了解AWS DynamoDB中的属性与Ruby

7
我似乎无法理解 AWS Ruby SDK 文档中 DynamoDB 的概念(或更具体地说是 DynamoDB 数据模型的概念)。
具体来说,我一直在阅读:http://docs.aws.amazon.com/AWSRubySDK/latest/frames.html#!AWS/DynamoDB.html 请注意:我也已经阅读了数据模型文档,但仍然不理解;我希望能够通过 Ruby 中的一个正确示例来澄清我的困惑。
在下面的代码片段中,我创建了一个名为“my_books”的表,它具有名为“item_id”的主键,并且它是一个 Hash 键(而不是一个 Hash/Range 组合)...
dyn = AWS::DynamoDB::Client::V20120810.new
# => #<AWS::DynamoDB::Client::V20120810>

dyn.create_table({
  :attribute_definitions => [
    { :attribute_name => "item_id", :attribute_type => "N" }
  ],
  :table_name => "my_books",
  :key_schema => [
    { :attribute_name => "item_id", :key_type => "HASH" },
  ],
  :provisioned_throughput => {
    :read_capacity_units  => 10,
    :write_capacity_units => 10
  }
})
# => {:table_description=>{:attribute_definitions=>[{:attribute_name=>"item_id", :attribute_type=>"N"}], :table_name=>"my_books", :key_schema=>[{:attribute_name=>"item_id", :key_type=>"HASH"}], :table_status=>"ACTIVE", :creation_date_time=>2014-11-24 16:59:47 +0000, :provisioned_throughput=>{:number_of_decreases_today=>0, :read_capacity_units=>10, :write_capacity_units=>10}, :table_size_bytes=>0, :item_count=>0}}

dyn.list_tables
# => {:table_names=>["my_books"]}

dyn.scan :table_name => "my_books"
# => {:member=>[], :count=>0, :scanned_count=>0}

我接下来会尝试添加一个新的项目到表格中。我的理解是,我需要为item_id(这是主键)指定数值,然后再为我要添加到表格中的新项目/记录/文档指定其他属性...

dyn.put_item(
  :table_name => "my_books",
  :item => {
    "item_id" => 1,
    "item_title" => "My Book Title",
    "item_released" => false
  }
)

但是最后一条命令返回以下错误:

expected hash value for value at key item_id of option item

虽然我不太明白哈希值由什么生成,但我尝试进行如下操作:

dyn.put_item(
  :table_name => "my_books",
  :item => {
    "item_id" => { "N" => 1 },
    "item_title" => "My Book Title",
    "item_released" => false
  }
)

但现在它返回以下错误...
expected string value for key N of value at key item_id of option item

我尝试了不同的变化,但似乎无法弄清楚这是如何工作的?


编辑/更新: 根据Uri Agassi的建议 - 我将值从1更改为"1"。我不太确定为什么这必须被引用,因为我已经定义类型为数字而不是字符串,但好吧,让我们接受这个并继续。


尝试 ..."item_id" => { "N" => "1" },... - 根据错误信息,期望一个字符串值。 - Uri Agassi
1个回答

6
我终于弄清楚了大部分我需要了解的DynamoDB数据模型和使用Ruby SDK的内容。
以下是我的示例代码,希望能帮助其他人,我在这里有一个完整的示例:https://gist.github.com/Integralist/9f9f2215e001b15ac492#file-3-dynamodb-irb-session-rb
# https://github.com/BBC-News/alephant-harness can automate the below set-up when using Spurious
# API Documentation http://docs.aws.amazon.com/amazondynamodb/latest/APIReference/API_Operations.html
# Ruby SDK API Documentation http://docs.aws.amazon.com/AWSRubySDK/latest/frames.html#!AWS/DynamoDB/Client/V20120810.html

require "aws-sdk"
require "dotenv"
require "spurious/ruby/awssdk/helper"

Spurious::Ruby::Awssdk::Helper.configure
# => <AWS::Core::Configuration>

Dotenv.load(
  File.join(
    File.dirname(__FILE__), "config", "development", "env.yaml"
  )
)
# => {"AWS_REGION"=>"eu-west-1", "AWS_ACCESS_KEY_ID"=>"development_access", "AWS_SECRET_ACCESS_KEY"=>"development_secret", "DYNAMO_LU"=>"development_lookup", "DYNAMO_SQ"=>"development_sequence", "SQS_QUEUE"=>"development_queue", "S3_BUCKET"=>"development_bucket"}

dyn = AWS::DynamoDB::Client.new :api_version => "2012-08-10"
dyn = AWS::DynamoDB::Client::V20120810.new
# => #<AWS::DynamoDB::Client::V20120810>

dyn.create_table({
  # This section requires us to define our primary key 
  # Which will be called "item_id" and it must be a numerical value
  :attribute_definitions => [
    { :attribute_name => "item_id", :attribute_type => "N" }
  ],
  :table_name => "my_books",
  # The primary key will be a simple Hash key (not a Hash/Range which requires both key types to be provided)
  # The attributes defined above must be included in the :key_schema Array
  :key_schema => [
    { :attribute_name => "item_id", :key_type => "HASH" }
  ],
  :provisioned_throughput => {
    :read_capacity_units  => 10,
    :write_capacity_units => 10
  }
})
# => {:table_description=>{:attribute_definitions=>[{:attribute_name=>"item_id", :attribute_type=>"N"}], :table_name=>"my_books", :key_schema=>[{:attribute_name=>"item_id", :key_type=>"HASH"}], :table_status=>"ACTIVE", :creation_date_time=>2014-11-24 16:59:47 +0000, :provisioned_throughput=>{:number_of_decreases_today=>0, :read_capacity_units=>10, :write_capacity_units=>10}, :table_size_bytes=>0, :item_count=>0}}

dyn.list_tables
# => {:table_names=>["my_books"]}

dyn.scan :table_name => "my_books"
# => {:member=>[], :count=>0, :scanned_count=>0}

dyn.put_item(
  :table_name => "my_books",
  :item => {
    "item_id" => { "N" => "1" }, # oddly this needs to be a String and not a strict Integer?
    "item_title" => { "S" => "My Book Title"},
    "item_released" => { "B" => "false" }
  }
)
# Note: if you use an "item_id" that already exists, then the item will be updated.
#       Unless you use the "expected" conditional feature

dyn.put_item(
  :table_name => "my_books",
  :item => {
    "item_id" => { "N" => "1" }, # oddly this needs to be a String and not a strict Integer?
    "item_title" => { "S" => "My Book Title"},
    "item_released" => { "B" => "false" }
  },
  # The :expected key specifies the conditions of our "put" operation.
  # If "item_id" isn't NULL (i.e. it exists) then our condition has failed.
  # This means we only write the value when the key "item_id" hasn't been set.
  :expected => {
    "item_id" => { :comparison_operator => "NULL" }
  }
)
# AWS::DynamoDB::Errors::ConditionalCheckFailedException: The conditional check failed

dyn.scan :table_name => "my_books"
# => {:member=>[{"item_id"=>{:n=>"1"}, "item_title"=>{:s=>"My Book Title"}, "item_released"=>{:b=>"false"}}], :count=>1, :scanned_count=>1}

dyn.query :table_name => "my_books", :consistent_read => true, :key_conditions => {
  "item_id" => {
    :comparison_operator => "EQ",
    :attribute_value_list => [{ "n" => "1" }]
  },
  "item_title" => {
    :comparison_operator => "EQ",
    :attribute_value_list => [{ "s" => "My Book Title" }]
  }
}
# => {:member=>[{"item_id"=>{:n=>"1"}, "item_title"=>{:s=>"My Book Title"}, "item_released"=>{:b=>"false"}}], :count=>1, :scanned_count=>1}

dyn.query :table_name => "my_books", 
  :consistent_read => true, 
  :select => "SPECIFIC_ATTRIBUTES",
  :attributes_to_get => ["item_title"],
  :key_conditions => {
  "item_id" => {
    :comparison_operator => "EQ",
    :attribute_value_list => [{ "n" => "1" }]
  },
  "item_title" => {
    :comparison_operator => "EQ",
    :attribute_value_list => [{ "s" => "My Book Title" }]
  }
}
# => {:member=>[{"item_title"=>{:s=>"My Book Title"}}], :count=>1, :scanned_count=>1}

dyn.delete_item(
  :table_name => "my_books",
  :key => {
    "item_id" => { "n" => "1" }
  }
)
# => {:member=>[], :count=>0, :scanned_count=>0}

谢谢,你救了我的一天。 - Guillaume

网页内容由stack overflow 提供, 点击上面的
可以查看英文原文,
原文链接