I am trying to find the best way to include a referenced model on what is essentially a compound key.
I have ChecklistItem (a list of things to do daily) and then ChecklistChecks (which ties the ChecklistItem together with a User for a particular day. These checklists can either be for all Stores (with a null store_id) or for a particular Store.
This pulls all of the ChecklistItems and their associated checks:
ChecklistItem.includes(:checklist_checks).where(store_id: [nil,#store.id], list_type: 'open')
The problem is that there would be checks from multiple days in there. What I need is to pull all of the ChecklistItems and any checks from a specific day.
I tried adding conditions like this:
ChecklistItem.includes(:checklist_checks).where(store_id: [nil,#store.id], list_type: 'open', checklist_checks: {store_id: #store.id, report_date: #today})
The problem is that will only pull ChecklistItems that have an associated ChecklistCheck.
It is generating SQL that is essentially:
SELECT
checklist_items.*,
checklist_checks.*
FROM
checklist_items
LEFT OUTER JOIN
checklist_checks
ON
checklist_checks.checklist_item_id = checklist_items.id
WHERE
checklist_items.list_type = 'open'
AND
checklist_checks.store_id = 1
AND
checklist_checks.report_date = '2015-05-03'
AND
(checklist_items.store_id = 1 OR checklist_items.store_id IS NULL)
I think the problem is that the conditions on checklist_checks is in the WHERE clause. If I could move them to the ON clause of the join, everything would work.
Is there a Rails way to end up with something like this?
SELECT
checklist_items.*,
checklist_checks.*
FROM
checklist_items
LEFT OUTER JOIN
checklist_checks
ON
checklist_checks.checklist_item_id = checklist_items.id
AND
checklist_checks.store_id = 1
AND
checklist_checks.report_date = '2015-05-03'
WHERE
checklist_items.list_type = 'open'
AND
(checklist_items.store_id = 1 OR checklist_items.store_id IS NULL)
UPDATE:
I found this: enter link description here
It suggests using find_by_sql and then passing the result array and model to be included to ActiveRecord::Associations::Preloader.new.preload
I tried that, and my find_by_sql pulls the right stuff, but the id column is nil in the resulting objects.
#store = Store.find(1)
#today = Date.today - 1.days
#open_items = ChecklistItem.find_by_sql(["SELECT checklist_items.*, checklist_checks.* FROM checklist_items LEFT OUTER JOIN checklist_checks ON checklist_checks.checklist_item_id = checklist_items.id AND checklist_checks.store_id = ? AND checklist_checks.report_date = ? WHERE checklist_items.list_type='open' AND (checklist_items.store_id=? OR checklist_items.store_ID IS NULL)", #store.id, #today, #store_id])
ActiveRecord::Associations::Preloader.new.preload(#open_items, :checklist_checks)
> #open_items.first.name
=> "Turn on the lights"
> #open_items.first.id
=> nil
A solution using Arel to generate a custom join clause:
class ChecklistItem < ActiveRecord::Base
has_many :checklist_checks
# ...
def self.superjoin(date, store_id)
# build the ON clause for the join
on = Arel::Nodes::On.new(
Arel::Nodes::Equality.new(ChecklistChecks.arel_table[:checklist_item_id], ChecklistItem.arel_table[:id]).\
and(ChecklistItem.arel_table[:store_id].eq(1)).\
and(ChecklistChecks.arel_table[:report_date].eq(date))
)
joins(Arel::Nodes::OuterJoin.new(ChecklistChecks.arel_table, on))
.where(store_id: [nil, store_id], list_type: 'open' )
end
end
I bundled it up into a model method to make it easier to test in the rails console.
irb(main):117:0> ChecklistItem.superjoin(1,2)
ChecklistItem Load (0.5ms) SELECT "checklist_items".* FROM "checklist_items" LEFT OUTER JOIN "checklist_checks" ON "checklist_checks"."checklist_item_id" = "checklist_items"."id" AND "checklist_items"."store_id" = 1 AND "checklist_checks"."report_date" = 1 WHERE (("checklist_items"."store_id" = 2 OR "checklist_items"."store_id" IS NULL)) AND "checklist_items"."list_type" = 'open'
=> #<ActiveRecord::Relation []>
Related
This is for Rails 5 / Active Record 5 using PostgreSQL.
Let's say I have two models: Product and Widget. A Product has_many widgets and a Widget has_many products via a join table called products_widgets.
I want to write a query that finds all products that are associated with both the widget with id=37 and the widget with id=42.
I actually have a list of ids, but if I can write the above query, I can solve this problem in general.
Note that the easier version of this query is to find all widgets that are associated with either the widget with id=37 or the widget with id=42, which you could write as follows:
Product.joins(:products_widgets).where(products_widgets: {widget_id: [37, 42]})
But that isn't what I need.
As a starter: in pure SQL, you could phrase the query with exists conditions:
select p.*
from product p
where
exists (
select 1
from products_widgets pw
where pw.product_id = p.product_id and pw.widget_id = 37
)
and exists (
select 1
from products_widgets pw
where pw.product_id = p.product_id and pw.widget_id = 42
)
In active Record, we can try and use the raw subqueries directly in where conditions:
product
.where('exists(select 1 from products_widgets where product_id = product.product_id and widget_id = ?)', 37)
.where('exists(select 1 from products_widgets where product_id = product.product_id and widget_id = ?)', 42)
I think that using .arel.exist might also work:
product
.where(products_widgets.where('product_id = product.product_id and widget_id = ?', 37).arel.exists)
.where(products_widgets.where('product_id = product.product_id and widget_id = ?', 42).arel.exists)
Order has_many Items is the relationship.
So let's say I have something like the following 2 orders with items in the database:
Order1 {email: alpha#example.com, items_attributes:
[{name: "apple"},
{name: "peach"}]
}
Order2 {email: beta#example.com, items_attributes:
[{name: "apple"},
{name: "apple"}]
}
I'm running queries for Order based on child attributes. So let's say I want the emails of all the orders where they have an Item that's an apple. If I set up the query as so:
orders = Order.joins(:items).where(items: {name:"apple"})
Then the result, because it's pulling at the Item level, will be such that:
orders.count = 3
orders.pluck(:email) = ["alpha#exmaple.com", "beta#example.com", "beta#example.com"]
But my desired outcome is actually to know what unique orders there are (I don't care that beta#example.com has 2 apples, only that they have at least 1), so something like:
orders.count = 2
orders.pluck(:email) = ["alpha#exmaple.com", "beta#example.com"]
How do I do this?
If I do orders.select(:id).distinct, this will fix the problem such that orders.count == 2, BUT this distorts the result (no longer creates AR objects), so that I can't iterate over it. So the below is fine
deduped_orders = orders.select(:id).distinct
deduped_orders.count = 2
deduped_orders.pluck(:email) = ["alpha#exmaple.com", "beta#example.com"]
But then the below does NOT work:
deduped_orders.each do |o|
puts o.email # ActiveModel::MissingAttributeError: missing attribute: email
end
Like I basically want the output of orders, but in a unique way.
I find using subqueries instead of joins a bit cleaner for this sort of thing:
Order.where(id: Item.select(:order_id).where(name: 'apple'))
that ends up with this (more or less) SQL:
select *
from orders
where id in (
select order_id
from items
where name = 'apple'
)
and the in (...) will clear up duplicates for you. Using a subquery also clearly expresses what you want to do–you want the orders that have an item named 'apple'–and the query says exactly that.
use .uniq instead of .distinct
deduped_orders = orders.select(:id).uniq
deduped_orders.count = 2
deduped_orders.pluck(:email) = ["alpha#exmaple.com", "beta#example.com"]
If you want to keep all the attributes of orders use group
deduped_orders = orders.group(:id).distinct
deduped_orders.each do |o|
puts o.email
end
#=> output: "alpha#exmaple.com", "beta#example.com"
I think you just need to remove select(:id)
orders = Order.joins(:items).where(items: {name:"apple"}).distinct
orders.pluck(:email)
# => ["alpha#exmaple.com", "beta#example.com"]
orders = deduped_orders
deduped_orders.each do |o|
puts o.email # loop twice
end
Because I'm newer to Rails than I am SQL, I took a shortcut in order to build a fairly complicated (to me, anyways) query.
Good news: I have the query working
Bad news: I'm stuck with hackish code AND can't figure out how to drop dynamic variables into it (which should be easy if it as in Rails-friendly format).
Here's my query (broken up for ease of reading):
ActiveRecord::Base.connection.select_all( "SELECT DISTINCT ps.* FROM merchants merch, product_sales ps, users u, user_merchant_relations umr, memberships m
INNER JOIN marketing_efforts me ON me.id = ps.marketing_effort_id
LEFT OUTER JOIN member_sales ms ON me.id = ms.marketing_effort_id
LEFT OUTER JOIN crm_sales cs ON me.id = cs.marketing_effort_id
WHERE ((me.start < '2013-05-26' AND me.end > '2013-05-26') OR (me.start < '2013-05-26' AND me.end IS NULL)) AND (me.num_avail IS NULL OR me.num_avail > 0) AND (me.gender IS NULL OR me.gender = u.gender OR u.gender IS NULL) AND ((u.birthdate > me.birthdate_start AND u.birthdate < me.birthdate_end) OR (u.birthdate > me.birthdate_start AND me.birthdate_end IS NULL) OR (me.birthdate_start IS NULL AND me.birthdate_end IS NULL) OR u.birthdate IS NULL) AND ((ms.member_id = m.member_id AND m.user_id = u.id) OR ms.member_id IS NULL) AND (cs.crm_tier_id = umr.crm_tier_id OR cs.crm_tier_id IS NULL) AND u.id = 3 AND merch.id = 1")
Anywhere '2013-05-26' appears I need to drop in Time.now. The u.id and merch.id at the very end also need to accept variables.
I thought I could simply do: .S=?...Q=?....L=?...", value, value, value) but that's not working.
Update
I'm here:
ProductSale.joins('INNER JOIN marketing_efforts ON marketing_efforts.id = product_sales.marketing_effort_id').joins('LEFT OUTER JOIN member_sales ON marketing_efforts.id = member_sales.marketing_effort_id').joins('LEFT OUTER JOIN crm_sales ON marketing_efforts.id = crm_sales.marketing_effort_id')
So I'm taking care of all the joins, but the problem I'm running into is that, with the raw SQL, I could easily define all the tables I need to access ... e.g.:
FROM merchants merch, product_sales ps, users u, user_merchant_relations umr, memberships m
Whereas I'm not sure how to do this with Rails. As a result, I just get FROM product_sales :
ProductSale Load (0.3ms) SELECT "product_sales".* FROM "product_sales" INNER JOIN ....
So when I try to say, for example, users.id = 1 I get:
ActiveRecord::StatementInvalid: SQLite3::SQLException: no such column: users.id:
Any thoughts? Thank you.
Perhaps not the best way of doing it, but it works and solved the main issue (inability to pass dynamic variables):
ProductSale.find_by_sql [ " QUE=?...R=?..Y=? ", value, value, value]
Is it possible to make a query like this with rails?
SELECT items.*, (select count(*) from ads where item_id = items.id) as adscount WHERE 1
And then access this field like so?
#item.adscount
For example for each item there is a hundred ads or so.
And in items index view I need to show how many ads each item has.
For now I only found out how to make a unique query for every item, for example:
select count(*) from ads where item_id = 1
select count(*) from ads where item_id = 2
select count(*) from ads where item_id = 3
etc
Edit:
Made a counter cache column.
That gave me huge performance improvement.
One solution is to use Scopes. You can either have it be a special query, like
class Item < ActiveRecord::Base
scope :with_adscount, select("items.*, (select count(*) from ads where item_id = items.id) as adscount")
end
Then in the controller, or where ever you query from, you can use it like so:
#items = Item.with_adscount
#items.each do |item|
item.adscount
end
Or, you can put it as the default scope using:
class Item < ActiveRecord::Base
default_scope select("items.*, (select count(*) from ads where item_id = items.id) as adscount")
end
Use Rails Counter Cache http://railscasts.com/episodes/23-counter-cache-column
Then you will be able to use it like #item.ads.count and this will not produce any database queries.
And in addition (if you really want to use it your way) you can create a model method
def adscount
self.ads.count
end
You can do it at the controller like this
a = MyModel.find_by_sql('SELECT *, (select count(*) from table) as adscount FROM table').first
a.adscount #total
a.column1 #value of a column
a.column2 #value of another column
There are 3 tables payment_logs, sourcings and purchasings in our rails app. A payment_log belongs to either sourcing or purchasing but not both at the same time. There is a col project_id in both sourcing and purchasing. We want to pick up all payment_logs with its project_id = project_id_search (project_id_search passed from a search page). Also we need a ActiveRecord as resultset returned. Here is the individual query, assuming payment_logs holds the ActiveRecord result set:
pick all payment_logs with its sourcing's project_id = project_id_search
payment_logs = payment_logs.joins(:sourcing).where("sourcings.project_id = ?", project_id_search)
pick all payment_logs with its purchasing's project_id = project_id_search
payment_logs = payment_logs.(:purchasing).where("purchasings.project_id = ?", project_id_search)
We need to union 1 and 2 in order to pick up all the payment_logs whose project_id = project_id_search. What's the right way to accomplish it? We did not find union in rails and find_by_sql returns an array which is not what we want. Thanks.
payment_logs.where(["
payment_logs.sourcing_id IN (
SELECT id FROM sourcings WHERE sourcings.project_id = ?
)
OR payment_logs.purchasing_id IN
(
SELECT id FROM purchasings WHERE purchasings.project_id = ?
)", project_id_search, project_id_search])
Lot of SQL, but it should work
Option 2 (two SQL requests ...) :
payment_logs = []
payment_logs << PaymentLog.joins(:sourcing).where("sourcings.project_id" => project_id_search)
payment_logs << PaymentLog.joins(:purchasing).where("purchasings.project_id" => project_id_search)
payment_logs.uniq! #In case some records have both a sourcing and a purchasing
Option 3, with the squeel gem : https://github.com/ernie/squeel
PaymentLog.where{(source_id.in Sourcing.where(:project_id => project_id_search)) | (purchasing_id.in Purchasing.where(:project_id => project_id_search))}
I like this solution :)
Also, whenever you have a doubt on the generated SQL, from the console or anywhere else, you can add .to_sql at the end of an ActiveRecord query to double check the generated SQL