Just as a quick check, I replaced the new parse_datetime function with the
old, and all is well.
Brad
On Friday, August 31, 2012 3:49:32 PM UTC-5, Brad Miller wrote:
>
> Here is the value that printed:
>
> 2012-07-05 02:36:50
>
> This field is declared in the model as: Field('timestamp','datetime'),
>
> The old parse_datetime function seems to skip over the date part before it
> does the split on the -
>
> Here's the old one:
>
> def parse_datetime(self, value, field_type):
> if not isinstance(value, datetime.datetime):
> (y, m, d) = map(int,str(value)[:10].strip().split('-'))
> time_items = map(int,str(value)[11:19].strip().split(':')[:3])
> if len(time_items) == 3:
> (h, mi, s) = time_items
> else:
> (h, mi, s) = time_items + [0]
> value = datetime.datetime(y, m, d, h, mi, s)
> return value
>
> Brad
>
> On Friday, August 31, 2012 3:33:22 PM UTC-5, Massimo Di Pierro wrote:
>>
>> This is the code that causes the problem to you:
>>
>> def parse_datetime(self, value, field_type):
>> if not isinstance(value, datetime.datetime):
>> ....
>> elif '-' in value:
>> value,tz = value.split('-')
>>
>> This function is called when parsing data retrieved from database. The
>> reason it is failing is that data is corrupted (contains more than one dash
>> and that is not allowed by ISO standard). What database (my guess is
>> sqlite, which allows you to change a string field into a datetime while
>> leaving data in there). Can you add a print statement and see what the data
>> looks like:
>>
>> def parse_datetime(self, value, field_type):
>> if not isinstance(value, datetime.datetime):
>> print value
>> ....
>>
>> Let me know it may help understand how the data got in there.
>>
>> On Friday, 31 August 2012 14:53:16 UTC-5, Brad Miller wrote:
>>>
>>> I just upgraded my app to the latest 2.0.3 (stable) and while checking
>>> some of my pages got the following error:
>>>
>>> Traceback (most recent call last):
>>> File "/Users/bmiller/Beta/web2py/gluon/restricted.py", line 209, in
>>> restricted
>>> exec ccode in environment
>>> File
>>> "/Users/bmiller/Beta/web2py/applications/runestone/controllers/admin.py",
>>> line 160, in <module>
>>> File "/Users/bmiller/Beta/web2py/gluon/globals.py", line 185, in
>>> <lambda>
>>> self._caller = lambda f: f()
>>> File "/Users/bmiller/Beta/web2py/gluon/tools.py", line 2780, in f
>>> return action(*a, **b)
>>> File
>>> "/Users/bmiller/Beta/web2py/applications/runestone/controllers/admin.py",
>>> line 116, in studentactivity
>>> db.useinfo.sid, count, last, groupby=db.useinfo.sid, orderby=count)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 8692, in select
>>> return adapter.select(self.query,fields,attributes)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 2029, in select
>>> return super(SQLiteAdapter, self).select(query, fields, attributes)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 1529, in select
>>> return self._select_aux(sql,fields,attributes)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 1510, in
>>> _select_aux
>>> return processor(rows,fields,self._colnames,cacheable=cacheable)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 1886, in parse
>>> fields[j].type,blob_decode)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 1718, in
>>> parse_value
>>> return self.parsemap[key](value,field_type)
>>> File "/Users/bmiller/Beta/web2py/gluon/dal.py", line 1753, in
>>> parse_datetime
>>> value,tz = value.split('-')
>>> ValueError: too many values to unpack
>>>
>>> This is the result of this this query code, which continues to work
>>> under 1.99.7
>>>
>>> count = db.useinfo.id.count()
>>> last = db.useinfo.timestamp.max()
>>> res = db(db.useinfo.course_id==course.course_id).select(
>>> db.useinfo.sid, count, last, groupby=db.useinfo.sid,
>>> orderby=count)
>>>
>>> I'm assuming this is some kind of regression since the same code is
>>> working in 1.99.7, but maybe I'm doing something that I could do but no
>>> longer should??
>>>
>>> Brad
>>>
>>>
--
