wrap some web2py 
components. Let's say I have simple.py:

from gluon.html import B

def my_b(txt):
return B(txt)

To successfully import this module, my pwd needs to be the directory where 
web2py is installed. Otherwise I get the following error:

>>> import web2pys.simple
Traceback (most recent call last):
  File "", line 1, in 
  File "/my/modules/web2pys/simple.py", line 1, in 
from gluon.html import B
ImportError: No module named gluon.html

This means that my application is running fine whenever I run it with with 
built in Rockett server (python web2py.py), but serving with apache via 
mod_wsgi does not work. I set the python path in mod_wsgi, so that my 
modules can be found:

WSGIDaemonProcess web2py-https user=www-data group=www-data display-name=%{
GROUP} python-path=/my/modules

But no idea how make gluon accessible to mod_wsgi.

How can I "install" web2py in my system, so that gluon components can be 
imported normally, like any other python package?

Thanks,
Daniel

-- 

[web2py] Running some code at the end of the request

[Daniel Gonzalez]

Hi,

I would like to run some code when the request has been processed, but I do 
not want to modify all my controllers for this.
To run something at the beginning of every request, I just need to put my 
code on the models/db.py.
Is it possible to run code at the end of the request processing too?

Thanks,
Daniel

-- 

[web2py] Re: Invalidated RAM cache from outside web2py to handle refreshing of asynchronous data

[Daniel Gonzalez]

Thanks Niphlod. With that I can proceed to implementation.

I am also considering another option: use memcached. This is not integrated 
with web2py (as far as I know), but would maybe simplify my implementation.

Do you think the performance impact of using memcached (which is accessed 
via tcp) compared to using web2py's cache.ram (which lives in the web2py 
memory map) is very big?

On Friday, January 11, 2013 3:06:40 PM UTC+1, Niphlod wrote:
>
> 1. if you use cache.ram you can't access the ram that your web2py process 
> is using from outside that process. 
>In that case you can code a controller that does the trick and call 
> it, e.g., with curl from the commandline.
> 2. same thing, curl from the commandline, or use urllib2, or something 
> that basically does a request to your webserver.
>
> Il giorno venerdì 11 gennaio 2013 14:53:27 UTC+1, Daniel Gonzalez ha 
> scritto:
>>
>> Thanks, that makes sense. In that context, I have two questions:
>>
>>1. How do I call "cache.ram.clear" from outside web2py. What web2py 
>>modules should I include?
>>2. How can I call a controller disguised as a user? I think this 
>>would be the most robust way, if it is possible ...
>>
>>
>> On Friday, January 11, 2013 2:42:37 PM UTC+1, Niphlod wrote:
>>>
>>> Cache is always "related" to a key. In your function, that key is 
>>> request.env.path_info
>>> Let's assume your key is "*the_key*"
>>> If you need to "invalidate" the cache, you should use cache.ram.clear("*
>>> the_key*") or cache.ram("*the_key*", None). From here one, there are 2 
>>> paths ahead of you:
>>> - you have an empty cache and it will be filled at the 1st request 
>>> coming from an user after you called clear()
>>> - you want to fill that key after clear() ASAP
>>>
>>> In the 1st case, after calling clear() you don't have to do nothing.
>>>
>>> In the 2nd case, either you call the controller "disguised as a user" or 
>>> you process the data and store it using cache.ram("*the_key*", thevalue)
>>>
>>> Il giorno venerdì 11 gennaio 2013 13:07:08 UTC+1, Daniel Gonzalez ha 
>>> scritto:
>>>>
>>>> Hi,
>>>>
>>>> I am serving some data which is expensive to compute with 
>>>> a @service.jsonrpc controller. I want to cache the result in RAM, using 
>>>> something like:
>>>>
>>>> @cache(request.env.path_info,time_expire=1000,cache_model=cache.ram)
>>>>
>>>> My data will not change very often, so I want to have a big ttl 
>>>> (actually, I would like a ttl "forever", is this possible?)
>>>>
>>>> But my main problem is this: in my architecture, web2py does not know 
>>>> when new data is available (the data is in a third-party database, 
>>>> completely out of control of web2py). Changes to the data happen 
>>>> asynchronously, according to business processes.
>>>>
>>>> I have three pieces:
>>>>
>>>>1. web2py serving the data to the clients, via jsonrpc
>>>>2. a small script which knows (long-polling) when changes in the 
>>>>third-party database occur
>>>>3. a library to process the information in the third-party 
>>>>database. Calling this must be avoided at all costs, *except* when we 
>>>> have 
>>>>actual changes (which is why I want to add the cache)
>>>>
>>>> The easiest implementation for me would be to be able, from the 
>>>> external script, to invalidate the web2py cache whenever I detect changes 
>>>> relevant to the active sessions (not the whole cache, just the part 
>>>> related 
>>>> to my jsonrpc controller)
>>>>
>>>> Is this possible? Maybe by calling a web2py controller which is in 
>>>> charge of invalidating the cache?
>>>>
>>>> What about the session? I assume the web2py cache is session-related. 
>>>> My script should call the web2py controller using the correct session, but 
>>>> I do not know how to handle this.
>>>>
>>>> Thanks,
>>>> Daniel
>>>>
>>>

-- 

[web2py] Re: Invalidated RAM cache from outside web2py to handle refreshing of asynchronous data

[Daniel Gonzalez]

Thanks, that makes sense. In that context, I have two questions:

   1. How do I call "cache.ram.clear" from outside web2py. What web2py 
   modules should I include?
   2. How can I call a controller disguised as a user? I think this would 
   be the most robust way, if it is possible ...


On Friday, January 11, 2013 2:42:37 PM UTC+1, Niphlod wrote:
>
> Cache is always "related" to a key. In your function, that key is request.
> env.path_info
> Let's assume your key is "*the_key*"
> If you need to "invalidate" the cache, you should use cache.ram.clear("*
> the_key*") or cache.ram("*the_key*", None). From here one, there are 2 
> paths ahead of you:
> - you have an empty cache and it will be filled at the 1st request coming 
> from an user after you called clear()
> - you want to fill that key after clear() ASAP
>
> In the 1st case, after calling clear() you don't have to do nothing.
>
> In the 2nd case, either you call the controller "disguised as a user" or 
> you process the data and store it using cache.ram("*the_key*", thevalue)
>
> Il giorno venerdì 11 gennaio 2013 13:07:08 UTC+1, Daniel Gonzalez ha 
> scritto:
>>
>> Hi,
>>
>> I am serving some data which is expensive to compute with 
>> a @service.jsonrpc controller. I want to cache the result in RAM, using 
>> something like:
>>
>> @cache(request.env.path_info,time_expire=1000,cache_model=cache.ram)
>>
>> My data will not change very often, so I want to have a big ttl 
>> (actually, I would like a ttl "forever", is this possible?)
>>
>> But my main problem is this: in my architecture, web2py does not know 
>> when new data is available (the data is in a third-party database, 
>> completely out of control of web2py). Changes to the data happen 
>> asynchronously, according to business processes.
>>
>> I have three pieces:
>>
>>1. web2py serving the data to the clients, via jsonrpc
>>2. a small script which knows (long-polling) when changes in the 
>>third-party database occur
>>3. a library to process the information in the third-party database. 
>>Calling this must be avoided at all costs, *except* when we have actual 
>>changes (which is why I want to add the cache)
>>
>> The easiest implementation for me would be to be able, from the external 
>> script, to invalidate the web2py cache whenever I detect changes relevant 
>> to the active sessions (not the whole cache, just the part related to my 
>> jsonrpc controller)
>>
>> Is this possible? Maybe by calling a web2py controller which is in charge 
>> of invalidating the cache?
>>
>> What about the session? I assume the web2py cache is session-related. My 
>> script should call the web2py controller using the correct session, but I 
>> do not know how to handle this.
>>
>> Thanks,
>> Daniel
>>
>

-- 

[web2py] Invalidated RAM cache from outside web2py to handle refreshing of asynchronous data

[Daniel Gonzalez]

Hi,

I am serving some data which is expensive to compute with 
a @service.jsonrpc controller. I want to cache the result in RAM, using 
something like:

@cache(request.env.path_info,time_expire=1000,cache_model=cache.ram)

My data will not change very often, so I want to have a big ttl (actually, 
I would like a ttl "forever", is this possible?)

But my main problem is this: in my architecture, web2py does not know when 
new data is available (the data is in a third-party database, completely 
out of control of web2py). Changes to the data happen asynchronously, 
according to business processes.

I have three pieces:

   1. web2py serving the data to the clients, via jsonrpc
   2. a small script which knows (long-polling) when changes in the 
   third-party database occur
   3. a library to process the information in the third-party database. 
   Calling this must be avoided at all costs, *except* when we have actual 
   changes (which is why I want to add the cache)
   
The easiest implementation for me would be to be able, from the external 
script, to invalidate the web2py cache whenever I detect changes relevant 
to the active sessions (not the whole cache, just the part related to my 
jsonrpc controller)

Is this possible? Maybe by calling a web2py controller which is in charge 
of invalidating the cache?

What about the session? I assume the web2py cache is session-related. My 
script should call the web2py controller using the correct session, but I 
do not know how to handle this.

Thanks,
Daniel

-- 

[web2py] Re: Establishin a session with contrib.WebClient

[Daniel Gonzalez]

I found a workaround: first send a login request without cookies to *get* 
the session cookie from web2py, and then do a second login request with the 
cookie, to give web2py a chance to set the session data. This sets up my 
tester with a valid session cookie, which corresponds to valid session data 
in web2py, and my testing can start.

This is a dirty way of handling the redirect (actually, it is *not* 
handling the redirect, which is still handled internally - twice - in 
urrlib2), but since I do not yet know how to modify WebClient to properly 
handle redirects, I will have to settle with this for the time being.

On Monday, December 24, 2012 2:38:48 PM UTC+1, Daniel Gonzalez wrote:
>
> Trying to answer some of my questions.
> The redirects are handled by the client. In the case of WebClient probably 
> by the urrlib2 library.
> Now my question is: how do I tell the urllib2 library used by WebClient to 
> properly setup the cookies when following a redirect?
>
> Thanks,
> Daniel
>

-- 

[web2py] Re: Using contrib.webclient to test JSONRPC interfaces

[Daniel Gonzalez]

Mmm, thanks. I have not taken a look at that yet. I will.
Nevertheless, I like the idea of having an integrated client to handle all 
kind of requests (including jsonrpc), which makes it easier for me for 
statistics and performance gathering.

On Monday, December 24, 2012 6:44:59 AM UTC+1, Massimo Di Pierro wrote:
>
> Hello Daniel,
>
> getting back at this. Why use webclient for this instead of using 
> gluon/contrib/simplejsonrpc.py?  its seems to me it does not belong there.
>
> On Monday, 17 December 2012 14:01:18 UTC-6, Daniel Gonzalez wrote:
>>
>> I have slightly modified the WebClient to support jsonrpc. I am not sure 
>> everything is correctly covered, but it is suiting my needs.
>>
>> In case you are interested, you can see the changes here:
>>
>> https://github.com/gonvaled/web2py/tree/webclient_add_jsonrpc
>>
>> The biggest problem I had was 
>> that opener.addheaders.append((key,str(value))) is not working as (I) 
>> expected: the content-type is not rewritten:
>>
>> http://stackoverflow.com/questions/13920211/not-possible-to-set-content-type-to-application-json-using-urllib2
>>
>> On Monday, December 17, 2012 6:38:30 PM UTC+1, Daniel Gonzalez wrote:
>>>
>>> Hi,
>>>
>>> I am trying to understand how to use the included WebClient to test a 
>>> JSONRPC interface. One showstopper for me at the moment is that I see the 
>>> following in WebClient.post:
>>>
>>> # time the POST request
>>> data = urllib.urlencode(data)
>>> t0 = time.time()
>>> self.response = opener.open(self.url,data)
>>> self.time = time.time()-t0
>>>
>>> When doing JSONRPC accesses I do not want the data to be urlencoded. 
>>> Actually, I will prepare the data with json.dump, and I want WebClient to 
>>> POST it transparently (I guess this goes in the body of the POST request). 
>>> Is this at all possible? Is there an example of WebClient usage 
>>> for JSONRPC interfaces somewhere?
>>>
>>> Thanks,
>>> Daniel Gonzalez
>>>
>>

-- 

[web2py] Re: Establishin a session with contrib.WebClient

[Daniel Gonzalez]

Trying to answer some of my questions.
The redirects are handled by the client. In the case of WebClient probably 
by the urrlib2 library.
Now my question is: how do I tell the urllib2 library used by WebClient to 
properly setup the cookies when following a redirect?

Thanks,
Daniel

-- 

[web2py] Establishin a session with contrib.WebClient

[Daniel Gonzalez]

Hi,

I am trying to test my application with WebClient. My application has a 
non-standard login process, based on a couple of redirects.
When login manually in the browser, everything works fine. Taking a look at 
the requests, I see that after the first redirect a session cookie appears 
in the request.

When doing the same POST request with WebClient, the session cookie does 
not appear after the first redirect.

I wonder if the redirects are handled internally in web2py, or if they 
communicate with the client (browser / WebClient).

Eventually the cookie is sent back in the reply to WebClient, but in 
between redirects the session is not yet set (only when testing with 
WebClient), which makes my test fail because session specific data which is 
setup during the redirect phase can not be configured.

How to handle the redirects with WebClient?

Thanks,

Daniel Gonzalez

-- 

[web2py] Web2py thread management

[Daniel Gonzalez]

Hi,

How are threads managed in web2py? I am stress testing with a multithreaded 
programm, and I see that web2py spawns 14 "worker threads".
These threads are not created/destroyed, but seem to run forever.

python,8698 web2py.py -a 
  |-{python},9005
  |-{python},9023
  |-{python},9042
  |-{python},9043
  |-{python},9044
  |-{python},9050
  |-{python},9051
  |-{python},9053
  |-{python},9054
  |-{python},9055
  |-{python},9057
  |-{python},9058
  |-{python},9059
  `-{python},9060

I have several questions:

   - How do these threads map to the requests? Does each thread serve a 
   request?
   - That means web2py can maximally cope with 14 parallel requests?
   - Is this number (14) configurable?
   - Are the objects defined in the models initialized at thread creation, 
   and then reused for all requests?

Thanks,

Daniel

-- 

[web2py] Re: Using contrib.webclient to test JSONRPC interfaces

[Daniel Gonzalez]

Indeed Massimo, that is the file.

Some comments:

   - I am forcing Content-Type to 'application/json-rpc' in WebClient.post
   - The data must be already json encoded
   - The opener.addheaders is still used, but since I am using 
   a urllib2.Request (to *really* force the Content-Type, which can not be 
   done with addheaders, as you can see in the previously linked StackOverflow 
   question), I would say the headers set with addheaders are not really 
   active.
   - That means self.detault_headers are probably not used
   - for jsonrpc post I have disabled form processing
   - and last but not least, my changes are based on 2.0.2. I assume no big 
   changes have been done for the WebClient since then, but I have not 
   verified.

I am not sure what are the implicactions of all these changes. Maybe some 
corrections are in order.

On Tuesday, December 18, 2012 4:27:26 AM UTC+1, Massimo Di Pierro wrote:
>
>  you point me to the file? Is it this one?
>
>
> https://github.com/gonvaled/web2py/blob/webclient_add_jsonrpc/gluon/contrib/webclient.py
>
>
> On Monday, 17 December 2012 14:01:18 UTC-6, Daniel Gonzalez wrote:
>>
>> I have slightly modified the WebClient to support jsonrpc. I am not sure 
>> everything is correctly covered, but it is suiting my needs.
>>
>> In case you are interested, you can see the changes here:
>>
>> https://github.com/gonvaled/web2py/tree/webclient_add_jsonrpc
>>
>> The biggest problem I had was 
>> that opener.addheaders.append((key,str(value))) is not working as (I) 
>> expected: the content-type is not rewritten:
>>
>> http://stackoverflow.com/questions/13920211/not-possible-to-set-content-type-to-application-json-using-urllib2
>>
>> On Monday, December 17, 2012 6:38:30 PM UTC+1, Daniel Gonzalez wrote:
>>>
>>> Hi,
>>>
>>> I am trying to understand how to use the included WebClient to test a 
>>> JSONRPC interface. One showstopper for me at the moment is that I see the 
>>> following in WebClient.post:
>>>
>>> # time the POST request
>>> data = urllib.urlencode(data)
>>> t0 = time.time()
>>> self.response = opener.open(self.url,data)
>>> self.time = time.time()-t0
>>>
>>> When doing JSONRPC accesses I do not want the data to be urlencoded. 
>>> Actually, I will prepare the data with json.dump, and I want WebClient to 
>>> POST it transparently (I guess this goes in the body of the POST request). 
>>> Is this at all possible? Is there an example of WebClient usage 
>>> for JSONRPC interfaces somewhere?
>>>
>>> Thanks,
>>> Daniel Gonzalez
>>>
>>

-- 

[web2py] Reusing expensive objects

[Daniel Gonzalez]

Hi,

My application uses some objects which are quite slow to setup. I create 
this object in the models file, so that they are available for all 
controllers.

Specifically, most of my processing is delegated to a celery queue (based 
on rabbitmq). The interface to this queue is wrapped by a class 
(MessageQueueClient) which I instantiate in the model file db.py. If I 
understand things correctly, this object will be instantiated for each 
request coming to web2py.

Would it be possible to reuse these objects, so that successive requests do 
not need to create them again? How could this be done?

Thanks,
Daniel

-- 

[web2py] web2py performance, without db

[Daniel Gonzalez]

Hi,

I am trying to get some metrics on my installation. For that I have 
disabled the database (db=None) and I am requesting a very simple 
controller:

def hello1():
return "Hello World"

I have a multi-threaded test program which is sending requests to web2py 
(using WebClient). I have let my test run for several iterations, and I got 
the following metrics:

CLIENTS ROUNDS ELLAPSED (s)TOT REQ   AVG (ms)  REQ/s
  8 72 17.969576 31.197  32.05
  9 25  6.973225 30.991  32.27
  6  4  0.748 24 31.179  32.07
  7 17  3.645119 30.634  32.64
  7 58 12.340406 30.393  32.90
  3 73  6.751219 30.827  32.44
 10 12  3.399120 28.329  35.30
  4 32  3.941128 30.791  32.48
  1 27  0.742 27 27.496  36.37
  8 58 14.156464 30.508  32.78
  4 31  3.574124 28.819  34.70
  8 14  3.274112 29.232  34.21
  9 68 18.722612 30.591  32.69
  9 81 22.070729 30.275  33.03
  4 47  6.031188 32.079  31.17
  7 75 15.491525 29.506  33.89
  1 90  2.495 90 27.719  36.08
  8 38  9.495304 31.234  32.02
  2 60  4.000120 33.335  30.00
  2 81  5.057162 31.214  32.04

(CLIENTS is the number of active threads, ROUNDS is the number of requests 
that each thread performs)

As you see, I am getting quite stable speed of between 30 and 36 requests/s.
This is fine, but 32 req/s is quite low. My system is (this is my 
development system, an HP 635):

   - web2py 2.0.2
   - 2 cores, AMD E-450 
   - 4 GB RAM
   - Linux 2.6.32-5-686
   - CrunchBang Linux statler
   - Python 2.7.2
   - I am using Rocket
   - No apache (or any other frontend). Directly accessing localhost:8000
   - web2py and my test program are sharing the same machine.

I would like to optimize these parameters before moving to production. Is 
there something obvious that I could try to improve these metrics?
Is 30 req/s an expected value for such a simple test? I expected more than 
100 req/s.

Now that I am writing and thinking about this, I am not even sure if the 
limiting factor is web2py or my test program ... How could I know which 
party is "saturating"?

Thanks,

Daniel

-- 

[web2py] Re: Using contrib.webclient to test JSONRPC interfaces

[Daniel Gonzalez]

I have slightly modified the WebClient to support jsonrpc. I am not sure 
everything is correctly covered, but it is suiting my needs.

In case you are interested, you can see the changes here:

https://github.com/gonvaled/web2py/tree/webclient_add_jsonrpc

The biggest problem I had was 
that opener.addheaders.append((key,str(value))) is not working as (I) 
expected: the content-type is not rewritten:
http://stackoverflow.com/questions/13920211/not-possible-to-set-content-type-to-application-json-using-urllib2

On Monday, December 17, 2012 6:38:30 PM UTC+1, Daniel Gonzalez wrote:
>
> Hi,
>
> I am trying to understand how to use the included WebClient to test a 
> JSONRPC interface. One showstopper for me at the moment is that I see the 
> following in WebClient.post:
>
> # time the POST request
> data = urllib.urlencode(data)
> t0 = time.time()
> self.response = opener.open(self.url,data)
> self.time = time.time()-t0
>
> When doing JSONRPC accesses I do not want the data to be urlencoded. 
> Actually, I will prepare the data with json.dump, and I want WebClient to 
> POST it transparently (I guess this goes in the body of the POST request). 
> Is this at all possible? Is there an example of WebClient usage 
> for JSONRPC interfaces somewhere?
>
> Thanks,
> Daniel Gonzalez
>

-- 

Re: [web2py] Specify password strengh, with error feedback

[Daniel Gonzalez]

Thanks, this was very helpful.

On Monday, December 17, 2012 4:05:02 AM UTC+1, Massimo Di Pierro wrote:
>
> Which you can find here:
>
>
> http://web2py.com/examples/static/epydoc/web2py.gluon.validators.IS_STRONG-class.html
>
> On Sunday, 16 December 2012 17:49:51 UTC-6, Jonathan Lundell wrote:
>>
>> On 16 Dec 2012, at 3:47 PM, Daniel Gonzalez  wrote:
>>
>> Is is possible to specify what kind of password is allowed? For example, 
>> let's say I want:
>>
>>- length >=8
>>- at least 1 uppercase
>>- at least 1 lowercase
>>- at least 1 number
>>- at least 1 char in a pre-defined set()
>>
>> I would also like to give the end-user information about why the password 
>> did not pass the cut, like "At least one uppercase is needed".
>>
>> How would this be done in web2py?
>>
>>
>> Have a look at IS_STRONG. 
>>
>

-- 

[web2py] Using contrib.webclient to test JSONRPC interfaces

[Daniel Gonzalez]

Hi,

I am trying to understand how to use the included WebClient to test a 
JSONRPC interface. One showstopper for me at the moment is that I see the 
following in WebClient.post:

# time the POST request
data = urllib.urlencode(data)
t0 = time.time()
self.response = opener.open(self.url,data)
self.time = time.time()-t0

When doing JSONRPC accesses I do not want the data to be urlencoded. 
Actually, I will prepare the data with json.dump, and I want WebClient to 
POST it transparently (I guess this goes in the body of the POST request). 
Is this at all possible? Is there an example of WebClient usage 
for JSONRPC interfaces somewhere?

Thanks,
Daniel Gonzalez

-- 

[web2py] Moving from 2.0.2 to R-2.3.1: undefined with globals in models

[Daniel Gonzalez]

Hi,

My application is using 2.0.2 I have some global variables defined in db.py 
which I am using in my controllers.
Now I have tried to move to the latest web2py version (2.3.1), and those 
globals can not be accessed from the controllers.

Has there been any important change in this regard? Do I need to "import" 
the models in my controllers now, in order to access the global variables 
defined there?

Thanks,
Daniel

-- 

[web2py] Specify password strengh, with error feedback

[Daniel Gonzalez]

Hi,

Is is possible to specify what kind of password is allowed? For example, 
let's say I want:

   - length >=8
   - at least 1 uppercase
   - at least 1 lowercase
   - at least 1 number
   - at least 1 char in a pre-defined set()

I would also like to give the end-user information about why the password 
did not pass the cut, like "At least one uppercase is needed".

How would this be done in web2py?

Thanks,
Daniel

-- 

[web2py] Re: Prevent calling compute when doing an update

[Daniel Gonzalez]

Mmmm ... I see. I actually went another route: I am passing all existing 
parameters to the update_record function, changing only the password. By 
passing the existing parameters, it seems the compute function is not 
called.

On Saturday, December 15, 2012 3:26:41 PM UTC+1, Massimo Di Pierro wrote:
>
> You can set 
>
> db.table.field.compute = lambda: 
>
> only in the actions that need it.
>
> On Saturday, 15 December 2012 04:26:33 UTC-6, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I am implementing a password update feature. I am doing something like 
>> this:
>>
>> def change_pass(email, new_pass):
>> my_query = db[auth.settings.table_user_name].email == email
>> my_set   = db(my_query)
>> my_set.update(password = new_pass)
>>
>> The problem is that this is calling one of my compute definitions in one 
>> of the fields.
>>
>> Field('third_party_entity', compute=
>> big_and_ugly_code_only_intended_for_new_entries)
>>
>> My compute is not intended to be used for updates, but only for new 
>> record creations.
>> I do not want update to call the compute functions. Actually, I do not 
>> want update to do *anything* except updating the field that I 
>> am explicitly giving. Is this possible?
>>
>> Thanks,
>> Daniel Gonzalez
>>
>

-- 

[web2py] Prevent calling compute when doing an update

[Daniel Gonzalez]

Hi,

I am implementing a password update feature. I am doing something like this:

def change_pass(email, new_pass):
my_query = db[auth.settings.table_user_name].email == email
my_set   = db(my_query)
my_set.update(password = new_pass)

The problem is that this is calling one of my compute definitions in one of 
the fields.

Field('third_party_entity', compute=
big_and_ugly_code_only_intended_for_new_entries)

My compute is not intended to be used for updates, but only for new record 
creations.
I do not want update to call the compute functions. Actually, I do not want 
update to do *anything* except updating the field that I 
am explicitly giving. Is this possible?

Thanks,
Daniel Gonzalez

-- 

[web2py] Re: Forcing minus for converted parameter names in form helpers

[Daniel Gonzalez]

Thanks Niphlod, this is working. For the record:

def get_bootstrap_radio(options):
'''
   1
   2
   3
   '''
i = []

for value in options:
i.append(BUTTON(value, _type="button", _class="btn"))
return DIV(*i, _class='btn-group', **{'_data-toggle': "buttons-radio" })



On Tuesday, November 20, 2012 9:50:48 AM UTC+1, Niphlod wrote:
>
> this is beginning to be a FAQ
>
> http://web2py.com/books/default/chapter/29/05?search=data-
>
> On Tuesday, November 20, 2012 9:37:34 AM UTC+1, Daniel Gonzalez wrote:
>>
>> Hi,
>>
>> I am trying to create a radio component in a custom web2py form, using 
>> twitter bootstrap styles, as explained 
>> here<http://twitter.github.com/bootstrap/javascript.html#buttons>
>> .
>> In order for the button group to behave like a radio button, we need to 
>> specify the 'data-toggle="buttons-radio"' modifier.
>> I have tried to do this in web2py with '_data_toggle="buttons-radio"', 
>> but web2py is generating data_toggle instead of data-toggle (underscore 
>> instead of minus).
>> Is there a way to tell web2py to convert the underscores in the parameter 
>> names to minus instead of underscore?
>>
>> This is my code:
>>
>> def get_bootstrap_radio(options):
>> '''
>>1
>>2
>>3
>>'''
>> i = []
>> for value in options:
>>  i.append(BUTTON(value, _type="button", _class="btn"))
>> return DIV(*i, _class='btn-group', _data_toggle="buttons-radio")
>>
>> Thanks,
>> Daniel
>>
>

-- 

[web2py] Forcing minus for converted parameter names in form helpers

[Daniel Gonzalez]

Hi,

I am trying to create a radio component in a custom web2py form, using 
twitter bootstrap styles, as explained 
here
.
In order for the button group to behave like a radio button, we need to 
specify the 'data-toggle="buttons-radio"' modifier.
I have tried to do this in web2py with '_data_toggle="buttons-radio"', but 
web2py is generating data_toggle instead of data-toggle (underscore instead 
of minus).
Is there a way to tell web2py to convert the underscores in the parameter 
names to minus instead of underscore?

This is my code:

def get_bootstrap_radio(options):
'''
   1
   2
   3
   '''
i = []
for value in options:
 i.append(BUTTON(value, _type="button", _class="btn"))
return DIV(*i, _class='btn-group', _data_toggle="buttons-radio")

Thanks,
Daniel

-- 

[web2py] Showing a fixed URL for a web2py application

[Daniel Gonzalez]

Hi,

In my application, I would like the URL being shown to the user to be 
always the same (http://www.myapp.com), no matter what page the user is 
currently browsing.
The idea is that the user should not be confused about the internal 
structure of my application. The links in the page would remain as they 
are, but the browser would show always the same address.

Is this at all possible with web2py? What side-effects could this have?

Thanks,
Daniel

-- 

[web2py] Implementing a self-modifying form in web2py

[Daniel Gonzalez]

Hello,

I am trying to implement a contact form (similar to the one in gmail) with 
web2py. I have problems implementing the following feature: the user should 
be able to expand the form, by adding new entries to the contact (for other 
telephone numbers, or other information). The user is presented at first 
with a simple form, and then a "Add" button should show a new input field. 
This field should be:

   - added *without* forcing a reload of the page
   - it should be processed by web2py when the form is submitted.

I guess this is only possible to be implemented with complicated jQuery 
logic (or similar client-side programming). Or is there a simple way to 
implement this in web2py?

Thanks,
Daniel

-- 

[web2py] Display a list of dictionaries as an html table

[Daniel Gonzalez]

Hello,

I am generating some data on the fly (no database), which I would like to 
display using a controller / view. The data is organized as a list of 
dictionaries, like this:

headers = [ 'header1', 'header2' ]
data = [
  { 'col1' : 'valA', 'col2' : 'valB' },
  { 'col1' : 'valC', 'col2' : 'valD },
]  

How can I code a controller to return an already prepared table for my view?

Thanks,
Daniel

-- 

[web2py] Using PYTHONOPTIMIZE with web2py

[Daniel Gonzalez]

Hi,

I am running web2py via apache wsgi. What is the suggested way to run the 
python interpreter with PYTHONOPTIMIZE?
Are there any known incompatibilities between web2py and PYTHONOPTIMIZE?

Thanks,
Daniel

-- 

[web2py] Re: Submitting a form with a link

[Daniel Gonzalez]

I have found the following method:

A(T('Forgot your password?'), _href='#', _onclick="document['%s'].submit()" 
% (SIGNIN_FORM_ID)

The problem is that I can not specify another href (I *must*  put '#', 
otherwise it does not work).
And, since the form is submitted to the same function as if the normal 
"submit" button was pressed, I am not able to distinguish between both 
situations.


   - How can I submit the form to another url, by using an A element?
   - Or, in case that is not possible, how can I check which way the form 
   has been submitted?
   
Thanks,
Daniel

On Monday, September 10, 2012 10:03:12 AM UTC+2, Daniel Gonzalez wrote:
>
> Hi,
>
> I am preparing a customized log-in form. I want to add a link to recover 
> the password, which is implemented with the A(..) element.
> In the controller function where I want to send an email, I need to access 
> the email given by the user. I do that with request.vars.email.
> The problem is that, since the form was not submitted, the email is empty.
>
> How can I submit the form with an A() element?
> Please note that I *do not* want to have an extra button with "Lost 
> Password" to submit the form. I really want it to be a link.
>
> Is this possible?
>
> Thanks,
> Daniel
>

--