csv_seed 0.0.4

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 61ca14d9500c412955f6f5c8553ade681bcaec17
+  data.tar.gz: a105521118230c11eea2c66be1f788fc63c950bb
+SHA512:
+  metadata.gz: c00d8d9934a4af33932d161346de46b88aa6372057ccc476935a9dd14f556343e3e17b95760412d200293e69dd9ef6831871026f7ec1ccb9ecaa6ebe148de3ac
+  data.tar.gz: 681a45c8f395f312159116a2fcbc231ae6e1ad4e143ffaa4554f70cfdc325ea1d33446a093a7be53b1ba4f9001792101237d34be1b0efe8bd2b9728b6e3681eb

data/README.md

@@ -0,0 +1,8 @@
+# csv_seed
+Import csv data to projects which are dependant on ActiveRecord.
+
+Why a new seed gem?
+- Csv is easier to read and update than fixtures or seeds.
+- Associations in csv tables can be applied to real databases.
+
+...

data/lib/csv_seed/importer.rb

@@ -0,0 +1,67 @@
+module CsvSeed
+  class Import
+    attr_accessor :name, :tables
+
+    def initialize(name)
+      @name = (name || '')
+    end
+
+    # thor csv:seed --use points --except Activity ActivityForm
+    def run(only, except)
+      # TODO 检查参数
+      @tables = read_csv
+      say_end = []
+      models = only.present? ? only : (@tables.keys - (except or []))
+      run_tables = @tables.values.select {|t| models.include? t.model}
+      # @tables.keep_if {|k| models.include? k}
+      puts "\n\n*** run_tables: #{run_tables.map(&:model).join(', ')}\n\n\n"
+      ActiveRecord::Base.transaction do
+        run_tables.each do |table|
+          say_end += table.execute_commands("destroy_all")
+        end
+        run_tables.each do |table|
+          say_end += table.execute_commands
+        end
+      end
+
+      puts "\n"*3
+      say_end.each {|say| puts say}
+    end
+
+    def read_csv()
+      tables = {}
+      current_model = ""
+      Dir.glob(Rails.root.join("db","seeds", @name, "*.csv")).each do |csv|
+        lines = CSV.read(csv)
+        lines.each do |line|
+          next if line[0].nil?
+
+          if line[0][0] == "#"
+            cmd, *args = *line[0][1..-1].split('#')
+            case cmd
+            when "model"
+              current_model = args.first
+              tables[current_model] = Table.new(current_model, tables)
+            when "command"
+              tables[current_model].commands << {cmd: args.first, args: args[1..-1]}
+            when "key"
+              tables[current_model].key = args.first
+            when "uploader"
+              tables[current_model].upload_field_names = args
+            end
+            next
+          end
+
+          # Imported tables should have the 'id' attribute.
+          if line[0] == 'id'
+            tables[current_model].field_names = line.compact[1..-1]
+            next
+          end
+          # 根据字段表长来截，compact会吞掉内容为空的字段。注意这儿有id。
+          tables[current_model].add_record line[0..tables[current_model].field_names.size]
+        end
+      end
+      tables
+    end
+  end
+end

data/lib/csv_seed/record.rb

@@ -0,0 +1,73 @@
+module CsvSeed
+  class Record
+    attr_accessor :table, :id, :real_id, :content
+
+    def initialize(table, attributes_line)
+      @table    = table
+      @id       = attributes_line[0]
+      @content  = Hash[@table.field_names.zip attributes_line[1..-1]]
+    end
+
+    def insert
+      return if @real_id.present?
+
+      # 先去掉外键和上传字段，外键后面统一处理
+      # @table.foreign_keys.keys +
+      params = @content.except *(@table.upload_field_names)
+      # 如果有enum字段，事先要先转为integer
+      params = do_type_cast(params) if @table.table_class.defined_enums.present?
+      r = @table.table_class.new params
+      @table.upload_field_names.each do |field|
+        File.open(Uploader.file_path(@content[field])) do |f|
+          # r[field] = f
+          # 上面这句无论如何不行，没有错也没上传。这是直接赋值，但是 carrierwave 应该改了这个赋值方法。
+          r.send field+'=', f
+        end
+      end
+      print "*** insert model: #{@table.model}, params: #{params}"
+      r.save!
+      # byebug if @table.model == 'Gift'
+      @real_id = r.id
+      @table.count += 1
+      puts ", *** real_id: #{@real_id}\n"
+
+      @table.foreign_keys.each do |k, m|
+        pk = @table.primary_keys[k]
+        puts "              model: #{@table.model}, foreign_key: #{k} -> #{m}, fk_content: #{@content[k]}, pk: #{pk}"
+        next if @content[k].nil? # 外键为空
+        next if pk != 'id' && pk != 'polymorphic' # 关系表主键不是 'id' 并且不是 polymorphic
+        m = @content[k.gsub('_id', '_type')] if pk == 'polymorphic'
+        f_record = @table.tables[m].find(@content[k])
+        f_record.insert if !f_record.real_id
+        puts "*** linked #{k}: #{f_record.real_id}"
+        r[k] = f_record.real_id
+      end
+      r.save!
+    end
+
+    # #command#update#2#2#code
+    def update(k, keys, excepts)
+      puts "*** [Record#update] params: #{k}, #{keys}, #{excepts}"
+      condition = @content.select {|k,v| keys.split(',').include? k}
+      obj = @table.table_class.where(condition).first
+      if obj
+        content = excepts.present? ? @content.reject {|k,v| excepts.split(',').include? k} : @content
+        obj.update! content
+        puts "*** id: #{k}, obj_id:#{obj.id}, condition: #{condition}, a record of #{@table.model} has been updated.\n"
+        true
+      else
+        false
+      end
+    end
+
+    private
+    def do_type_cast(params)
+      result = params.clone
+      @table.table_class.defined_enums.keys.each do |field|
+        # nil表示缺省，缺省并不一定是0。nil.to_i == 0
+        result[field] = params[field].to_i if params[field]
+      end
+      result
+    end
+  end
+end

data/lib/csv_seed/table.rb

@@ -0,0 +1,110 @@
+module CsvSeed
+  class Table
+    attr_accessor :model, :field_names, :foreign_keys, :primary_keys, :records, :count, :commands, :tables, :key, :upload_field_names
+
+    def initialize(model, tables)
+      @model = model
+      @count = 0
+      keys = table_class.reflect_on_all_associations(:belongs_to)
+      @foreign_keys = Hash[keys.map {|k| [k.foreign_key, k.class_name]}]
+      # polymorphic 的时候 association_primary_key 会出错。
+      @primary_keys = Hash[keys.map {|k| [k.foreign_key, k.polymorphic? ? 'polymorphic' : k.association_primary_key]}]
+      @tables = tables
+      @commands = []
+      @field_names = []
+      @upload_field_names = []
+      @key = 'id'
+      @key_h = {}
+    end
+
+    def table_class
+      # puts "*** table_class's model: #{@model}"
+      @model.constantize
+    end
+
+    def add_record(attributes_line)
+      @records ||= {}
+      r = Record.new(self, attributes_line)
+      pk = r.id
+      # 如果有key的时候，这张表一般是用作指示的
+      if @key != 'id'
+        r.real_id = table_class.where(get_params_from_key(r)).first.try(:id)
+        puts "*** [Table#add_record] Model: #{@model}, key: #{@key} -- id: #{r.id}, real_id: #{r.real_id}"
+        puts "*** [Table#add_record] sql: #{table_class.where(get_params_from_key(r)).to_sql}"
+        pk = r.content[@key]
+      end
+      @records[r.id] = r
+      @key_h[pk] = r
+    end
+
+    # 根据 id 或 key 来找
+    def find(k)
+      puts "*** [Table#find] Model: #{@model}, k: #{k}"
+      @records[k] || @key_h[k]
+    end
+
+    def get_params_from_key(record)
+      return {id: 0} if @key == 'id'
+      @key.split(',').each_with_object({}) do |attr_name, h|
+        # 如果外键里有 id 的话，按这个找是找不到记录的。除非是 code 才能找到。
+        h[attr_name.to_sym] = record.content[attr_name]
+      end
+    end
+
+    def execute_commands(only_command = nil)
+      say_later = []
+      done = []
+      @commands.each do |item|
+        cmd, args = *item.values
+        next if only_command.present? and only_command != cmd
+        log_args = args.present? ? " args: #{args}" : ''
+        puts "*** #{self.model}.#{cmd}#{log_args}\n"
+        if cmd == 'destroy_all'
+          say_later << "*** #{self.table_class.count} records have been deleted from #{self.model}\n"
+          self.table_class.destroy_all
+          @records.values.each {|r| r.real_id = nil}
+          done << item
+          next
+        elsif cmd == 'create_all'
+          self.records.values.each {|r| r.insert}
+          done << item
+          say_later << "*** #{self.count} records have been imported into model #{self.model}. \n"
+          next
+        end
+
+        if %w(append delete update).include? cmd
+          # keys 是查询条件包含的字段表列表，以逗号分隔。当cmd为append的时候没有
+          ibegin, iend, keys, excepts = args[0].to_i, args[1].to_i, args[2], args[3]
+          if (ibegin > iend || ibegin * iend <= 0)
+            puts "*** #{cmd} args error: #{args}\n"
+            raise "*** Import Error."
+          end
+
+          # #command#append#5#5
+          if cmd == 'append'
+            # byebug
+            (ibegin..iend).each {|k| self.records[k.to_s].insert}
+            done << item
+            say_later << "*** #{iend - ibegin + 1} records have been imported into model #{self.model}. \n"
+          end
+
+          # #command#update#4#4#code
+          if cmd == 'update'
+            count = 0
+            (ibegin..iend).each do |k|
+              count += 1 if self.records[k.to_s].update(k, keys, excepts)
+            end
+            done << item
+            say_later << "*** #{count} records of model #{self.model} have been updated.\n"
+          end
+          next
+        end
+
+        puts "*** cmd:#{cmd} not supported\n"
+      end
+      @commands -= done
+      puts "*** #{self.model}.commands left: #{@commands}\n"
+      say_later
+    end
+  end
+end

data/lib/csv_seed/uploader.rb

@@ -0,0 +1,8 @@
+# TODO 没有 2 个外键情况的处理，如：id, code
+module CsvSeed
+  module Uploader
+    def self.file_path(path)
+      "#{Rails.root}/db/seed/#{path}"
+    end
+  end
+end

data/lib/csv_seed/version.rb

@@ -0,0 +1,3 @@
+module CsvSeed
+  VERSION = '0.0.4'
+end

data/lib/csv_seed.rb

@@ -0,0 +1,12 @@
+require 'csv'
+require 'csv_seed/importer'
+require 'csv_seed/record'
+require 'csv_seed/table'
+require 'csv_seed/uploader'
+require 'csv_seed/version'
+
+module CsvSeed
+  def self.load_tasks
+    Dir[File.expand_path('../tasks/*.thor', __FILE__)].each { |ext| load ext }
+  end
+end

metadata

@@ -0,0 +1,78 @@
+--- !ruby/object:Gem::Specification
+name: csv_seed
+version: !ruby/object:Gem::Version
+  version: 0.0.4
+platform: ruby
+authors:
+- chaofan
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2016-12-14 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: thor
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.14'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.14'
+- !ruby/object:Gem::Dependency
+  name: thor-rails
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.0.1
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.0.1
+description: Import csv data to projects which are dependant on ActiveRecord.
+email: jiangchaofan@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- README.md
+- lib/csv_seed.rb
+- lib/csv_seed/importer.rb
+- lib/csv_seed/record.rb
+- lib/csv_seed/table.rb
+- lib/csv_seed/uploader.rb
+- lib/csv_seed/version.rb
+homepage: https://github.com/chaofan/csv_seed
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.5.1
+signing_key: 
+specification_version: 4
+summary: Import csv tables to rails
+test_files: []