I've created a custom validator in Rails 3.0 which validates whether a combination of columns is unique within a table. The entire code of the validation is:
class UniqueInProjectValidator < ActiveModel::EachValidator
def validate_each(object, attribute, value)
unless object.class.where("project_id = ? AND #{attribute} = ?", object.project_id, value).empty?
if object.new_record?
object.errors[attribute] << (options[:message] || "must be unique in each project")
else
orig_rec = object.class.find(object.id)
if value != orig_rec.method(attribute).call || object.project_id != orig_rec.project_id
object.errors[attribute] << (options[:message] || "must be unique in each project")
end
end
end
end
Note that it is not easy to recognize what the if statements do, so I was hoping to be able to replace the unless conditional with a def attribute_and_project_exist? method and the second if statement with a def attribute_or_project_changed? method. However when creating those methods, the arguments from validates_each do not pass because of encapsulation.
Now the question: Is there a way to somehow cleanly allow those variables to be accessed by my two newly created methods as one can do with column names in a model, or am I stuck with the options of either passing each argument again or leaving the hard to read conditional statements?
Thanks in advance!
I suppose you could clean it up a bit with one variable, one lambda, and one "return as soon as possible":
def validate_each(object, attribute, value)
# If there is no duplication then bail out right away as
# there is nothing to check. This reduces your nesting by
# one level. Using a variable here helps to make your
# intention clear.
attribute_and_project_exists = object.class.where("project_id = ? AND #{attribute} = ?", object.project_id, value).empty?
return unless attribute_and_project_exists
# This lambda wraps up your second chunk of ugly if-ness and saves
# you from computing the result unless you have to.
attribute_or_project_changed = lambda do
orig_rec = object.class.find(object.id)
value != orig_rec.method(attribute).call || object.project_id != orig_rec.project_id
end
# Note that || short-circuits so the lambda will only be
# called if you have an existing record.
if object.new_record? || attribute_or_project_changed.call
object.errors[attribute] << (options[:message] || "must be unique in each project")
end
end
I don't know how much better that is than your original but the logic and control flow is a lot clearer to me due to the nicer chunking.
Related
I'm using Rails 4 and have an Article model that has answer, side_effects, and benefits as attributes.
I am trying to create a before_save method that automatically looks at the side effects and benefits and creates links corresponding to another article on the site.
Instead of writing two virtually identical methods, one for side effects and one for benefits, I would like to use the same method and check to assure the attribute does not equal answer.
So far I have something like this:
before_save :link_to_article
private
def link_to_article
self.attributes.each do |key, value|
unless key == "answer"
linked_attrs = []
self.key.split(';').each do |i|
a = Article.where('lower(specific) = ?', i.downcase.strip).first
if a && a.approved?
linked_attrs.push("<a href='/questions/#{a.slug}' target=_blank>#{i.strip}</a>")
else
linked_attrs.push(i.strip)
end
end
self.key = linked_attrs.join('; ')
end
end
end
but chaining on the key like that gives me an undefined method 'key'.
How can I go about interpolating in the attribute?
in this bit: self.key you are asking for it to literally call a method called key, but what you want, is to call the method-name that is stored in the variable key.
you can use: self.send(key) instead, but it can be a little dangerous.
If somebody hacks up a new form on their browser to send you the attribute called delete! you don't want it accidentally called using send, so it might be better to use read_attribute and write_attribute.
Example below:
def link_to_article
self.attributes.each do |key, value|
unless key == "answer"
linked_attrs = []
self.read_attribute(key).split(';').each do |i|
a = Article.where('lower(specific) = ?', i.downcase.strip).first
if a && a.approved?
linked_attrs.push("<a href='/questions/#{a.slug}' target=_blank>#{i.strip}</a>")
else
linked_attrs.push(i.strip)
end
end
self.write_attribute(key, linked_attrs.join('; '))
end
end
end
I'd also recommend using strong attributes in the controller to make sure you're only permitting the allowed set of attributes.
OLD (before I knew this was to be used on all attributes)
That said... why do you go through every single attribute and only do something if the attribute is called answer? why not just not bother with going through the attributes and look directly at answer?
eg:
def link_to_article
linked_attrs = []
self.answer.split(';').each do |i|
a = Article.where('lower(specific) = ?', i.downcase.strip).first
if a && a.approved?
linked_attrs.push("<a href='/questions/#{a.slug}' target=_blank>#{i.strip}</a>")
else
linked_attrs.push(i.strip)
end
end
self.answer = linked_attrs.join('; ')
end
I created a simple web form where users can enter some search criteria to look for venues e.g. a price range. When a user clicks "find" I use active record to query the database. This all works very well if all fields are filled in. Problems occur when one or more fields are left open and therefore have a value of null.
How can I work around this in my controller? Should I first check whether a value is null and create a query based on that? I can imagine I end up with many different queries and a lot of code. There must be a quicker way to achieve this?
Controller:
def search
#venues = Venue.where("price >= ? AND price <= ? AND romance = ? AND firstdate = ?", params[:minPrice], params[:maxPrice], params[:romance], params[:firstdate])
end
You may want to filter out all of the blank parameters that were sent with the request.
Here is a quick and DRY solution for filtering out blank values, triggers only one query of the database, and builds the where clause with Rails' ActiveRecord ORM.
This approach safeguards against SQL-injection, as pointed out by #DanBrooking. Rails 4.0+ provides "strong parameters." You should use the feature.
class VenuesController < ActiveRecord::Base
def search
# Pass a hash to your query
#venues = Venue.where(search_params)
end
private
def search_params
params.
# Optionally, whitelist your search parameters with permit
permit(:min_price, :max_price, :romance, :first_date).
# Delete any passed params that are nil or empty string
delete_if {|key, value| value.blank? }
end
end
I would recommend to make method in Venue
def self.find_by_price(min_price, max_price)
if min_price && max_price
where("price between ? and ?", min_price, max_price)
else
all
end
end
def self.find_by_romance(romance)
if romance
where("romance = ?", romance)
else
all
end
end
def self.find_by_firstdate(firstdate)
if firstdate
where("firstdate = ?", firstdate)
else
all
end
end
And use it in your controller
Venue
.find_by_price(params[:minPrice], params[:maxPrice])
.find_by_romance(params[:romance])
.find_by_firstdate(params[:firstdate])
Another solution to this problem, and I think a more elegant one, is using scopes with conditions.
You could do something like
class Venue < ActiveRecord::Base
scope :romance, ->(genre) { where("romance = ?", genre) if genre.present? }
end
You can then chain those, which would work as an AND if there is no argument present, then it is not part of the chain.
http://guides.rubyonrails.org/active_record_querying.html#scopes
Try below code, it will ignore parameters those are not present
conditions = []
conditions << "price >= '#{params[:minPrice]}'" if params[:minPrice].present?
conditions << "price <= '#{params[:maxPrice]}'" if params[:maxPrice].present?
conditions << "romance = '#{params[:romance]}'" if params[:romance].present?
conditions << "firstdate = '#{params[:firstdate]}'" if params[:firstdate].present?
#venues = Venue.where(conditions.join(" AND "))
I have a web app that uses Mongoid's find_or_create_by method.
poll = Poll.find_or_create_by(fields)
Before we go into production, I am trying to run through the failure scenarios, and it occurred to me that multiple users could try to access this resource using this method. Is there any likelihood that it could create multiple instances of the same object? What can I do to prevent that?
Disclaimer: I'm new to Mongoid and Rails so I could be totally wrong.
Looking at modifiable.rb and many.rb it doesn't appear to lock on any resource.
It looks to me like it simply does a "where().first" query and then if that does not return anything it performs a "create" query:
def find_or(method, attrs = {}, &block)
where(attrs).first || send(method, attrs, &block)
end
For find_or_create_by, "send" would call "create_document":
def create_document(method, attrs = nil, &block)
klass.__send__(method,
selector.reduce(attrs || {}) do |hash, (key, value)|
unless key.to_s =~ /\$/ || value.is_a?(Hash)
hash[key] = value
end
hash
end, &block)
end
Conclusion: To me this appears to be a convenience method, do not expect it to be "thread-safe". I wish there was more documentation available on this method.
sometimes it is not worth to implement a thread-safe code in MRI. because most of the time we have to go multiprocess deployments. so the easiest solution is database level uniqueness validations.
class Person
include Mongoid::Document
field :ssn
index({ ssn: 1 }, { unique: true })
end
so when save failed rollback the transaction. keep in mind that model level uniqueness validation is also error-prone.
I have a class that I use to contain select menu options for property types. It works fine. However, I need to be able to verify the selection and perform specific logic based on the selected option. This needs to happen in my Ruby code and in JavaScript.
Here is the class in question:
class PropertyTypes
def self.[](id)
##types[id]
end
def self.options_for_select
##for_select
end
private
##types = {
1 => "Residential",
2 => "Commercial",
3 => "Land",
4 => "Multi-Family",
5 => "Retail",
6 => "Shopping Center",
7 => "Industrial",
8 => "Self Storage",
9 => "Office",
10 => "Hospitality"
}
##for_select = ##types.each_pair.map{|id, display_name| [display_name, id]}
end
What is the best way to verify the selection? I need to perform specific logic and display user interface elements based on each type of property type.
Since I am storing the id, I would be verifying that the id is a particular property type. Something like:
PropertyTypes.isResidential?(id)
Then this method would look like this:
def self.isResidential?(id)
##types[id] == "Residential"
end
But now I am duplicating the string "Residential".
For JavaScript, I assume I would make an ajax call back to the model to keep the verification code DRY, but this seems like over kill.
Do I need to manually create a verification method for each property type or can I use define_method?
This seems so basic yet I am confused and burned out on this problem.
Thanks
===
Here's my solution:
class << self
##types.values.each do |v|
# need to remove any spaces or hashes from the found property type
v = v.downcase().gsub(/\W+/, '')
define_method "is_#{v}?", do |i|
type_name = ##types[i]
return false if type_name == nil #in case a bogus index is passed in
type_name = type_name.downcase().gsub(/\W+/, '')
type_name == v
end
end
end
It sounds like you can benefit from some Ruby meta-programming. Try googling "ruby method_missing". You can probably do something quick & dirty along the lines of:
class PropertyTypes
def method_missing(meth, *args, &block)
if meth.to_s =~ /^is_(.+)\?$/
##types[args.first] == $1
else
super
end
end
end
On the ruby side you could also use something like this to define dynamically these methods:
class << self
##types.values.each do |v|
define_method "is_#{v}?", do |i|
##types[i] == v
end
end
end
I have an activity model which has_many participants and I'd like to ensure that a participant always exists when updating an activity and its participants. I have the following method in my activity model which does the trick:
def must_have_participant
if self.participants.size == 0 || self.participants.size == self.participants.to_ary.find_all{ |p| p.marked_for_destruction? }.count
self.errors[:base] << I18n.t(:msg_activity_must_have_participant)
end
end
The problem is that the participants are lazy loaded if I'm simply updating the activity on its own which I'd like to avoid. I've tried the following alternative, however, loaded? returns false when removing all participants using the :_destroy flag.
def must_have_participant
if self.new_record? || self.participants.loaded?
if self.participants.size == 0 || self.participants.size == self.participants.to_ary.find_all{ |p| p.marked_for_destruction? }.count
self.errors[:base] << I18n.t(:msg_activity_must_have_participant)
end
end
end
Is there an alternative to loaded? that I can use to know whether the participants are going to be updated?
I did something like this in a recent validation that I created. I searched for the original record and checked the original value against the new value. No guarantees my code will work for you but here is my code for your application:
orig_rec = self.find(id)
if participant_ids.size != orig_rec.participant_ids.size
Note that I checked the size of participant_ids instead of fetching all the participant records and checking the size of them. That should be more efficient.
I don't know if there is some kind of built in function to do this or not in ruby, I'll be curious to see what someone who is more rails specific may suggest.
For reference I've amended the method like so:
def must_have_participant
if self.new_record? || self.association(:participants).loaded?
if self.participants.size == 0 || self.participants.size == self.participants.select{ |p| p.marked_for_destruction? }.size
self.errors[:base] << I18n.t(:msg_must_have_participant)
end
end
end