Inserting multiple records from one table (model) to another

[John Mcleod]

Hello all,
I've been getter better with Rails but I'm still just learning.

Here's what I have.
I have one table (imports) and I read in a csv into this table.
After doing some preliminary editing to each record I wish to insert all
records into another table (projects) in one shot.
I could create a new method in the projects controller and then call the
"create", but doesn't that only insert one record?

I need a little clarity on this.

I know how to delete all records from a table. I wonder if it's as
simple?

Thank you for any help.

JohnM
--
Posted via http://www.ruby-forum.com/.

[Marnen Laibow-Koser]

John Mcleod wrote:
> Hello all,
> I've been getter better with Rails but I'm still just learning.
>
> Here's what I have.
> I have one table (imports) and I read in a csv into this table.
> After doing some preliminary editing to each record I wish to insert all
> records into another table (projects) in one shot.
> I could create a new method in the projects controller and then call the
> "create", but doesn't that only insert one record?

Yes. In this case, if the tables have the same structure, you might
want to use SQL's INSERT SELECT syntax (see your DB for details).

But having two tables with identical structure is smelly. Why not just
use a flag and a named_scope?

>
> I need a little clarity on this.
>
> I know how to delete all records from a table. I wonder if it's as
> simple?

Rails doesn't abstract this AFAIK, in part because this is very rarely
needed.

>
> Thank you for any help.
>
> JohnM

Best,
--
Marnen Laibow-Koser
http://www.marnen.org
mar...@marnen.org

[John Mcleod]

Thanks for the quick reply.

>But having two tables with identical structure is smelly. Why not just
> use a flag and a named_scope?

I'm not sure what you mean by this.

Yes, both tables have the identical structure but my idea is too delete
all data from the "import" table after the insert into the "projects"
table. This way the user has a fresh table to get info from a csv file.

John

[Marnen Laibow-Koser]

John Mcleod wrote:
> Thanks for the quick reply.
>
>>But having two tables with identical structure is smelly. Why not just
>> use a flag and a named_scope?
> I'm not sure what you mean by this.

Then go look up named_scope -- or, in this case, default_scope.

>
> Yes, both tables have the identical structure but my idea is too delete
> all data from the "import" table after the insert into the "projects"
> table. This way the user has a fresh table to get info from a csv file.

There's no reason to have a fresh table. Again, just flag records as
appropriate.

[John Mcleod]

Your right.
I could not deal with the additional table.

Thanks for the insight.

John

[Marnen Laibow-Koser]

John Mcleod wrote:
> Your right.
> I could not deal with the additional table.

You could -- it's just poor design in most cases.

>
> Thanks for the insight.

You're welcome.

>
> John

[Ar Chron]

Marnen Laibow-Koser wrote:
>
> You could -- it's just poor design in most cases.
>

Most cases yes, but in some cases a parallel import table is not a bad
thing.

If you put all your data into one table, then all the records going into
that table have to pass all the model constraints immediately.

From John's initial post, I read "after doing some preliminary editing"
to imply that the records, as imported from CSV, may not satisfy all the
constraints of the application.

It may be in his interest to keep these "dirty" records separate from
the "clean" records so he does not have to relax any constraints on the
mainstream application data, or complicate his existing model
constraints by mixing a flag check into the middle of things.

Ultimately, it'll be whatever works best for John.

[Marnen Laibow-Koser]

Ar Chron wrote:
> Marnen Laibow-Koser wrote:
>>
>> You could -- it's just poor design in most cases.
>>
>
> Most cases yes, but in some cases a parallel import table is not a bad
> thing.
>
> If you put all your data into one table, then all the records going into
> that table have to pass all the model constraints immediately.

That's true, of course. But in that case, the tables don't have the
same structure, so the smell doesn't apply.

>
> From John's initial post, I read "after doing some preliminary editing"
> to imply that the records, as imported from CSV, may not satisfy all the
> constraints of the application.
>
> It may be in his interest to keep these "dirty" records separate from
> the "clean" records so he does not have to relax any constraints on the
> mainstream application data,

You may be right. I hadn't thought of that from the initial
description.

> or complicate his existing model
> constraints by mixing a flag check into the middle of things.

default_scope is not complicated!

>
> Ultimately, it'll be whatever works best for John.

Best,

--
Marnen Laibow-Koser
http://www.marnen.org
mar...@marnen.org

[John Mcleod]

You are correct.
I must do some editing to incoming data before it's "imported" into the
"projects" table.

I'm having more problems importing data from one table to another. When
editing is complete, I click a button to "save" to projects table.
My button is very up front and easy.
<%= button_to "Add Records to Projects", :action => "addIrbProjects",
:id => import.id %>

It's just setting up the projects Controller that has me confused.
I'm in the projects controller reading a "imports" table.

John

[John Mcleod]

Hello all,
I've been getter better with Rails but I'm still just learning.

Here's what I have.
I have one table (imports) and I read in a csv into this table.

I must do some preliminary editing to each record, afterwhich, I wish to

insert all records into another table (projects) in one shot.

I've created a button below the table

<%= button_to "Add Records to Projects", :action => 'add_new_projects'
%>

I need a little clarity on this.

I know how to delete all records from a table. I wonder if it's as
simple?

Thank you for any help.

JohnM

[Marnen Laibow-Koser]

You already asked this at http://www.ruby-forum.com/topic/198008 /
http://groups.google.com/group/rubyonrails-talk/browse_thread/thread/7f09e32544d1b4f1/ede1159a796d9e64
, and got some answers. Let's keep further discussion to that thread.

Best,
--
Marnen Laibow-Koser
http://www.marnen.org
mar...@marnen.org

--
Posted via http://www.ruby-forum.com/.

[Jillian Galloway]

I think you have to loop through the list saving one record each time, but I could be wrong. If anyone's got a better way I'd love to know what it is. :)

[John Mcleod]

Thank you for the reply.
I've been struggling with this for some time.

Recently, I have this...

- controller -
def create_csv_projects
@imports = Import.all
@projects = Project.all

@imports.each { |import|
if @projects.create!(params)
flash['notice'] = 'CSV projects have been successfully created
in Projects Database.'
redirect_to :controller => 'projects', :action => 'index'
end
}
end

Unfortunately, it not working.

John

Jillian Galloway wrote:
> I think you have to loop through the list saving one record each time,
> but I
> could be wrong. If anyone's got a better way I'd love to know what it
> is. :)
>
>
>
> On Tue, Oct 27, 2009 at 2:37 PM, John Mcleod <

--
Posted via http://www.ruby-forum.com/.

[Marnen Laibow-Koser]

Jillian Galloway wrote:
> I think you have to loop through the list saving one record each time,
> but I
> could be wrong.

And so you are. It is never appropriate to put an SQL query inside a
loop.

> If anyone's got a better way I'd love to know what it
> is. :)
>

Use something like ar-extensions if possible. If not, then build an SQL
query string directly.

>
>
> On Tue, Oct 27, 2009 at 2:37 PM, John Mcleod <

Best,

--
Marnen Laibow-Koser
http://www.marnen.org
mar...@marnen.org