[Rails] Validate in a model using a controller def

I don't know what you mean by "def" but if you want to do conditional
validations have a look at the Rails Guide on validations and it will
show you how.

Colin

What I mean by "def" is the basic CRUD that rails provide

def create
....
end

def update
...
end

def new
...
end

I want that the data validates in 2 cases (create / update)
because when I want to update , for example:

Evaluation 1
- Question 1

If I update question 1 and change the content, and then I save
it shows a error message that says "The number is being used" (and
that shows that
validates_uniqueness_of :qnumber, :scope works)
I want that it only validates it when I'm CREATING a new one, not when
I'm UPDATING

How can I do that?

That is not really a function of the controller action, it is just
whether you are creating a record or updating an existing one. Did
you read the guide as I suggested?

Colin

I'm doing it :)
but I read in another place, that its not safe.. or not recommended to
access a controller from a model
is that true?

Those def were created when I made rails g scaffold

Please don't top post, it makes it difficult to follow the thread.
Insert your reply at appropriate points in previous message. Thanks.

As I said you do not need to access the controller. It is entirely
within the model. The model knows whether it is being asked to create
a new record or to update an existing one. Look at section 4.4 in the
guide (:on)

Colin

Hi Colin. I asked about this before in this list. There are no preferences in this list, so that is where the list stands.

As for top posting, that is an old school habit. Top posting is indeed effective for many people. This message can be clear with historical references down below. It isn't imperative as the text I am adding to this thread, which you are reading now. So, it is secondary to my reply. It IS included below, but it isn't necessary as this message is directed towards yourself. You know what you put before, and it's in your head. If your theory was correct, then every message would have to be quoted in every post. If not, then your email application should have some way of organizing individual messages in order for you to follow the long presentation. A problem with quoting and inserting intermixed is that it gets ugly. A problem with bottom posting is the need to scroll every message that comes into view.

My point is that people have different preferences, and the bottom posting preference only belongs to some. Please don't expect everybody to follow your preference. It is not carved in stone, and can be managed in the preferences of most major email applications for that reason.

----
it seems pretty clear that the logical flow of top to bottom favors bottom posting for anything beyond a short reply that in essence terminates the discussion.

While I am reasonably neutral on the topic myself - the issue really becomes one for those who are taking the time to reply and on a topic that has several replies, the top postings tend to confuse who said what/when. You should keep in mind that those who reply - especially those who are knowledgeable such as Frederick and Colin, I would want to do anything that makes it easy for them to engage/stay engaged but hey, that's just me.

Craig

Actually, it doesn't. The topic is old and argued in very many places.

http://www.google.com/search?q=top+posting+vs+bottom+posting

Arguing how people should post online is futile and simply an expression of someone's preferences. That's it. It can be argued into the ground like it has in the past. Unless a list owner demands it, it's a topic that will never be solved.

My preference is top posting because when I follow a thread, I don't have to start over every time I come back to the thread. While I don't really care myself, it's fine by me that people can post above or below. My mail application can segregate quotations just fine, and I don't have any problem getting past it. I don't even notice if it's top or bottom. I personally don't like having to scroll down several pages to see someone else say "ya I've experienced that result as well". It's a waste of time, and my right hand isn't getting any younger every time I use that wheel.

Don't do it then, I wasn't arguing, just asking nicely.

Colin

See, you failed to read the quotation on the bottom. It was referencing what was quoted down below. Many people don't like this format of interspersing. They find it highly confusing.
People discussing something where contrary opinions are at hand, are in somewhat of a debate. Presentations in such a debate are indeed arguments, and are not meant to convey angry diatribe. That's what I meant.

I'm sure you're trying to be interestingly ironic, but you're shooting
yourself in the foot a little. Yes, of course, it may be "preference"
to "quote" from references at the bottom; but normally in this
instance, it's common to put some form of indicator to the footnote
(such as a number in square-brackets, matched to another at the
reference). Just writing at the top and saying that everything you
write is referencing everything at the bottom is rather glib.
Many people find computer programming highly confusing... I'm not
going to stop encouraging them to get better at it...


On 5 November 2011 21:58, BeeRich wrote:

Because you interspersed this one, it really was *much* easier to follow ;-)

No, it just *is* ignorance, whether I say so out loud or sit quietly
in the wings. Lots of preferences are born out of ignorance; sure,
lots of people get very heated about this one, but if you cancel the
noise on all sides, top/bottom posted replies of any length cause
confusion, while interspersed replies offer an opportunity of
discussing like a "normal" conversation (yet may well still give rise
to some confusion on occasion).

Given the choice of loads of confusion, or little; I'll choose the lesser.
But again; this is contradictory, because you're asking people *not*
to ask other people not to top post? (eek... triple-negative - never
good ;-)


A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

night all.

Well, you're just wrong then. I'm not trying to be ironic at all. I actually don't like hipsters.

And I disagree. I'm responding to what I am quoting. Hopefully what's quotes is what the current reply is commenting about. That's why one would quote something, as a reference. If not, why would it be quoted at all?
For you. Some people don't find it easier. :-P
For your expectations, sure. I don't mind it. THAT is my point.
So now it's a developers issue, and other people who simply do not know how to use email?
It's dead because everybody expects everybody else to wrap around their own needs.
No, I'm saying don't bitch when people do things their way. Please, with all the ways this thread has been quoted and formatted, you still haven't read what I have put. I'm continuing to repeat myself here, and you still don't get it. Would you like a different font? Want me to explain it yet again? If anything, you're demonstrating that in any way shape or form, you still won't get it, and formatting has nothing to do with the effectiveness of a thread or a point.

What would be ultimately respectful to people is if they read the posts instead of expecting others to format properly.

I'm talking about the confusion upon *all* readers totalled together -
not just the confusion of an individual reader.
When lots of people read and reply with top or bottom posted email,
there's lots of confusion. When another lot read and reply with
interleaving, there's generally less confusion.
I'm suggesting that the developers add the functionality of thread
organisation to attempt to compensate for the situation that has
resulted from people not being able to compose messages nicely.
I'm not implying that it's the developers' *fault* there's top-posting
(unless they're the developers of Outlook ;-)
And you don't seem to have "got it" either. No-one has "bitched". It
was suggested to a poster not to top-post, because the people (or at
least a large portion of them) that are attempting to help would find
it easier to help if their messages were composed differently.

So in this event, if the poster continues to abide by their
"preference", they're being deliberately ignorant of the simple
requests of the people that are most likely to help them. A little
courtesy goes a long way.
Now, now. It's all been very civil so far. If you want to veer toward
snide, I'll abstain.
So you agree... they're not formatting "properly" ATM? ;-)

Three types of ways you can format a reply.

- top posting
- bottom posting
- interspersed

Having expectations of everybody else conforming to what you prefer, is going to shut down people's participation.
But that's the situation. You can only assume that with a list where there is nothing formally posted to instruct people, and with a population in a list that's healthy, that there are all three types of posters around.

Interleaving makes sense to you. Other people say top posting makes sense. Other people say bottom posting makes sense. Arguments all around. That's why it's a dead topic, because there is never a winner.
You just insist that your preference is right.
…and then you continue to call them ignorant, because they don't agree with you. And you wonder why topics never progress.

Yes, but I specifically don't understand how "Which is along the same
lines as not expecting people to have people format emails the way you
want them" has any relation to with me saying that I'm not going to
stop helping people improve their coding where possible.
Don't worry though - it's probably not worth the distraction (unless
it was very important, in which case, please do clarify)
I "assume" based on reading all the threads (and I read 'em all...),
and seeing the amount and quality of people who post in different
ways. Most of the requests are poorly written, most of the best
assistance is well written.
Yes... people say all sorts of cr@p about all sorts of stuff. But not
everyone's opinion is worth the same. Look at the quality of posts
from people, and judge their "opinions" by that.
I don't need to consider it - no-one has ever posted a response to a
reply here saying "please don't interleave your response, it's too
easy to see what line you're responding to". Why spend time
considering something which has never happened?

The only "easy" that top or bottom-posting-without-any-trimming is is
easy for the sender - because it's quick and lazy.
I'm calling their behaviour ignorant, and I've said they're
demonstrating ignorance - we can certainly all behave ignorantly from
time-to-time.

It's a good thing there's beer in the World.

I don't understand what you're saying here.
I'm talking about the confusion upon *all* readers totalled together -
not just the confusion of an individual reader.
When lots of people read and reply with top or bottom posted email,
there's lots of confusion. When another lot read and reply with
interleaving, there's generally less confusion. I'm suggesting that
the developers add the functionality of thread organisation to attempt
to compensate for the situation that has resulted from people not
being able to compose messages nicely.
I'm not implying that it's the developers' *fault* there's top-posting
(unless they're the developers of Outlook ;-)And you don't seem to
have "got it" either. No-one has "bitched". It was suggested to a
poster not to top-post, because the people (or at least a large
portion of them) that are attempting to help would find it easier to
help if their messages were composed differently.
So in this event, if the poster continues to abide by their
"preference", they're being deliberately ignorant of the simple
requests of the people that are most likely to help them. A little
courtesy goes a long way.
Now, now. It's all been very civil so far. If you want to veer toward
snide, I'll abstain.
So you agree... they're not formatting "properly" ATM? ;-)

(re-posted as a top-post as an experiment on the clarity of the format...)

And a reply to demonstrate how stupid a bottom post can look.

yes... I've agreed with that already...

Hehe, interspersed can get rather lengthy as well.

My preferences are top-posting and interspersed. I don't mind editing a good topic.

I reiterate: if you change the subject, change the Subject!

In case my subtlety is lost on you, lemme 'splain: if you start
writing about something other than what the conversation was
originally about, take a few seconds to change the Subject line of the
message. (Or if you're writing directly on a web forum, it may be
called the Title, or Re, or something else, but still you probably
understand what I mean.)

Because certain people didn't do that (even though *I did* and thereby
created a perfectly good thread in which to sidetrack the nonsense),
we've now got all this repetitive dead-horse-beating about
top/bottom/interspersed posting, polluting the thread in which
JavierQQ was actually trying to get some help, and *some* of us were
actually trying to help him. Making him, and the rest of us, wade
through it (if at this point anybody but the post-order zealots is
still reading this), is even MORE anti-productive, than whatever
posting-placement you might think is most wrong.

Think of it like a poorly named variable, method, or class. We see
the subject line (go take a moment to look at it!), and expect it to
be something about RoR validations, models, and controller defs,
whatever those are. But instead we find the same old crap people have
been arguing about for literally decades, about top posting. Like if
Rails gives you a stack trace that says that you have an error in your
method Category::rename_to_standard, and you find it has nothing to do
with categories, naming, or standards, but does something utterly
unrelated like maybe calculating a ballistic trajectory (and the whole
application had nothing to do with ballistics), as part of an
Easter-egg game that some overly clever but underly supervised bored
programmer decided to stick into some gem you're using.

Yes I know that by posting this message I am contributing to said
noise -- but as my high school calculus teacher said, sometime you've
got to make something uglier before you can make it pretty.

NOW GET IT THE FSCK OUT OF HERE! :-P

-Dave

+1
Since you have not quoted anyone we don't know who you are referring
to, but I assume it is BeeRich since it was he that changed the
subject.

Though I can't see where you iterated this in the first place in order
to re-iterate it now.

Colin

----
he was actually proving the point of why top posting is a problem while
trying to argue the opposite. He was answering my point which was
sandwiched down below.

Some people will never get it - probably because they don't want to.

Craig

Read the thread.

Read the thread.

I think perhaps I was being too subtle :)
I will find a larger sledgehammer next time.

Colin

You say in another post that you can't be expected to recall
everyone's personal preference for type of reply, yet you seem to
imply here that you have no problem remembering all the contents of
every post in every thread you follow.
Absolutely - those who "bottom post" without snipping are just as
ignorant than those that top-post. Neither group is taking any effort
to contribute to a conversation, they're just blurting out their
thoughts in one paragraph, and slapping 'send'.

Because you say so? Again, check the intensity of the google's hitlist I sent before.
If it's a quick thread, then yes indeed, I do remember the last several comments. Second, like I have said, I regularly quote below and top post. Keeping the thread as well is expected. Why do you think email applications have things like thread organization?
Honestly, I think this is a dead topic. It's never going to go away. I just wanted to say that asking people to post a certain way is a bit much.

I'm sorry if I'm doing something wrong,
I'm new in this group and I don't want to bother anyone,
I'm just hitting the reply button and that's all.

Many users, particularly on technical lists, prefer that a reply is
inserted at appropriate points into the previous message, so that the
full post reads like a question and answer sequence. This requires
you to hit reply then scroll through the previous message inserting
bits of text at the appropriate points. Whilst doing this you may
remove any redundant bits of the previous message. It is not a rule
(at least on this list) but many prefer it that way.

Colin

I don't expect everybody to follow my preferences, but if I am
spending my time helping someone then I feel at liberty to ask them to
post in a manner that makes it easier for me to follow what is going
on.

Colin

OK…but then you would also expect other people to expect you to keep a log of how people's preferences should be, with you responding accordingly? That's unreasonable.

validations are fundamentally ignorant of what it is that has caused
save to be called (i.e. it doesn't know what action, controller etc.
There might not even be a controller if the code was being run from a
background task).

The validation methods do take an on option (:on => :create or :on
=> :update) but you should understand that this refers to whether
active record is create or updating an object and has nothing to do
with what controller action was involved.

However, validates_uniqueness_of should already be adding a condition
to the query so that when updating an existing object it doesn't find
itself, so I'm not sure why you're running into this

Fred

Because as an example
I have
Evaluation 1
- Question 1
- Question 2

If I edit question 2 and change the number it will allow me to do that
and it will show this
Evaluation 1
- Question 1
- Question 1

And that's what I want to avoid

I tried validates_uniqueness_of :qnumber, :scope
=> :evaluation_id, :on => :create
and it works, but when I edit, I can do what I mention

re-ordering for top -> bottom
----
what is the field/column that is 'Question 1'? Is it one of the fields that you only validate on 'create'? You should probably remove ', :on => :create' because that won't protect uniqueness on 'update'

Craig

I'm showing evaluation as it shows on my view

EVALUATION

Question Content Created by
Number

1 (..........)
(..........) Edit
2 (..........)
(..........) Edit
3 (..........)
(..........) Edit

In my form question

Question number (_______________________________)
Content (_______________________________)
Created by (_______________________________)

Save

And for example what If I edit "3" and I change it to "2", it
shouldn't let someone do that

On Nov 4, 2011, at 10:20 AM, JavierQQ wrote:

still not entirely clear but I would think that you need to change...
to this...

validates_uniqueness_of :qnumber

I'm doing that because I can create several evaluations and each
evaluation has an id, that's why I'm scoping by evaluation_id
I found in section 15 of the validation guide, something about
Proc.new
What I want to do with update is that when I hit save it first search
if the number exists(with a where() I guess), if exist it shouldn't
allow me to update
if not, it will let me update the question
Do you suggest to use Proc.new?

Thanks

You are still being too vague about your requirement, for me anyway.
Can you write the validation test as pseudo code, being absolutely
clear about each reference there? Ignore everything about the meaning
of the database fields and just write the validation requirement in
terms of data in the database and the new data about to be saved,
referencing just database fields.

Colin

On Fri, Nov 4, 2011 at 16:27, Colin Law wrote
to JavierQQ <jquarites@gmail.com>:
I think I grok what he wants and can express it briefly: the
*combination* of evaluation_id and qnumber, must be unique.

If that's right, then the Rails Guide on validation (and callbacks)
has a close example:

validates :name, :uniqueness => { :scope => :year,
:message => "should happen once per year" }

but the OP says that:

validates_uniqueness_of :qnumber, :scope => :evaluation_id

isn't working, and IIUC should be equivalent to

validates :qnumber, :uniqueness => { :scope => :evaluation_id }

Just in case I don't UC:

JavierQQ, try this:

validates :qnumber, :uniqueness => { :scope => :evaluation_id,
:message => "must be unique within each evaluation" }

in the model, and let us know what happens. Don't worry about what
controller it's going through to get there.

-Dave

PS: As for top vs. bottom posting, another point of list etiquette
that some people have been violating is that if you change the
subject, change the Subject. :-P

Thanks for your reply and sorry for my late response =)

I'm using Typus, I know that theres a group for typus but I think that
this problem can happen everywhere
I have a view of evaluations

EVALUATIONS

Name View questions Add question
Math view add (edit)
Science view add (edit)

EVALUATION : Math
QUESTIONS

Number View alternatives Add alt.
1 view add (edit)
2 view add (edit)
3 view add (edit)

with
validates_uniqueness_of :qnumber, :scope => :evaluation_id

I made that there's only one question number "1" on each evaluation
the only problem that I have (as I mentioned before) is that, when I
edit the second question (# 2)
and I change its content, it doesn't let me to do it, because they
found that the number (# 2) has been already taken

The other problem is what happens if I want to edit
the second question (# 2) and I change the number to (# 3)
with that validation it let me do that, but what if I edit the second
question (#2)
and I put (#1).... there will be two questions with the same number
(#1)

I was searching for a validation for it, I've been reading the
validation guide and found section 15 about
Proc.new.... it could help but I don't understand it well (I'm really
new with rails)
It works! ... well as far as I tested :), thanks a lot

I've been reading the messages about top-posting and I'm really really
sorry about any trouble I could have caused
Thanks a lot Colin for keep on trying to help me, I'll follow your
advice about how to post something

Don't worry about it, it was my fault for asking you. Some people get
very excited about this sort of thing.
I think Dave Aronson's suggestions were more use than mine :)

Colin