However, this calls the following, updating all fields:
UPDATE users SET last_login = ‘2006-05-02 13:27:41’, hashed_password
= …
Is there any way I can tell ActionRecord to only update the one field?
I don’t really want the hashed_password (and other) fields sent back to
the database if possible. If there is a reason rails does it like this
that I’m missing, it would be good to hear what.
However, this calls the following, updating all fields:
UPDATE users SET last_login = ‘2006-05-02 13:27:41’, hashed_password
= …
Is there any way I can tell ActionRecord to only update the one field?
I don’t really want the hashed_password (and other) fields sent back to
the database if possible. If there is a reason rails does it like this
that I’m missing, it would be good to hear what.
Is there any way I can tell ActionRecord to only update the one field?
I don’t really want the hashed_password (and other) fields sent back to
the database if possible. If there is a reason rails does it like this
that I’m missing, it would be good to hear what.
I believe I tried this, but the save fails (tested with .save or
logger.info ‘failed :(’). Is there any way to find out why?
I am assuming it fails because there is a validator on the password
attribute, which is an attr_accessor that is only filled during create
or change_password.
authenticated_user.save_without_validation
I would say you definitely don’t want to do this. While I think it’s
a good thing that save_without_validation is there in ActiveRecord, I
don’t think I would ever use it in a production application. In fact,
I might only use it in a Unit Test, but even there, I’m not sure when
I would use it.
The good news is, there are ways to update attributes without needing
to perform a save action on the whole model object. The OP should be
able to just do:
The good news is, there are ways to update attributes without needing
to perform a save action on the whole model object. The OP should be
able to just do:
I am assuming it fails because there is a validator on the password
attribute, which is an attr_accessor that is only filled during create
or change_password.
Except, using update_attributes will still do a SQL Update for the
whole record. From the log of a little test app I have:
Book Update (0.003085) UPDATE books SET “author” = ‘Bill Writer
klj ;lkj’, “title” = ‘new title’ WHERE id = 2
That was after I did in the console:
b = Book.find(2)
b.update_attribute(:title, “new title”)
Hmm… well. I stand (or type corrected. I wasn’t aware that it
did a full object update, simply including the new data. I though it
did just a single column. I suppose I can see why that’s the case,
but the AWDR book certainly leads you to believe it will simply
update that attribute.
That being said, I still don’t like the idea of designing a model
method that deliberately defeats the model’s validation. Just feels
wrong to me, but it definitely looks like that is the way to go, if
that’s what you have a need to do…
you mention that you wouldn’t use save_without_validation and then
propose using update_attribute.
I’m guessing that you’re not aware that update_attribute (the
singular version of the method) does not trigger validation - check
the API docs for more information.
you mention that you wouldn’t use save_without_validation and then
propose using update_attribute.
Yes I did. I said that, because I don’t like the idea of explicitly
telling the entire model to save itself, but skip all the validation
that I’ve added to make sure that a full model object is valid. At
least, not in my production code.
I’m guessing that you’re not aware that update_attribute (the
singular version of the method) does not trigger validation - check
the API docs for more information.
Actually, I did read that, and that’s pretty much what I was trying
to say in my message. However, I managed to confuse “save the whole
model object” with “run validation on the model object when saving”.
As you (and the API docs point out, those aren’t the same thing.
Thanks for clearing up what I managed to get quite confused about.
I suppose it’s a style/philosophical thing, but I have no problem
with the implicit “skipping” of validation that comes from using
update_attribute. That’s a designed feature of that method. But
save_without_validation just gives me a weird feeling. Not sure I can
explain it better than that…
-Brian
This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.
corrected. I wasn’t aware that it