I have this chain of models:
And I have a list of suboptions. I would like to group_by the suboptions by each of the above fathers.
So, it would be divided into Categories, which would be divided into activities, which would be divided into options. Do you see?
How can I do this??
Suboptions.all.group_by(???)
# or
Suboptions.all.order_by(????)
# or ????
Example:
Suboptions
+----+-----------+
| id | option_id |
+----+-----------+
| 1 | 1 |
| 2 | 2 |
| 3 | 2 |
| 4 | 4 |
| 5 | 3 |
+----+-----------+
Options
+----+-------------+
| id | activity_id |
+----+-------------+
| 1 | 1 |
| 2 | 2 |
| 3 | 1 |
| 4 | 3 |
+----+-------------+
Activities
+----+-------------+
| id | category_id |
+----+-------------+
| 1 | 1 |
| 2 | 2 |
| 3 | 1 |
+----+-------------+
Categories
+----+------+
| id | name |
+----+------+
| 1 | cat1 |
| 2 | cat2 |
+----+------+
then the search should return the suboptions with ids in the following order: [1,5,4,2,3]
( try to visualize it from the category to the suboption )
Depending on the size of the data set, sounds like you want to do a join, followed by an order
Suboption.joins(option: {activity: :category}).order('categories.name')
Will that work for you?
I am doing an inner join (at least I think that's what the code is doing) but my search is returning the same result multiple times. I think I have something wrong with my join.
Tags
+------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+------------+--------------+------+-----+---------+----------------+
| id | int(11) | NO | PRI | NULL | auto_increment |
| name | varchar(255) | YES | | NULL | |
| color | varchar(255) | YES | | NULL | |
| article_id | int(11) | YES | MUL | NULL | |
| created_at | datetime | NO | | NULL | |
| updated_at | datetime | NO | | NULL | |
+------------+--------------+------+-----+---------+----------------+
Articles
+------------+--------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+------------+--------------+------+-----+---------+----------------+
| id | int(11) | NO | PRI | NULL | auto_increment |
| title | varchar(255) | YES | | NULL | |
| info | text | YES | | NULL | |
| created_at | datetime | NO | | NULL | |
| updated_at | datetime | NO | | NULL | |
+------------+--------------+------+-----+---------+----------------+
Each article I have tagged and it returns multiple results based on the tags. So if the article has 3 tags this results in 3 records being returned. Even though each article should only have 1 returned?
class Article < ApplicationRecord
has_many :tags, dependent: :destroy
validates :title, presence: true
def self.search(search)
if search
joins(:tags).where('title LIKE :search OR tags.name LIKE :search', search: "%#{search}%")
else
all
end
end
end
Use .distinct or .group. There is .uniq alias too starting from rails 4.0.2.
Example:
joins(:tags).where('title LIKE :search OR tags.name LIKE :search', search: "%#{search}%").distinct
joins(:tags).where('title LIKE :search OR tags.name LIKE :search', search: "%#{search}%").
group('article_id')
I'm using the PublicActivity gem: https://github.com/pokonski/public_activity
All of the models I use PublicActivity to track use the column edition_id. And I'm wondering how I can scope by that column since it's polymorphic relation.
Eg PublicActivity looks like so:
PublicActivity::Activity.limit(50)
+----+--------------+----------------+----------+------------+--------------------+------------+--------------+----------------+---------------------------+---------------------------+
| id | trackable_id | trackable_type | owner_id | owner_type | key | parameters | recipient_id | recipient_type | created_at | updated_at |
+----+--------------+----------------+----------+------------+--------------------+------------+--------------+----------------+---------------------------+---------------------------+
| 1 | 42 | QuizMaster | 6 | User | quiz_master.update | {} | | | 2013-04-12 17:33:14 +0100 | 2013-04-12 17:33:14 +0100 |
| 2 | 25 | Place | 6 | User | place.update | {} | | | 2013-04-12 17:42:42 +0100 | 2013-04-12 17:42:42 +0100 |
| 3 | 25 | Event | 6 | User | event.update | {} | | | 2013-04-12 17:45:08 +0100 | 2013-04-12 17:45:08 +0100 |
| 4 | 20 | QuizMaster | 6 | User | quiz_master.update | {} | | | 2013-04-12 17:49:09 +0100 | 2013-04-12 17:49:09 +0100 |
| 5 | 20 | QuizMaster | 6 | User | quiz_master.update | {} | | | 2013-04-12 17:50:51 +0100 | 2013-04-12 17:50:51 +0100 |
+----+--------------+----------------+----------+------------+--------------------+------------+--------------+----------------+---------------------------+---------------------------+
Where edition_id is on the polymorphic trackable relation.
What I would like to do it something like:
PublicActivity::Activity.limit(50).includes(:trackable)# where trackable edition = 1
I'm not sure how or even if it's possible to join or include or preload a polymorphic model across multiple "trackable" types.
I don't think this is possible directly with SQL because of the polymorphism, but you could do something with ruby pretty easily.
PublicActivity::Activity.limit(50).select {|c| c.trackable.edition_id = 1 }
It won't be as efficient as a SQL query but using Rails preload it'll probably be good enough depending on the number of records you're fetching.
PublicActivity::Activity.limit(50).preload(:trackable).select {|c| c.trackable.edition_id = 1 }
I populate book items into my Book model,
But I found there are many items have the same id.
So, how to create unique id for items. To prevent many items from having the same id ?
Here is the book model code
# encoding: utf-8
class Book < ActiveRecord::Base
attr_accessible :name, :isbn ,:price ,:comment ,:author ,:sale_type ,:publisher ,:sn ,:category
attr_accessible :location, :category, :release_date
validates_uniqueness_of :sn
Here are the part of my items
irb(main):058:0> Book.all[1..10]
+-----+------+-----+-----+------+------+-----+------+-----+-----+------+------+-----+------+
| id | pric | com | cre | upda | rele | loc | sn | isb | aut | sale | name | cat | publ |
+-----+------+-----+-----+------+------+-----+------+-----+-----+------+------+-----+------+
| 118 | 4543 | 作 | 201 | 2013 | 2006 | --- | 2124 | 978 | 趙 | prom | 求索 | 商 | 聯經 |
| 118 | 872 | 馬 | 201 | 2013 | 2013 | --- | 2124 | 978 | 黎 | prom | 告別 | 政 | 聯經 |
| 118 | 2105 | 某 | 201 | 2013 | 2012 | --- | 2124 | 978 | 吳 | prom | 複眼 | 政 | 夏日 |
| 118 | 301 | 作 | 201 | 2013 | 2006 | --- | 2124 | 978 | 王 | norm | 天香 | 歷 | 麥田 |
| 118 | 411 | 少 | 201 | 2013 | 2008 | --- | 2124 | 978 | 韓 | norm | 鞋癖 | 商 | 聯經 |
| 119 | 3751 | 有 | 201 | 2013 | 2010 | --- | 2124 | 978 | 紀 | prom | 私家 | 體 | 印刻 |
| 119 | 3361 | 文 | 201 | 2013 | 2010 | --- | 2124 | 978 | 林 | fix_ | 我不 | 體 | 印刻 |
| 119 | 1140 | 何 | 201 | 2013 | 2012 | --- | 2124 | 978 | 邁 | norm | 正義 | 體 | 雅言 |
| 119 | 888 | 一 | 201 | 2013 | 2007 | --- | 2124 | 978 | 福 | fix_ | 生命 | 商 | 究竟 |
| 119 | 3283 | 近 | 201 | 2013 | 2011 | --- | 2124 | 978 | 芮 | norm | 海拉 | 政 | 遠流 |
+-----+------+-----+-----+------+------+-----+------+-----+-----+------+------+-----+------+
here the rake code to generate my data
16 bk = Book.new(:sn => real_sn,:name => book_name, :isbn=>isbn,
17 :price =>Random.rand(200..5000), :location=>location, :category=>["商業","歷史","體育","政治"].sample,
18 :author => author, :sale_type => [:fix_priced, :normal, :promotion].sample, :publisher => publisher,
19 :release_date => rand(10.years).ago, :comment => comment
20 )
Columns in the table I use the Postgre DB
Column | Type | Modifiers
--------------+-----------------------------+----------------------------------------------------
id | integer | not null default nextval('books_id_seq'::regclass)
price | integer |
comment | text |
created_at | timestamp without time zone | not null
updated_at | timestamp without time zone | not null
release_date | text |
location | text |
sn | bigint |
isbn | bigint |
author | text |
sale_type | text |
name | text |
category | text |
publisher | text |
Indexes:
"books_pkey" PRIMARY KEY, btree (id)
The code above, does not save any record in the database, it just instantiate objects of Book model. You should either save the object after initialization bk.save or use the create method instead of new.
bk = Book.new(:sn => real_sn,:name => book_name, :isbn=>isbn,
:price =>Random.rand(200..5000), :location=>location,
:category=>["商業","歷史","體育","政治"].sample, :author => author,
:sale_type => [:fix_priced, :normal, :promotion].sample,
:publisher => publisher, :release_date => rand(10.years).ago,
:comment => comment)
bk.save
Or alternatively you can use the create method
bk = Book.create(:sn => real_sn,:name => book_name, :isbn=>isbn,
:price =>Random.rand(200..5000), :location=>location,
:category=>["商業","歷史","體育","政治"].sample, :author => author,
:sale_type => [:fix_priced, :normal, :promotion].sample,
:publisher => publisher, :release_date => rand(10.years).ago,
:comment => comment)
Once saved in the database, it will automatically gain a unique id.
I have a very simple migration:
class RemoveAuthorIdFromBooks < ActiveRecord::Migration
def change
remove_column :books, :author_id
end
end
But I get the following error:
Mysql2::Error: Error on rename of './mysite_staging/#sql-3b1_3c78' to './mysite_staging/books' (errno: 150): ALTER TABLE `books` DROP `author_id`
This is the description of the table:
+------------------+---------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+------------------+---------------+------+-----+---------+----------------+
| id | int(11) | NO | PRI | NULL | auto_increment |
| author_id | int(11) | NO | MUL | NULL | |
| title | varchar(255) | NO | | NULL | |
| teaser | varchar(500) | NO | | NULL | |
| description | varchar(2000) | YES | | NULL | |
| cover_image | varchar(255) | NO | | NULL | |
| publication_date | date | NO | | NULL | |
| enabled | tinyint(1) | NO | | 1 | |
| created_at | datetime | NO | | NULL | |
| updated_at | datetime | NO | | NULL | |
| excerpt | text | YES | | NULL | |
| featured | tinyint(1) | YES | | NULL | |
| site_id | int(11) | YES | | NULL | |
+------------------+---------------+------+-----+---------+----------------+
Any clues?
If anyone still facing this with Rails 4 and above, then you could do the following
remove_reference(:books, :author, index: true, foreign_key: true)
For some reason a foreign key constraint was breaking the drop sentence.
I did the following:
show create table books;
Looked at the foreign key name and then:
alter table books drop foreign key books_ibfk_1;
Then rake db:migrate worked.