#django

yo

django binary field are somehow corrupting my

encrypted bytearrays

Is there another field i should be using other than binaryfields to store this data?

the BinaryField is just fine, your code is probably doing something bad

and even if not, you should try to figure out what is wrong instead of switching to something else

I'm having a unicode problem. I'm trying to put all my data in unicode - but there are parts of the django/admin function that seem to be trying to force things to str - and crashing.

rando305: I strongly doubt that

apollo13: i’ve checked teh encrypt and decrypt functions in teh django shell

they work fine

polydaic: aha

just with variables

i’m getting the error because the expected digest is not equal to the actual digest

whatever that means

please see the topic…

When I try to delete an entry in /admin - and it is trying to pull up all the dependencies, it crashes. Even though I can look at the student, edit and delete the unicode stuff.

rando305: please see the topic

rando305: post your code

apollo13: anyways my encryption and decryption functions work great. I’ve got them working fine in other parts of the code

the only thing different here is that i’m storing the keys in the binary fields

and retrieving them

why waste my time instead of doing what the topic asks? :(

polydaic - I switched readers. What is the best way to do a gist? I'm on a ubuntu machine. Thanks

hmm

maybe it is because the binary field has a default length

i've got problem with psycopg2 and django .. when i'm trying to run dbshell postgres ask me for password

rando305:

rando305: gist.github.com

my code is here: https://github.com/rando305/education_operation...

i've no access to pg_hba.conf

rando305: copy paste the problem section to gist.github.com

make a new gist

and then share the link

postgres is on the same host

kushal:

kuter: hmm?

why not

are you on a PaaS?

polydaic: you are using mysql right?

apollo13: yep

surprise surprise

https://gist.github.com/rando305/83287c8cfd1743... - here is the error

polydaic: because it is restricted.

rando305: we need the traceback, not just the error

so its using VARBINARY

hmm

polydaic: that is not what I ment, but if a database silently truncates data it is mysql :þ

which is 8k bytes long

in postgres that would have been an error…

It says ascii codec can't encode '...' - but I didn't ask for ascii anywhere. This is all the django/admin code.

or not cause it supports more than 8k most likely

rando305: …

- ok - I think I know what the problem is. I was trying to convert to Python 3. Went from __unicode__ back to __str__. Is there a way to future proof my code?

(when I actually can go to Python 3)?

yes, read the django docs about writing py2 & py3 compatible code

yes - I had. But I don't understand the def __unicode__() - back to __str__(). The string doesn't work in Python 2

as I said before, show the code in question and the ______full_____ traceback

Well the code in question is in admin. But I'll look at fixing my __str__ declarations to something that won't force ascii.

*sigh*

with code I ment __your__ code, it is not the admin causing the problem even though it triggers it, the problem is still in your code

apollo13: wait yea

the size of my pk went from 3279 to 64

hmm

yes - I posted my github. I expect that it is in the string definitions (I used to have them as __unicode__ ) and didn't switch them back. I'm looking now to see how to do something that will work for 2 and 3

rando305: you posted the error, not the traceback

ok - I've updated the gist

I guess just go back and replace all the __str__ back to __unicode__ and then switch back - the alternative looks like a bit of a hack.

or you just read the django docs on how to support both at the same time?!

I really don't see any hacks there…

if i have a dict in my template, and I want to get the value for a key (which in this case is a string), I can do {{ dict.key }}. What if the key has a space in it?

most likely not without a custom filter

Figured. No problem, thanks

yeah - told me to write a function that behaved differently based on the system configuration. It would require that all my u".." commands be replaced with u(..). It probably wasn't going to fix the __str__ declarations in my models - which was why admin was crashing.

rando305: uhm, wait -- which docs exactly are you reading?

django docs hopefully don't tell you to use u()

But django is still true. Replaced the __str__ declarations back to __unicode__ and everything works again. I'll have to switch it all back when I can migrate to 3.

*sigh*

seriously, you haven't been listening to anything I say

Yes - but I've been working at the same time. Do you have a specific link. I just went to the first one that came up. It probably wasn't the best.

I told you to look at the django docs, how hard can it be to find a link there?

first link for "django python 3" on google https://docs.djangoproject.com/en/1.7/topics/py...

sometimes I really wonder what people search for

ok. I'll look there.

I was thinking it was more a python thing that django. I didn't put django in the search.

I will

ok, for the next time assume that when I say "django docs" I actually mean them for a reason

This was actually the link I used a week or so ago to convert to 3.

Thanks for resending me there. It looks like they have a decorator that fixes things.

yes we do ;)

apollo13: hmm

polydaic: mhm?

it seems like mysql is trucanting

and like

idk bitshifting?

they don’t match up

its removing ilke 10 bytes

so as much as I hate mysql I'd be surprised about shifting

apollo13: how do i prevent this

other than moving to postgres

Sorry apollo13 - I didn't realize that 'they' was 'you'! :-)

at least the truncating part

increase your column's length/size?

ghosticus: its not doing it at a fixed spot

polydaic: dunno, #mysql probably knows how you can configure mysql to error out instead of truncating

polydaic: does it fail at nullbytes?