I get an ArgumentError for line #3 in my views/-/new.html.erb file that states:
"wrong number of arguments (1 for 3..4)"
<div class='form-group'>
<%= form.label :category %>
<%= form.select "category", options_from_collection_for_select([{1 => 'Food'}, {2 => 'Entertainment'}]) %>
</div>
The Application trace states:
app/views/events/new.html.erb:14:in block in _app_views_events_new_html_erb__1569841425540097418_70204987081640'
app/views/events/new.html.erb:5:in_app_views_events_new_html_erb__1569841425540097418_70204987081640'
erb:14 is line #3 above, and erb:5 is
<%= form_for #event do |form| %>
The error message is being thrown by options_from_collection_for_select([{1 => 'Food'}, {2 => 'Entertainment'}])
[{1 => 'Food'}, {2 => 'Entertainment'}] is an array being passed as one argument to options_from_collection_for_select method; hence the error message.
The correct form for calling the options_from_collection_for_select helper method is
options_from_collection_for_select(collection, value_method, text_method, selected = nil)
See more details and usage examples at http://apidock.com/rails/ActionView/Helpers/FormOptionsHelper/options_from_collection_for_select
I think you'd be better off using options_for_select, which will work with the arguments you've provided and just replace the options_from_collection_for_select in your code. The documentation for which can be found here.
You need to decide whether you want the number (1, 2 etc.) or the word (Food, Entertainment etc.) to be the "value" which reaches your back-end. For example:
options_for_select({'Food' => 1, 'Entertainment' => 2})
The output of the above in HTML would be as follows:
<option value="1">Food</option>
<option value="2">Entertainment</option>
The options_from_collection_for_select method requires 3 arguments. You've only provided one: [{1 => 'Food'}, {2 => 'Entertainment'}]. Your second argument should be a method/attribute on each object in the collection which represents the "value", the third should be the label for that value.
Related
I have a field call query_data defined as text in my MySQL database.
In my model I defined this field as serialize :query_data, JSON.
The JSON format I would like to save and retrieve look like that:
{:items => [
{:id => 1},
{:id => 2},
{:id => 3}
]}
I have a collection (in that case, called items) that contain an array of objects.
I was wondering, what's the best way to add or delete an Item.
Ex: remove {:id => 2} from my items list and add `{:id => 4} to it
Ruby on Rails has some nice methods to move seamlessly between JSON and Ruby.
thing = {:items => [
{:id => 1},
{:id => 2},
{:id => 3}
]}
thing.to_json # "{\"items\":[{\"id\":1},{\"id\":2},{\"id\":3}]}"
thing.to_json is essentially what's happening in the serializer. If you want them back to Ruby, you can just do:
#items = #thing.query_data
JSON.parse(#items) # "items"=>[{"id"=>1}, {"id"=>2}, {"id"=>3}]}
Now that we can easily move between the two, lets just use Ruby syntax to deal with adding and deleting keys.
thing = {:items => [
{:id => 1},
{:id => 2},
{:id => 3}
]}
thing[:items] = thing[:items].append({:id => 4}) # adding a new item
thing[:items] = thing[:items].select { |item| item[:id] != 2 } # removing an item
First: The second argument to serialize should be the class of object you're storing in the field. You should have serialize :query_data, Hash instead.
Besides that, there aren't really any established best practices for working with serialized data. It really just depends too much on the structure of your data. You might as well ask, "what's the best way to add or delete an item from a hash?"
But since this is a hash you should make sure to keep dirty attributes in mind. If you were to do something like:
items = my_model.query_data[:items]
items.reject! {|item| item[:id] == 2}
items += {id: 4}
then the model wouldn't know that query_data changed and should be updated on save.
my_model.changed?
# => false
my_model.save
# Won't actually save changes to db.
To avoid this, you can:
A) Make sure you only ever set my_model.query_data directly
B) Explicitly call my_model.query_data_will_change! after changing that field so that it will be properly updated on save.
Base on #veridian-dynamics (thanks for your help!) Here what I did.
Model:
class MyModel < ApplicationRecord
serialize :item_data, JSON
end
Controller:
class ItemController < ApplicationController
before_action :authenticate_user!
def add_item
begin
mymodel = MyModel.find_or_create_by(id: param[:model_id])
if mymodel .item_data.blank?
item = {:items => []}
else
item = mymodel.item_data.deep_symbolize_keys
end
bookmark_exist = item[:items].any? {|i| i[:id] == params[:id]}
if !bookmark_exist
item[:items] = item[:items ].append({id: params[:id]}) # adding a new item
end
mymodel.item_data = item
mymodel.save
return render :json => item, :status=> 200
rescue Exception => e
return render :json =>{:errors=>e.message}, :status=> 400
puts "ERROR: #{e.message}"
end
end
def delete_item
begin
mymodel = MyModel.find_by(id: params[:model_id])
if mymodel.present? && mymodel.item_data.present?
item = mymodel.item_data.deep_symbolize_keys
item[:items] = (item[:items].select { |itm| itm[:id] != params[:id] }) # remove an item
mymodel.item_data = item
mymodel.save
return render :json => item, :status=> 200
end
rescue Exception => e
return render :json =>{:errors=>e.message}, :status=> 400
puts "ERROR: #{e.message}"
end
end
end
I want my partial below to get the classy class from it being injected through locals, but I keep getting undefined method for classy.
//view
<%= render layout: "layouts/partial", locals: {className: "classy"} do %>
...
<% end>
//partial
<div class="regular-div <%=className if className?%>"></div>
To check if a local variable is set use local_assigns.has_key?(:some_key) or local_assigns[:some_key] to safely access the local variable.
A nifty way of handling the common task of building a list of classes is:
module ApplicationHelper
# Takes an array or list of classes and returns a string
# Example:
# class_list('a', 'b', nil, 'c')
# => "a b"
# class_list(['a', 'b', nil, 'c'])
# => "a b c"
def class_list(*classes)
[*classes].flatten.compact.join(' ')
end
end
Then you can do:
<div class="<%= class_list('regular-div', local_assigns[:className]) %>"></div>
This is my code for mysql like query:
def search
params.permit!
#query = params[:query]
respond_to do |format|
#outlet = Outlet.select(:name).where("name like ?","%#{#query}%")
format.json { render json: #outlet }
end
end
It renders all of my data from table. It does not respond to the query. Do you have any ideas?
My route is:
namespace :api do
resources :outlets, :defaults => { :format => 'json'}
get 'outlets/auto_complete' => 'outlets#auto_complete', :defaults => { :format => 'json'}
post 'outlets/search' => 'outlets#search', :defaults => { :format => 'json' }
end
The development.log is
Started POST "/api/outlets/search" for 127.0.0.1 at 2015-05-30 16:56:22 +0530
Processing by Api::OutletsController#search as JSON
Parameters: {"outlet"=>{"query"=>"life"}}
[1m[35mOutlet Load (0.1ms)[0m SELECT `outlets`.`name` FROM `outlets` WHERE (name like '%%')
Completed 200 OK in 28ms (Views: 22.3ms | ActiveRecord: 1.7ms)
Looking at the log file and below trace :-
Parameters: {"outlet"=>{"query"=>"life"}}
I found the issue. You need to do #query = params[:outlet][:query].
It is because params[:query] is nil, so the resulting sql is
where name like '%%'
You do have a query parameter in params[:outlet][:query] which you could use without changing your view.
However, as you're not creating or updating an Outlet, and query probably isn't an attribute of the Outlet model, it doesn't really make sense to structure the form in this way.
Try using form_tag instead of form_for and don't pass it an instance of Outlet. Also use text_field_tag instead of form.text_field. This way params[:query] will be set, instead of being wrapped under params[:outlet].
The new form would look a bit like this:
<%= form_tag do %>
<%= text_field_tag :query %>
<%= submit_tag %>
<% end %>
I am using SQL Korma to run some simple examples on a DB and am trying to convert this into JSON using Cheshire.
This works well when I have only 1 record returned but throws an error when I have more than 1 result.
Here are the 2 functions:
(defn get-room [id]
(first (select room
(where {:id id})
(limit 1))))
(defn get-rooms []
(select room))
and data:
(def x get-rooms)
(def y (get-room 1))
X is of type testproj.models.db:
(x)
=> [{:created_on "2014-04-05 13:19:47", :id 1, :description "Room 1"} {:created_on "2014-04-05 13:20:17", :id 2, :description "Room 2"} {:created_on "2014-04-05 13:20:20", :id 3, :description "Room 3"}]
Because y is a Hashmap:
(pr-str y)
=> "{:created_on \"2014-04-05 13:19:47\", :id 1, :description \"Room 1\"}"
Trying to convert to Json:
(cheshire.core/generate-string x)
JsonGenerationException Cannot JSON encode object of class: class testproj.models.db$get_rooms: testproj.models.db$get_rooms#507501ff cheshire.generate/generate (generate.clj:147)
(cheshire.core/generate-string y)
=> "{\"created_on\":\"2014-04-05 13:19:47\",\"id\":1,\"description\":\"Room 1\"}"
Why is korma returning different types based on the amount of records (this would help me understand this better) and secondly - how should I go about this?
It seems you're missing a function call. Try this:
(cheshire.core/generate-string (x))
i want to upload CSV files through the activeadmin panel.
on the index page from the resource "product" i want a button next to the "new product" button with "import csv file".
i dont know where to start.
in the documentation is something about collection_action, but with the code below i have no link at the top.
ActiveAdmin.register Post do
collection_action :import_csv, :method => :post do
# Do some CSV importing work here...
redirect_to :action => :index, :notice => "CSV imported successfully!"
end
end
anyone here who use activeadmin and can import csv data?
Continuing from Thomas Watsons great start to the answer which helped me get my bearings before figuring the rest of it out.
The code blow allows not just CSV upload for the example Posts model but for any subsequent models thereafter. all you need to do is copy the action_item ands both collection_actions from the example into any other ActiveAdmin.register block and the functionality will be the same. hope this helps.
app/admin/posts.rb
ActiveAdmin.register Post do
action_item :only => :index do
link_to 'Upload CSV', :action => 'upload_csv'
end
collection_action :upload_csv do
render "admin/csv/upload_csv"
end
collection_action :import_csv, :method => :post do
CsvDb.convert_save("post", params[:dump][:file])
redirect_to :action => :index, :notice => "CSV imported successfully!"
end
end
app/models/csv_db.rb
require 'csv'
class CsvDb
class << self
def convert_save(model_name, csv_data)
csv_file = csv_data.read
CSV.parse(csv_file) do |row|
target_model = model_name.classify.constantize
new_object = target_model.new
column_iterator = -1
target_model.column_names.each do |key|
column_iterator += 1
unless key == "ID"
value = row[column_iterator]
new_object.send "#{key}=", value
end
end
new_object.save
end
end
end
end
note: this example does a check to see whether or not the first column is an ID column, it then skips that column as rails will assign an ID to the new object (see example CSV below for reference)
app/views/admin/csv/upload_csv.html.haml
= form_for :dump, :url=>{:action=>"import_csv"}, :html => { :multipart => true } do |f|
%table
%tr
%td
%label{:for => "dump_file"}
Select a CSV File :
%td
= f.file_field :file
%tr
%td
= submit_tag 'Submit'
app/public/example.csv
"1","TITLE EXAMPLE","MESSAGE EXAMPLE","POSTED AT DATETIME"
"2","TITLE EXAMPLE","MESSAGE EXAMPLE","POSTED AT DATETIME"
"3","TITLE EXAMPLE","MESSAGE EXAMPLE","POSTED AT DATETIME"
"4","TITLE EXAMPLE","MESSAGE EXAMPLE","POSTED AT DATETIME"
"5","TITLE EXAMPLE","MESSAGE EXAMPLE","POSTED AT DATETIME"
note: quotations not always needed
Adding a collection_action does not automatically add a button linking to that action. To add a button at the top of the index screen you need to add the following code to your ActiveAdmin.register block:
action_item :only => :index do
link_to 'Upload CSV', :action => 'upload_csv'
end
But before calling the collection action you posted in your question, you first need the user to specify which file to upload. I would personally do this on another screen (i.e. creating two collection actions - one being a :get action, the other being your :post action). So the complete AA controller would look something like this:
ActiveAdmin.register Post do
action_item :only => :index do
link_to 'Upload posts', :action => 'upload_csv'
end
collection_action :upload_csv do
# The method defaults to :get
# By default Active Admin will look for a view file with the same
# name as the action, so you need to create your view at
# app/views/admin/posts/upload_csv.html.haml (or .erb if that's your weapon)
end
collection_action :import_csv, :method => :post do
# Do some CSV importing work here...
redirect_to :action => :index, :notice => "CSV imported successfully!"
end
end
#krhorst, I was trying to use your code, but unfortunately it sucks on big imports. It eat so much memory =( So I decided to use own solution based on activerecord-import gem
Here it is https://github.com/Fivell/active_admin_import
Features
Encoding handling
Support importing with ZIP file
Two step importing (see example2)
CSV options
Ability to prepend CSV headers automatically
Bulk import (activerecord-import)
Ability to customize template
Callbacks support
Support import from zip file
....
Based on ben.m's excellent answer above I replaced the csv_db.rb section suggested with this:
require 'csv'
class CsvDb
class << self
def convert_save(model_name, csv_data)
begin
target_model = model_name.classify.constantize
CSV.foreach(csv_data.path, :headers => true) do |row|
target_model.create(row.to_hash)
end
rescue Exception => e
Rails.logger.error e.message
Rails.logger.error e.backtrace.join("\n")
end
end
end
end
While not a complete answer I did not want my changes to pollute ben.m's answer in case I did something egregiously wrong.
expanding on ben.m's response which I found very useful.
I had issues with the CSV import logic (attributes not lining up and column iterator not functioning as required) and implemented a change which instead utilizes a per line loop and the model.create method. This allows you to import a .csv with the header line matching the attributes.
app/models/csv_db.rb
require 'csv'
class CsvDb
class << self
def convert_save(model_name, csv_data)
csv_file = csv_data.read
lines = CSV.parse(csv_file)
header = lines.shift
lines.each do |line|
attributes = Hash[header.zip line]
target_model = model_name.classify.constantize
target_model.create(attributes)
end
end
end
end
So your imported CSV file can look like this (use to match up with model attributes):
importExample.csv
first_name,last_name,attribute1,attribute2
john,citizen,value1,value2
For large excel which takes time on normal process, I created a gem that process Excel sheets using an active job and display results using action cable(websockets)
https://github.com/shivgarg5676/active_admin_excel_upload
Some of the solutions above worked pretty well. I ran into challenges in practice that I solved here below. The solved problems are:
Importing CSV data with columns in different orders
Preventing errors caused by hidden characters in Excel CSVs
Resetting the database primary_key so that the application can continue to add records after the import
Note: I took out the ID filter so I could change IDs for what I'm working on, but most use cases probably want to keep it in.
require 'csv'
class CsvDb
class << self
def convert_save(model_name, csv_data)
csv_file = csv_data.read
csv_file.to_s.force_encoding("UTF-8")
csv_file.sub!("\xEF\xBB\xBF", '')
target_model = model_name.classify.constantize
headers = csv_file.split("\n")[0].split(",")
CSV.parse(csv_file, headers: true) do |row|
new_object = target_model.new
column_iterator = -1
headers.each do |key|
column_iterator += 1
value = row[column_iterator]
new_object.send "#{key.chomp}=", value
end
new_object.save
end
ActiveRecord::Base.connection.reset_pk_sequence!(model_name.pluralize)
end
end
end