thinkingdata-ruby 1.2.0 → 2.0.0

data/lib/thinkingdata-ruby/logger_consumer.rb

@@ -1,57 +0,0 @@
-require 'logger'
-require 'thinkingdata-ruby/errors'
-
-module TDAnalytics
-  # 将数据写入本地文件, 需配合 LogBus 将数据上传到服务器
-  # 由于 LogBus 有完善的失败重传机制，因此建议用户首先考虑此方案
-  class LoggerConsumer
-    # LoggerConsumer 构造函数
-    #   log_path: 日志文件存放目录
-    #   mode: 日志文件切分模式，可选 daily/hourly
-    #   prefix: 日志文件前缀，默认为 'tda.log', 日志文件名格式为: tda.log.2019-11-15
-    def initialize(log_path='.', mode='daily', prefix:'tda.log')
-      case mode
-      when 'hourly'
-        @suffix_mode = '%Y-%m-%d-%H'
-      when 'daily'
-        @suffix_mode = '%Y-%m-%d'
-      else
-        raise IllegalParameterError.new("#{mode} is unsupported for LoggerConsumer. Replaced it by daily or hourly")
-      end
-
-      raise IllegalParameterError.new("prefix couldn't be empty") if prefix.nil? || prefix.length == 0
-
-      @current_suffix = Time.now.strftime(@suffix_mode)
-
-      @full_prefix = "#{log_path}/#{prefix}."
-
-      _reset
-    end
-
-    def add(msg)
-      unless Time.now.strftime(@suffix_mode) == @current_suffix
-        @logger.close
-        @current_suffix = Time.now.strftime(@suffix_mode)
-        _reset
-      end
-      @logger.info(msg.to_json)
-    end
-  
-    # 关闭 logger
-    def close
-      @logger.close
-    end
-
-    private
-
-    # 重新创建 logger 对象. LogBus 判断新文件会同时考虑文件名和 inode，因此默认的切分方式会导致数据重传
-    def _reset
-      @logger = Logger.new("#{@full_prefix}#{@current_suffix}")
-      @logger.level = Logger::INFO
-      @logger.formatter = proc do |severity, datetime, progname, msg|
-        "#{msg}\n"
-      end
-    end
-
-  end
-end

data/lib/thinkingdata-ruby/tracker.rb

@@ -1,402 +0,0 @@
-require 'securerandom'
-require 'thinkingdata-ruby/errors'
-require 'thinkingdata-ruby/version'
-
-module TDAnalytics
-  # TDAnalytics::Tracker 是数据上报的核心类，使用此类上报事件数据和更新用户属性.
-  # 创建 Tracker 类需要传入 consumer 对象，consumer 决定了如何处理格式化的数据（存储在本地日志文件还是上传到服务端).
-  #
-  #   ta = TDAnalytics::Tracker.new(consumer)
-  #   ta.track('your_event', distinct_id: 'distinct_id_of_user')
-  #
-  # TDAnalytics 提供了三种 consumer 实现:
-  #   LoggerConsumer: 数据写入本地文件
-  #   DebugConsumer: 数据逐条、同步的发送到服务端，并返回详细的报错信息
-  #   BatchConsumer: 数据批量、同步的发送到服务端
-  #
-  # 您也可以传入自己实现的 Consumer，只需实现以下接口:
-  #   add(message): 接受 hash 类型的数据对象
-  #   flush: (可选) 将缓冲区的数据发送到指定地址
-  #   close: (可选) 程序退出时用户可以主动调用此接口以保证安全退出
-  class Tracker
-
-    LIB_PROPERTIES = {
-      '#lib' => 'ruby',
-      '#lib_version' => TDAnalytics::VERSION,
-    }
-
-    # SDK 构造函数，传入 consumer 对象
-    #
-    # 默认情况下，除参数不合法外，其他 Error 会被忽略，如果您希望自己处理接口调用中的 Error，可以传入自定义的 error handler.
-    # ErrorHandler 的定义可以参考 thinkingdata-ruby/errors.rb
-    #
-    # uuid 如果为 true，每条数据都会被带上随机 UUID 作为 #uuid 属性的值上报，该值不会入库，仅仅用于后台做数据重复检测
-    def initialize(consumer, error_handler = nil, uuid: false)
-      @error_handler = error_handler || ErrorHandler.new
-      @consumer = consumer
-      @super_properties = {}
-      @uuid = uuid
-    end
-
-    # 设置公共事件属性，公共事件属性是所有事件都会带上的属性. 此方法会将传入的属性与当前公共属性合并.
-    # 如果希望跳过本地格式校验，可以传入值为 true 的 skip_local_check 参数
-    def set_super_properties(properties, skip_local_check = false)
-      unless skip_local_check || _check_properties(:track, properties)
-        @error_handler.handle(IllegalParameterError.new("Invalid super properties"))
-        return false
-      end
-      properties.each do |k, v|
-        if v.is_a?(Time)
-          @super_properties[k] = _format_time(v)
-        else
-          @super_properties[k] = v
-        end
-      end
-    end
-
-    # 清除公共事件属性
-    def clear_super_properties
-      @super_properties = {}
-    end
-
-    # 上报事件. 每个事件都包含一个事件名和 Hash 对象的时间属性. 其参数说明如下:
-    #   event_name: (必须) 事件名 必须是英文字母开头，可以包含字母、数字和 _, 长度不超过 50 个字符.
-    #   distinct_id: (可选) 访客 ID
-    #   account_id: （可选) 账号ID distinct_id 和 account_id 不能同时为空
-    #   properties: （可选) Hash 事件属性。支持四种类型的值：字符串、数值、Time、boolean
-    #   time: （可选）Time 事件发生时间，如果不传默认为系统当前时间
-    #   ip: (可选) 事件 IP，如果传入 IP 地址，后端可以通过 IP 地址解析事件发生地点
-    #   skip_local_check: (可选) boolean 表示是否跳过本地检测
-    def track(event_name: nil, distinct_id: nil, account_id: nil, properties: {}, time: nil, ip: nil,first_check_id:nil, skip_local_check: false)
-      begin
-        _check_name event_name
-        _check_id(distinct_id, account_id)
-        unless skip_local_check
-          _check_properties(:track, properties)
-        end
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      data = {}
-      data[:event_name] = event_name
-      data[:distinct_id] = distinct_id if distinct_id
-      data[:account_id] = account_id if account_id
-      data[:time] = time if time
-      data[:ip] = ip if ip
-      data[:first_check_id] = first_check_id if first_check_id
-      data[:properties] = properties
-
-      _internal_track(:track, data)
-    end
-
-    #   上报事件数据可进行更新. 每个事件都包含一个事件名和事件ID以及 Hash 对象的时间属性. 其参数说明如下:
-    #   event_name: (必须) 事件名 必须是英文字母开头，可以包含字母、数字和 _, 长度不超过 50 个字符.
-    #   event_id:(必须) event_name + event_id 会作为一条事件的唯一键
-    #   distinct_id: (可选) 访客 ID
-    #   account_id: （可选) 账号ID distinct_id 和 account_id 不能同时为空
-    #   properties: （可选) Hash 事件属性。支持四种类型的值：字符串、数值、Time、boolean
-    #   time: （可选）Time 事件发生时间，如果不传默认为系统当前时间
-    #   ip: (可选) 事件 IP，如果传入 IP 地址，后端可以通过 IP 地址解析事件发生地点
-    #   skip_local_check: (可选) boolean 表示是否跳过本地检测
-    def track_overwrite(event_name: nil,event_id: nil, distinct_id: nil, account_id: nil, properties: {}, time: nil, ip: nil, skip_local_check: false)
-      begin
-        _check_name event_name
-        _check_event_id event_id
-        _check_id(distinct_id, account_id)
-        unless skip_local_check
-          _check_properties(:track_overwrite, properties)
-        end
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      data = {}
-      data[:event_name] = event_name
-      data[:event_id] = event_id
-      data[:distinct_id] = distinct_id if distinct_id
-      data[:account_id] = account_id if account_id
-      data[:time] = time if time
-      data[:ip] = ip if ip
-      data[:properties] = properties
-      _internal_track(:track_overwrite, data)
-    end
-
-
-
-    #   上报事件数据可进行覆盖. 每个事件都包含一个事件名和事件ID以及 Hash 对象的时间属性. 其参数说明如下:
-    #   event_name: (必须) 事件名 必须是英文字母开头，可以包含字母、数字和 _, 长度不超过 50 个字符.
-    #   event_id:(必须) event_name + event_id 会作为一条事件的唯一键
-    #   distinct_id: (可选) 访客 ID
-    #   account_id: （可选) 账号ID distinct_id 和 account_id 不能同时为空
-    #   properties: （可选) Hash 事件属性。支持四种类型的值：字符串、数值、Time、boolean
-    #   time: （可选）Time 事件发生时间，如果不传默认为系统当前时间
-    #   ip: (可选) 事件 IP，如果传入 IP 地址，后端可以通过 IP 地址解析事件发生地点
-    #   skip_local_check: (可选) boolean 表示是否跳过本地检测
-    def track_update(event_name: nil,event_id: nil, distinct_id: nil, account_id: nil, properties: {}, time: nil, ip: nil, skip_local_check: false)
-      begin
-        _check_name event_name
-        _check_event_id event_id
-        _check_id(distinct_id, account_id)
-        unless skip_local_check
-          _check_properties(:track_update, properties)
-        end
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      data = {}
-      data[:event_name] = event_name
-      data[:event_id] = event_id
-      data[:distinct_id] = distinct_id if distinct_id
-      data[:account_id] = account_id if account_id
-      data[:time] = time if time
-      data[:ip] = ip if ip
-      data[:properties] = properties
-      _internal_track(:track_update, data)
-    end
-
-    # 设置用户属性. 如果出现同名属性，则会覆盖之前的值.
-    #   distinct_id: (可选) 访客 ID
-    #   account_id: （可选) 账号ID distinct_id 和 account_id 不能同时为空
-    #   properties: （可选) Hash 用户属性。支持四种类型的值：字符串、数值、Time、boolean
-    def user_set(distinct_id: nil, account_id: nil, properties: {}, ip: nil)
-      begin
-        _check_id(distinct_id, account_id)
-        _check_properties(:user_set, properties)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_set,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-                      properties: properties,
-                      ip: ip,
-      )
-    end
-
-    # 设置用户属性. 如果有重名属性，则丢弃, 参数与 user_set 相同
-    def user_set_once(distinct_id: nil, account_id: nil, properties: {}, ip: nil)
-      begin
-        _check_id(distinct_id, account_id)
-        _check_properties(:user_setOnce, properties)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_setOnce,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-                      properties: properties,
-                      ip: ip,
-      )
-    end
-
-    # 追加用户的一个或多个列表类型的属性
-    def user_append(distinct_id: nil, account_id: nil, properties: {})
-      begin
-        _check_id(distinct_id, account_id)
-        _check_properties(:user_append, properties)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_append,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-                      properties: properties,
-                      )
-    end
-
-    # 删除用户属性, property 可以传入需要删除的用户属性的 key 值，或者 key 值数组
-    def user_unset(distinct_id: nil, account_id: nil, property: nil)
-      properties = {}
-      if property.is_a?(Array)
-        property.each do |k|
-          properties[k] = 0
-        end
-      else
-        properties[property] = 0
-      end
-
-      begin
-        _check_id(distinct_id, account_id)
-        _check_properties(:user_unset, properties)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_unset,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-                      properties: properties,
-      )
-    end
-
-    # 累加用户属性, 如果用户属性不存在，则会设置为 0，然后再累加
-    #   distinct_id: (可选) 访客 ID
-    #   account_id: （可选) 账号ID distinct_id 和 account_id 不能同时为空
-    #   properties: （可选) Hash 数值类型的用户属性
-    def user_add(distinct_id: nil, account_id: nil, properties: {})
-      begin
-        _check_id(distinct_id, account_id)
-        _check_properties(:user_add, properties)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_add,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-                      properties: properties,
-      )
-    end
-
-    # 删除用户，用户之前的事件数据不会被删除
-    def user_del(distinct_id: nil, account_id: nil)
-      begin
-        _check_id(distinct_id, account_id)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        return false
-      end
-
-      _internal_track(:user_del,
-                      distinct_id: distinct_id,
-                      account_id: account_id,
-      )
-    end
-
-    # 立即上报数据，对于 BatchConsumer 会触发上报
-    def flush
-      return true unless defined? @consumer.flush
-      ret = true
-      begin
-        @consumer.flush
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        ret = false
-      end
-      ret
-    end
-
-    # 退出前调用，保证 Consumer 安全退出
-    def close
-      return true unless defined? @consumer.close
-      ret = true
-      begin
-        @consumer.close
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        ret = false
-      end
-      ret
-    end
-
-    private
-
-    # 出现异常的时候返回 false, 否则 true
-    def _internal_track(type, properties: {}, event_name: nil, event_id:nil, account_id: nil, distinct_id: nil, ip: nil,first_check_id: nil, time: Time.now)
-      if account_id == nil && distinct_id == nil
-        raise IllegalParameterError.new('account id or distinct id must be provided.')
-      end
-
-      if type == :track || type == :track_update || type == :track_overwrite
-        raise IllegalParameterError.new('event name is empty') if event_name == nil
-        properties = {'#zone_offset': time.utc_offset / 3600.0}.merge(LIB_PROPERTIES).merge(@super_properties).merge(properties)
-      end
-
-      # 格式化 Time 类型
-      properties.each do |k, v|
-        if v.is_a?(Time)
-          properties[k] = _format_time(v)
-        end
-      end
-
-      data = {
-          '#type' => type,
-          '#time' => _format_time(time),
-          'properties' => properties,
-      }
-
-      data['#event_name'] = event_name if (type == :track || type == :track_update || :track_overwrite)
-      data['#event_id'] = event_id if (type == :track_update || type == :track_overwrite)
-      data['#account_id'] = account_id if account_id
-      data['#distinct_id'] = distinct_id if distinct_id
-      data['#ip'] = ip if ip
-      data['#first_check_id'] = first_check_id if first_check_id
-      data['#uuid'] = SecureRandom.uuid if @uuid
-
-      ret = true
-      begin
-        @consumer.add(data)
-      rescue TDAnalyticsError => e
-        @error_handler.handle(e)
-        ret = false
-      end
-
-      ret
-    end
-
-    # 将 Time 类型格式化为数数指定格式的字符串
-    def _format_time(time)
-      time.strftime("%Y-%m-%d %H:%M:%S.#{((time.to_f * 1000.0).to_i % 1000).to_s.rjust(3, "0")}")
-    end
-
-    def _check_event_id(event_id)
-      raise IllegalParameterError.new("the event_id or property cannot be nil") if event_id.nil?
-      true
-    end
-
-    # 属性名或者事件名检查
-    def _check_name(name)
-      raise IllegalParameterError.new("the name of event or property cannot be nil") if name.nil?
-
-      unless name.instance_of?(String) || name.instance_of?(Symbol)
-        raise IllegalParameterError.new("#{name} is invalid. It must be String or Symbol")
-      end
-      true
-    end
-
-    # 属性类型检查
-    def _check_properties(type, properties)
-      unless properties.instance_of? Hash
-        return false
-      end
-
-      properties.each do |k, v|
-        _check_name k
-        next if v.nil?
-        unless v.is_a?(Integer) || v.is_a?(Float) || v.is_a?(Symbol) || v.is_a?(String) || v.is_a?(Time) || !!v == v || v.is_a?(Array)
-          raise IllegalParameterError.new("The value of properties must be type in Integer, Float, Symbol, String, Array,and Time")
-        end
-
-        if type == :user_add
-          raise IllegalParameterError.new("Property value for user add must be numbers") unless v.is_a?(Integer) || v.is_a?(Float)
-        end
-        if v.is_a?(Array)
-          v.each_index do |i|
-            if v[i].is_a?(Time)
-              v[i] = _format_time(v[i])
-            end
-          end
-        end
-      end
-      true
-    end
-
-    # 检查用户 ID 合法性
-    def _check_id(distinct_id, account_id)
-      raise IllegalParameterError.new("account id or distinct id must be provided.") if distinct_id.nil? && account_id.nil?
-    end
-  end
-end

data/lib/thinkingdata-ruby/version.rb

@@ -1,3 +0,0 @@
-module TDAnalytics
-  VERSION = '1.2.0'
-end