mandy 0.4.2 → 0.4.3
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/bin/mandy-hadoop +3 -2
- data/lib/job.rb +8 -4
- data/lib/wrappers/mandy_wrapper.rb +1 -1
- metadata +1 -1
data/bin/mandy-hadoop
CHANGED
|
@@ -75,14 +75,15 @@ Mandy::Job.jobs.each_with_index do |job, i|
|
|
|
75
75
|
output = File.join(output_folder, "#{i+1}-#{job.name.downcase.gsub(/\W/, '-')}")
|
|
76
76
|
|
|
77
77
|
bootstrap_file = File.expand_path(File.join(File.dirname(__FILE__), '..', 'bootstrap.rb'))
|
|
78
|
+
inputreader = job.input_format == :xml ? "StreamXmlRecordReader,begin=<#{job.input_format_options[:xml_tag]} ,end=</#{job.input_format_options[:xml_tag]}>" : nil
|
|
78
79
|
|
|
79
80
|
command = %($HADOOP_HOME/bin/hadoop jar $HADOOP_HOME/contrib/streaming/hadoop-*-streaming.jar #{jobconf}\
|
|
81
|
+
-files "#{payload}","#{bootstrap_file}" \
|
|
80
82
|
-conf '#{config}' \
|
|
81
83
|
-input "#{input}" \
|
|
84
|
+
#{ inputreader.nil? ? '' : "-inputreader \"#{inputreader}\"" } \
|
|
82
85
|
-mapper "ruby bootstrap.rb #{File.basename(payload)} map #{filename} '#{job.name}'" \
|
|
83
86
|
-reducer "ruby bootstrap.rb #{File.basename(payload)} reduce #{filename} '#{job.name}'" \
|
|
84
|
-
-file "#{payload}" \
|
|
85
|
-
-file "#{bootstrap_file}" \
|
|
86
87
|
#{ cmdenv.nil? ? '' : "-cmdenv #{cmdenv}" }\
|
|
87
88
|
-output "#{output}")
|
|
88
89
|
|
data/lib/job.rb
CHANGED
|
@@ -12,7 +12,8 @@ module Mandy
|
|
|
12
12
|
|
|
13
13
|
attr_reader :settings
|
|
14
14
|
attr_reader :name
|
|
15
|
-
|
|
15
|
+
attr_reader :input_format_options
|
|
16
|
+
|
|
16
17
|
def initialize(name, &blk)
|
|
17
18
|
@name = name
|
|
18
19
|
@settings = {}
|
|
@@ -25,11 +26,14 @@ module Mandy
|
|
|
25
26
|
modules.each {|m| @modules << m}
|
|
26
27
|
end
|
|
27
28
|
alias_method :serialize, :mixin
|
|
28
|
-
|
|
29
|
-
def input_format(format)
|
|
29
|
+
|
|
30
|
+
def input_format(format=nil, options={})
|
|
31
|
+
return @input_format if format.nil?
|
|
32
|
+
|
|
30
33
|
@input_format = format
|
|
34
|
+
@input_format_options = options
|
|
31
35
|
end
|
|
32
|
-
|
|
36
|
+
|
|
33
37
|
def output_format(format)
|
|
34
38
|
@output_format = format
|
|
35
39
|
end
|
|
@@ -27,7 +27,7 @@ module Mandy
|
|
|
27
27
|
def put_files_on_hdfs(hdfs_path, input_files)
|
|
28
28
|
input_files = [input_files] unless input_files.is_a?(Array)
|
|
29
29
|
input_files.each do |input_file|
|
|
30
|
-
input_file_path = File.expand_path(input_file.path)
|
|
30
|
+
input_file_path = input_file.is_a?(File) ? File.expand_path(input_file.path) : input_file
|
|
31
31
|
base_filename = input_file_path.split("/").last
|
|
32
32
|
dest_file = ["input/#{hdfs_path}", base_filename].join("/")
|
|
33
33
|
run_command "mandy-put #{input_file_path} #{dest_file}"
|