Activerecord找不到使用CSV信息创建的记录

时间:2011-07-29 16:45:22

标签: ruby-on-rails ruby activerecord csv fastercsv

当我从CSV文件创建记录时,Activerecord无法使用Record#find_by找到它们。如果我通过rails控制台创建一个Record,它会按预期工作。这是我通过CSV创建的代码:

def create
  file = params[:record][:file].tempfile

  CSV.foreach file, :headers => true, :header_converters => :symbol do |row|
    Record.create row.to_hash
  end

  redirect_to records_url
end

以下是控制台的一些示例:

> Record.find_by_email "matching@asd.asd" 
=> nil     # Should return record created through the CSV file

> Record.create :email => "matching@asd.asd"
> Record.find_by_email "matching@asd.asd"
=> <Record id: 4, email: "matching@asd.asd">
> Record.find_all_by_email "matching@asd.asd"
=> [<Record id: 4, email: "matching@asd.asd">]    # Should return both

任何帮助将不胜感激。如果重要的话我就在Rails 3.1rc5上。

-

已更新为rows.to_hash输出

根据要求,调试rows.to_hash的输出:

{:email=>"Matching@asd.asd", :first_name=>"First", :last_name=>"Last"}
{:email=>"Notmatching@asd.asd", :first_name=>"Matching", :last_name=>"Name"}
{:email=>"asdasdasdasd@asd.asd", :first_name=>"asd", :last_name=>"asd"}

来自控制台的另一个例子,这进一步加剧了我的困惑:

> Record.find 14    # Record created by CSV
=> <Record id: 14, first_name: "First", last_name: "Last", email: "Matching@asd.asd", created_at: "2011-07-29 18:03:25", updated_at: "2011-07-29 18:03:25">
> Record.find(14).email
=> "Matching@asd.asd"
> Record.find_by_email Record.find(14).email
=> nil

-

更新了SQL输出

Record.find_by_email Record.find(14).email生成的SQL:

Record Load (0.3ms)  SELECT "records".* FROM "records" WHERE "records"."id" = ? LIMIT 1  [["id", 14]]
Record Load (0.3ms)  SELECT "records".* FROM "records" WHERE "records"."email" = 'Matching@asd.asd' LIMIT 1

-

试用SQLite控制台

sqlite> SELECT "records".* FROM "records" WHERE "records"."email" = 'Matching@asd.asd' LIMIT 1;
# This one should have returned the CSV record, but it returns nothing
sqlite> SELECT "records".* FROM "records" WHERE "records"."email" = 'nomatch@asd.asd' LIMIT 1;
5|2|match|this|nomatch@asd.asd|2011-07-29 17:13:13.821972|2011-07-29 17:13:13.821972

-

添加型号代码,查询结果,CSV输入

可能是人类已知的最激动人心的模型:

class Record < ActiveRecord::Base
  attr_accessible :email, :first_name, :last_name, :file
  attr_accessor :file
end

查询的更多结果:

sqlite> select * from records;
5|2|match|this|nomatch@asd.asd|2011-07-29 17:13:13.821972|2011-07-29 17:13:13.821972
9|3|first|last||2011-07-29 17:56:50.471766|2011-07-29 17:56:50.471766
10|6|first|last||2011-07-29 17:56:54.917432|2011-07-29 17:56:54.917432
17||First|Last|Matching@asd.asd|2011-07-29 19:43:23.843188|2011-07-29 19:43:23.843188
18||Matching|Name|Notmatching@asd.asd|2011-07-29 19:43:23.849001|2011-07-29 19:43:23.849001
19||asd|asd|asdasdasdasd@asd.asd|2011-07-29 19:43:23.852037|2011-07-29 19:43:23.852037

为了更好地衡量,测试CSV文件:

email,first name,last name
Matching@asd.asd,First,Last
Notmatching@asd.asd,Matching,Name
asdasdasdasd@asd.asd,asd,asd

2 个答案:

答案 0 :(得分:1)

这是你的问题。从attr_*模型中删除所有Record行。默认情况下,可以访问所有ActiveRecord个数据库字段。通过添加attr字段,您实际上会覆盖默认的Rails访问器,这就是为什么一切都表现得很奇怪。

要学习的课程:当你认为你的模型中“没什么有趣的”时......非常可疑:)祝你好运。删除这些行应该可以解决所有问题。

答案 1 :(得分:0)

原来是一个编码问题。

电子邮件,名字和姓氏作为blob被抛入SQLite,我猜这会导致一两个问题。添加:encoding => 'u'作为CSV.foreach的选项可以解决所有问题。