In my Rails 5 + Postgres app I make a query like this:
user = User.where("name = ?", name).first.email
So this gives me the email of the first user with the name.
But if no user with this names exists I get an error:
NoMethodError (undefined method `email' for nil:NilClass)
How can I check if I have any results before using the method?
I can think if various ways to do this using if-clauses:
user = User.where("name = ?", name).first
if user
user_email = user.email
end
But this does not seem to be the most elegant way and I am sure Rails has a better way.
You can use find_by, returns the object or nil if nothing is found.
user = User.find_by(name: name)
if user
...
end
That being said you could have still used the where clause if you're expecting more than one element.
users = User.where(name: name)
if users.any?
user = users.first
...
end
Then there is yet another way as of Ruby 2.3 where you can do
User.where(name: name).first&.name
The & can be used if you're not sure if the object is nil or not, in this instance the whole statement would return nil if no user is found.
I use try a lot to handle just this situation.
user = User.where("name = ?", name).first.try(:email)
It will return the email, or if the collection is empty (and first is nil) it will return nil without raising an error.
The catch is it'll also not fail if the record was found but no method or attribute exists, so you're less likely to catch a typo, but hopefully your tests would cover that.
user = User.where("name = ?", name).first.try(:emial)
This is not a problem if you use the Ruby 2.3 &. feature because it only works with nil object...
user = User.where("name = ?", name).first&.emial
# this will raise an error if the record is found but there's no emial attrib.
You can always use User.where("name = ?", name).first&.email, but I disagree that
user = User.where("name = ?", name).first
if user
user_email = user.email
end
is particularly inelegant. You can clean it up with something like
def my_method
if user
# do something with user.email
end
end
private
def user
#user ||= User.where("name = ?", name).first
# #user ||= User.find_by("name = ?", name) # can also be used here, and it preferred.
end
Unless you really think you're only going to use the user record once, you should prefer being explicit with whatever logic you're using.
Related
I'm running a query like the below:
Item.where("created_at >=?", Time.parse(params[:created_at])).where(status_id: params[:status_id])
...where the user can decide to NOT provide a parameter, in which case it should be excluded from the query entirely. For example, if the user decides to not pass a created_at and not submit it, I want to run the following:
Item.where(status_id: params[:status_id])
I was thinking even if you had a try statement like Time.try(:parse, params[:created_at]), if params[created_at] were empty, then the query would be .where(created_at >= ?", nil) which would NOT be the intent at all. Same thing with params[:status_id], if the user just didn't pass it, you'd have a query that's .where(status_id:nil) which is again not appropriate, because that's a valid query in itself!
I suppose you can write code like this:
if params[:created_at].present?
#items = Item.where("created_at >= ?", Time.parse(params[:created_at])
end
if params[:status_id].present?
#items = #items.where(status_id: params[:status_id])
end
However, this is less efficient with multiple db calls, and I'm trying to be more efficient. Just wondering if possible.
def index
#products = Product.where(nil) # creates an anonymous scope
#products = #products.status(params[:status]) if params[:status].present?
#products = #products.location(params[:location]) if params[:location].present?
#products = #products.starts_with(params[:starts_with]) if params[:starts_with].present?
end
You can do something like this. Rails is smart in order to identify when it need to build query ;)
You might be interested in checking this blog It was very useful for me and can also be for you.
If you read #where documentation, you can see option to pass nil to where clause.
blank condition :
If the condition is any blank-ish object, then #where is a no-op and returns the current relation.
This gives us option to pass conditions if valid or just return nil will produce previous relation itself.
#items = Item.where(status_condition).where(created_at_condition)
private
def status_condition
['status = ?', params[:status]] unless params[:status].blank?
end
def created_at_condition
['created_at >= ?', Time.parse(params[:created_at])] unless params[:created_at].blank?
end
This would be another option to achieve the desired result. Hope this helps !
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 before_action method like this:
def current_user
#current_user ||= User.find(:id => session[:id])
end
And I call a method like this:
def get_food user
food = Food.find(:id => user.id)
end
This is fine, but I want to add exception handling.
When the user is nil I want to use #current_user:
def get_food user
food = Food.find(if user is nil i want to use #current_user.id)
end
Of course, I can write it like this:
def get_food user
if user.nil?
food = Food.find(#current_user.id)
else
food = Food.find(user.id)
end
Or, is this the best way?
def get_food user
food = Food.find(user == nil? #current_user.id : user.id)
end
I'm curious is there a better way than adding a simple if statement inside the param?
The shortest one lines I can think of are something like this:
Food.find((user || current_user).id)
Food.find(user.try(:id) || current_user.id)
Food.find(user ? user.id : current_user.id)
Not sure if this is really an impovement in readability. I would prefer something like this:
def get_food(user)
user ||= current_user
Food.find(user.id)
end
You can use ternary operator to make it one line:
user ? Food.find(user.id) : Food.find(#current_user.id)
How about arrays
food = Food.where(id: [#current_user.try(:id),user.id]).first
You can try this:
food = Food.find(user.nil? ? #current_user.id : user.id)
What about default parameters?
def get_food(user = #current_user)
food = Food.find(user.id)
end
It will work if you call it without the parameter
something.get_food # notice the method is called with no params
If you want it working also if you pass nil, you should also add:
def get_food(user = #current_user)
food = Food.find((user || #current_user).id)
end
However is strange that foods and users have the same ids...
Maybe the correct query is:
food = Food.find_by_user_id((user || #current_user).id)
or, if users have more than just one food:
foods = Food.where(user: (user || #current_user)) # rails 4, :user => (user || #current_user) for rails 3
Food.find(user.id rescue #current_user.id)
Given the following associations:
User has_many Ultimate_Source
User has_many Budget_Source
How do I create the following method:
def foo(source)
user = User.find(1)
user.source.id
end
such that
foo(ultimate_sources)
returns:
user.ultimate_sources.id
thanks.
if source == :ultimate_sources or source == :budget_sources then the following should work.
user.send(source).id
if source is a string you can always convert it to a symbol using source.to_sym.
The send method takes a symbol representing the name of a method and sends the message to that object to execute that method and return what the method returns.
http://ruby-doc.org/core-2.1.2/Object.html#method-i-send
You can use the method .send:
def foo(source)
user = User.find(1)
user.send(source.to_s).id
end
Or the method .try (will not raise a NoMethodError if source is not a method of User):
def foo(source)
user = User.find(1)
user.try(source.to_s).id
end
But I really hope that source is not something coming from the user's input. What if I send delete as value of source variable? It would delete the user ...
I highly recommend you to limit the possible methods, in your case it could be something like this:
def foo(source)
user = User.find(1)
user.try("#{source.gsub('_sources', '')}_sources").try(:id)
end
This code version protects you to send destroy as the source value:
foo('ultimate_sources') # => user.ultimate_sources.id
foo('destroy') # => nil
foo('destroy_sources') # => nil
foo('budget') # => user.budget_sources.id
You could also put in a guard clause to be safe.
def foo(source)
return if (source != 'ultimate_sources') || (source != 'budget_sources')
user = User.find(1)
user.send(source).id
end
Ok so i have this helper
def current_company_title
(Company.find_by_id(params["company_id"]).name rescue nil) || (#companies.first.name rescue nil) current_user.company.name
end
Basically what I am achieving with this is the following ...
If the param["company_id"] exists then try to get the company and if not then
if #companies exists grab the first company name and if not then get the current users company name
This works but the rescues seem like a hack...any idea on another way to achieve this
Indeed rescue is kind of a hack, id' probably split it up into two methods and then use try to fetch the name if available: http://api.rubyonrails.org/classes/Object.html#method-i-try
def current_company
#current_company ||= Company.find_by_id(params[:company_id]) || #companies.try(:first) || current_user.try(:company)
end
def current_company_name
current_company.try(:name)
end
Company.find_by_id(params["company_id"]).name`
find and its derivates are meant to be used when you're sure-ish you'll have a positive result, and only in some cases (row was deleted, etc) errors. That's why it raises an exception. In your case, you're assuming it's gonna fail, so a regular where, which would return nil if no rows was found, would do better, and remove the first rescue
#companies.first.name rescue nil
could be replaced by
#companies.first.try(:name)
I'll let you check the api for more on the topic of try. It's not regular ruby, it's a Rails addition.
Less "magic", simple code, simple to read:
def current_company_title
company = Company.where(id: params["company_id"]).presence
company ||= #companies.try(:first)
company ||= current_user.company
company.name
end
Ps. Not a big fan of Rails' try method, but it solves the problem.
def current_company_title
if params["company_id"]
return Company.find_by_id(params["company_id"]).name
elsif #companies
return #companies.first.name
else
return current_user.company.name
end
end
The rescues are a hack, and will obscure other errors if they occur.
Try this:
(Company.find_by_id(params["company_id"].name if Company.exists?(params["company_id"]) ||
(#companies.first.name if #companies && #companies.first) ||
current_user.company.name
then you can extract each of the bracketed conditions to their own methods to make it more readable, and easier to tweak the conditions:
company_name_from_id(params["company_id"]) || name_from_first_in_collection(#companies) || current_user_company_name
def company_name_from_id(company_id)
company=Company.find_by_id(company_id)
company.name if company
end
def name_from_first_in_collection(companies)
companies.first.name if companies && companies.first
end
def current_user_company_name
current_user.company.name if current_user.company
end
[Company.find_by_id(params["company_id"]),
#companies.to_a.first,
current_user.company
].compact.first.name