csv_patch 1.0.2 → 1.1.0

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    NDkxYjBmMGNhMjY1ZTY5ZjNiZTEwNWI2YjFjZjUwNGRiZjg3OTVhMQ==
+    MmUwOTQxZTdjZDZlOGE3YjVlNjFiOGJmZjg4OWI3ZTU0ODk4ZjQzZA==
   data.tar.gz: !binary |-
-    YTJhNGEwZTdjNzE5MjRiNzg2MjY3Njg1ZGE4MWY2ZDQ2MDU0MDhkMg==
+    YTkyZmNjMTYzNmFjOTJjMDVlMWZkYmY0ZTJkZGQxMWU2MTJiOTFmMA==
 SHA512:
   metadata.gz: !binary |-
-    MDYyYTliMjU1MDYwMWE5NmM4MjdkNjhhNzNjYzI2ZGM1ZWI0YzhhOGUzOGU5
-    NDFlN2ZmMmJhMTdiMDJmMGNlNTViZTI3OTMzNmY2ZTE5M2FkYWI0Mzc5ZmFi
-    ZjEyYmQyNzhlNWQ3N2MyMDYyMjdjMWIxMDhhOTkwMzBlMGYzZDY=
+    OTk4MzUzOGQ5OWFjY2IyNjY4MDQ0ZmRlODYxYjlhYzYwZmM0MWIyM2E0ODYy
+    MTg2M2Q4YzAzMGVjOTM2OGFhNjlhODQ3YzdmYjQ1MWJlNDc2NDA5MDU2NmYy
+    YmEzMmY5NDJkNzRjMGQ3MzJlMmQ3MTI0ZTQyODRkMzQyNmRkMjI=
   data.tar.gz: !binary |-
-    MDNlYjkwYjg0YTkwNGU2YjYzZTBlYmUzZTk2ZWEzOGE0OGE2MjA5NmViZTVh
-    MjM1NjYyMDEwOTlkNTZmNTUzZTIwYWFiNzg0NTJkMmIxZTZlYTdhM2ZlNzc3
-    MjgxZDVlNjc1ZjhkZDc5NWE4N2RkNzk0NTQ0NjBiMjJkMjhiZGI=
+    MjcxY2E5OTg2ZTY4ZDI0ZWQyYTg0MjkzN2UzOTcxYTMxNGM2NGZjZWVlMmZi
+    YTQ1MTk3MzU4ZTZhZjRlNjZiZjk4MWZkZDAyMWJkMjllYTZlYzMwNzJlYmFm
+    NWQ1ODhjYTNkYWExYTVmM2U2MGNiN2FlZjdkZDBiZGI3ZjVjOWI=

data/VERSION

@@ -1 +1 @@
-1.0.2
+1.1.0

data/bin/csv_patch

@@ -20,7 +20,7 @@ OptionParser.new do |opts|
     options[:changes] = File.new(changes_file, 'r')
   end
 
-  opts.on('-bBATCH_SIZE', '--batch_size=BATCH_SIZE', 'Number of changes to read before applying a patch') do |batch_size|
+  opts.on('-bBATCH_SIZE', '--batch_size=BATCH_SIZE', 'Number of changes to read before applying a patch. Defaults to 500.') do |batch_size|
     options[:batch_size] = batch_size.to_i
   end
 

data/csv_patch.gemspec

@@ -2,16 +2,16 @@
 # DO NOT EDIT THIS FILE DIRECTLY
 # Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
 # -*- encoding: utf-8 -*-
-# stub: csv_patch 1.0.2 ruby lib
+# stub: csv_patch 1.1.0 ruby lib
 
 Gem::Specification.new do |s|
   s.name = "csv_patch"
-  s.version = "1.0.2"
+  s.version = "1.1.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.require_paths = ["lib"]
   s.authors = ["Lyudmil"]
-  s.date = "2015-10-28"
+  s.date = "2015-11-03"
   s.description = "Applies a list of changes in a given format to a CSV file"
   s.email = "lyudmilangelov@gmail.com"
   s.executables = ["csv_patch"]
@@ -23,6 +23,7 @@ Gem::Specification.new do |s|
     "bin/csv_patch",
     "csv_patch.gemspec",
     "lib/csv_patch.rb",
+    "lib/csv_patch/batches.rb",
     "lib/csv_patch/compression.rb",
     "lib/csv_patch/operation.rb",
     "lib/csv_patch/patch.rb",

data/lib/csv_patch/batches.rb

@@ -0,0 +1,62 @@
+require 'csv_patch/stream_batch'
+require 'csv_patch/patch'
+
+module CsvPatch
+
+  class Batches
+
+    DEFAULT_BATCH_SIZE = 500
+
+    def initialize options
+      @intermediate_files = []
+      @input, @output     = options[:input], options[:output]
+      @batches            = StreamBatch.new(options[:changes], options[:batch_size] || DEFAULT_BATCH_SIZE)
+    end
+
+    def execute
+      apply_changes_in_batches
+      close_intermediate_files
+    end
+
+    private
+
+    def apply_changes_in_batches
+      @batches.each { |batch| patch_for(batch).apply }
+    end
+
+    def patch_for batch
+      Patch.new(patch_options(batch))
+    end
+
+    def patch_options batch
+      { input: input_for_next_patch, output: output_for_next_patch(batch), changes: batch.changes }
+    end
+
+    def close_intermediate_files
+      @intermediate_files.each(&:close)
+    end
+
+    def input_for_next_patch
+      last_intermediate_file || @input
+    end
+
+    def last_intermediate_file
+      return if @intermediate_files.empty?
+
+      @intermediate_files.last.rewind
+      @intermediate_files.last
+    end
+
+    def output_for_next_patch batch
+      return @output if batch.last?
+      new_intermediate_file
+    end
+
+    def new_intermediate_file
+      @intermediate_files << Tempfile.new('csv_patch')
+      @intermediate_files.last
+    end
+
+  end
+
+end

data/lib/csv_patch/patch.rb

@@ -9,7 +9,7 @@ module CsvPatch
     TEMPFILE_NAME = 'csv_patch'
 
     def initialize options
-      @input, @output   = options[:input], options[:output]
+      @input, @output   = CSV.new(options[:input]), options[:output]
 
       @revision_result  = Tempfile.new(TEMPFILE_NAME)
       @revision         = Revision.new(options[:changes], @revision_result)

data/lib/csv_patch/revision.rb

@@ -4,24 +4,22 @@ module CsvPatch
 
   class Revision < Operation
 
-    EMPTY_LINE = "\n"
-
     def initialize changes, output_stream
       @output_stream  = output_stream
       @changes        = changes
 
-      header_line(EMPTY_LINE)
+      header_line([])
     end
 
     def header_line line
       return unless line
 
-      @columns = csv_values(line)
+      @columns = line
       mark_all_columns_empty
     end
 
     def replace_line line
-      emit replacement_line_for csv_values(line)
+      emit replacement_line_for(line)
     end
 
     def add_new_lines

data/lib/csv_patch/stream_batch.rb

@@ -1,36 +1,44 @@
 require 'json'
+require 'ostruct'
 
-class StreamBatch
+module CsvPatch
 
-  def initialize stream, batch_size
-    @stream     = stream
-    @batch_size = batch_size
-  end
+  class StreamBatch
 
-  def each
-    yield next_batch until stream_end?
-  end
+    def initialize stream, batch_size
+      @stream     = stream
+      @batch_size = batch_size
+    end
 
-  private
+    def each
+      yield next_batch until stream_end?
+    end
 
-  def next_batch
-    batch = {}
+    private
 
-    batch.merge!(next_change) until batch_full?(batch)
+    def next_batch
+      OpenStruct.new(changes: next_batch_of_changes, last?: stream_end?)
+    end
 
-    batch
-  end
+    def next_batch_of_changes
+      batch = {}
 
-  def next_change
-    JSON.parse(@stream.gets)
-  end
+      batch.merge!(next_change) until batch_full?(batch)
 
-  def batch_full? batch
-    batch.size >= @batch_size || stream_end?
-  end
+      batch
+    end
 
-  def stream_end?
-    @stream.eof?
-  end
+    def next_change
+      JSON.parse(@stream.gets)
+    end
+
+    def batch_full? batch
+      batch.size >= @batch_size || stream_end?
+    end
 
+    def stream_end?
+      @stream.eof?
+    end
+
+  end
 end

data/lib/csv_patch.rb

@@ -1,22 +1,9 @@
-require 'csv_patch/stream_batch'
-require 'csv_patch/patch'
+require 'csv_patch/batches'
 
 module CsvPatch
 
-  DEFAULT_BATCH_SIZE = 500
-
   def self.patch options
-    batches(options).each { |changes| apply_patch(options, changes) }
-  end
-
-  private
-
-  def self.batches options
-    StreamBatch.new(options[:changes], options[:batch_size] || DEFAULT_BATCH_SIZE)
-  end
-
-  def self.apply_patch options, changes
-    Patch.new(input: options[:input], output: options[:output], changes: changes).apply
+    Batches.new(options).execute
   end
 
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: csv_patch
 version: !ruby/object:Gem::Version
-  version: 1.0.2
+  version: 1.1.0
 platform: ruby
 authors:
 - Lyudmil
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-10-28 00:00:00.000000000 Z
+date: 2015-11-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: jeweler
@@ -52,6 +52,7 @@ files:
 - bin/csv_patch
 - csv_patch.gemspec
 - lib/csv_patch.rb
+- lib/csv_patch/batches.rb
 - lib/csv_patch/compression.rb
 - lib/csv_patch/operation.rb
 - lib/csv_patch/patch.rb