It looks like table inheritance is affected by the same issue:

monster = db.Table(db, 'monster', Field('color'), Field('size'))
db.define_table('thing', Field('name'), monster)


The following traceback is the same as the error using 
_enable_record_versioning() starting at *table = self.lazy_define_table*:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/chanson/py/py4web/apps/thing/__init__.py", line 7, in <module>
    from . models import db
  File "/home/chanson/py/py4web/apps/thing/models.py", line 9, in <module>
    db.define_table('thing', Field('name'), monster)
*  File 
"/home/chanson/.local/share/virtualenvs/py4web-recc_4-8/lib/python3.6/site-packages/pydal/base.py",
 
line 656, in define_table*
*    table = self.lazy_define_table(tablename, *fields, **kwargs)*
*  File 
"/home/chanson/.local/share/virtualenvs/py4web-recc_4-8/lib/python3.6/site-packages/pydal/base.py",
 
line 676, in lazy_define_table*
*    if field.requires is DEFAULT:*
*  File 
"/home/chanson/.local/share/virtualenvs/py4web-recc_4-8/lib/python3.6/site-packages/threadsafevariable/__init__.py",
 
line 16, in __get__*
*    return getattr(self.storage, "%s.%s" % (id(instance), id(self)))*
*  File "src/gevent/local.py", line 408, in 
gevent._local.local.__getattribute__*
AttributeError: 'gevent._local.local' object has no attribute 
'139976984026752.139976984930736'



On Tuesday, February 18, 2020 at 8:07:25 PM UTC-8, Carlos Hanson wrote:
>
> Greetings,
>
> I just started using py4web, and I'm very excited to do so. 
>
> Has there been any more thought on this issue? I was partially into a new 
> web2py app in which I use _enable_record_versioning(), and I am migrating 
> it to py4web.
>
> Thanks.
>
>
> On Saturday, August 17, 2019 at 10:34:21 PM UTC-7, Massimo Di Pierro wrote:
>>
>> Sorry for the late response. I just run into this myself. Do not have a 
>> good solution at this time, unless we make pydal aware of 
>> ThreadSafeVariable(s). Maybe we have to.
>>
>> On Wednesday, 31 July 2019 05:05:50 UTC-7, Manuel Vio wrote:
>>>
>>> Ok, I narrowed the problem a bit:
>>>
>>> In py4web/core.py, around line 207, there is this adjustment in order to 
>>> make some variables threadsafe:
>>>
>>> # make sure some variables in pydal are thread safe
>>> for _ in ['readable', 'writable', 'default', 'update', 'requires']:
>>> setattr(pydal.DAL.Field, _, threadsafevariable.ThreadSafeVariable())
>>>
>>> Turns out that PyDAL's _enable_record_versioning(), in order to create 
>>> the archive table, clones all original fields with their properties. Or, at 
>>> least, it tries to: it seems that copy() fails to correctly transfer 
>>> threadsafevariable informations when dealing with those variables.
>>> I guess that when cloned field's requires attribute is accessed those 
>>> information are not valid anymore and py4web throws the aforementioned 
>>> exception.
>>> I was able to circumvent this error adding an explicit copy of the 
>>> attribute in pydal, but still I'm not sure if this could eventually lead to 
>>> thread problems:
>>>
>>> for field in self:
>>> nfk = same_db or not field.type.startswith('reference')
>>> clone = field.clone(unique=False, type=field.type if nfk else 'bigint')
>>> clone.requires = copy.copy(field.requires)
>>> clones.append(
>>> clone
>>> )
>>>
>>>
>>> Any thougths?
>>>
>>

-- 
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.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/web2py/78646e72-8591-4e79-b572-6309850aac16%40googlegroups.com.

Reply via email to