patriot-hadoop 0.1.1 → 0.1.2
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/patriot_hadoop/command/hive.rb +15 -2
- data/lib/patriot_hadoop/ext/hive.rb +12 -3
- metadata +5 -6
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 3a75278502a8dd0d24ffca6600f006728ca4b97e
|
4
|
+
data.tar.gz: be5d6721503438d426a33b466bd9ccd7e6385c03
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 60b66f64532ad327acfa9a5804b1085788f4f07e2ac87b0cb8d2c210553287c8296d117add0ada7e60fbd4859f13a56109a27b3678e16f9667b8c48acaa037f7
|
7
|
+
data.tar.gz: d5ba491a269f4d55789c090fdb8eb88cd5c57a9b13e6cd6ff118c3c90b9781ede959db5720519733860d7c96fd2daa943933651c1423a924717a0d94574c3cb1
|
@@ -4,7 +4,7 @@ module PatriotHadoop
|
|
4
4
|
declare_command_name :hive
|
5
5
|
include PatriotHadoop::Ext::Hive
|
6
6
|
|
7
|
-
command_attr :hive_ql, :output_prefix, :exec_user, :props, :name_suffix
|
7
|
+
command_attr :hive_ql, :output_prefix, :compression, :exec_user, :props, :name_suffix
|
8
8
|
|
9
9
|
def job_id
|
10
10
|
job_id = "#{command_name}"
|
@@ -28,8 +28,8 @@ module PatriotHadoop
|
|
28
28
|
|
29
29
|
tmpfile = output_prefix + '.hql'
|
30
30
|
_create_hivequery_tmpfile(@hive_ql, tmpfile, opt)
|
31
|
+
output_file = _create_output_filename(output_prefix, @compression)
|
31
32
|
|
32
|
-
output_file = output_prefix + '.tsv'
|
33
33
|
execute_hivequery(tmpfile, output_file, @exec_user)
|
34
34
|
|
35
35
|
if File.zero?(output_file)
|
@@ -41,6 +41,19 @@ module PatriotHadoop
|
|
41
41
|
end
|
42
42
|
|
43
43
|
|
44
|
+
# true / 'gzip' / 'bzip2' are available.
|
45
|
+
def _create_output_filename(output_prefix, compression)
|
46
|
+
output_file = output_prefix + '.tsv'
|
47
|
+
case compression
|
48
|
+
when true, 'gzip'
|
49
|
+
output_file += '.gz'
|
50
|
+
when 'bzip2'
|
51
|
+
output_file += '.bz2'
|
52
|
+
end
|
53
|
+
return output_file
|
54
|
+
end
|
55
|
+
|
56
|
+
|
44
57
|
def _create_hivequery_tmpfile(hive_ql, tmpfile, opt={})
|
45
58
|
hive_ql = _add_udfs(hive_ql, opt[:udf]) if opt.has_key?(:udf)
|
46
59
|
hive_ql = "#{_set_hive_property_prefix(opt[:props])}#{hive_ql}" if opt.has_key?(:props)
|
@@ -16,7 +16,7 @@ module PatriotHadoop
|
|
16
16
|
|
17
17
|
|
18
18
|
def execute_hivequery(hql_file, output_file=nil, user=nil)
|
19
|
-
command = "hive -f \"#{hql_file}\""
|
19
|
+
command = "hive -f \"#{hql_file}\"#{_compress_option(File.extname(output_file))}"
|
20
20
|
unless user.nil?
|
21
21
|
if user !~ /^[a-z_][a-z0-9_]{0,30}$/
|
22
22
|
raise HiveException, "Invalid username"
|
@@ -27,8 +27,17 @@ module PatriotHadoop
|
|
27
27
|
end
|
28
28
|
|
29
29
|
|
30
|
+
def _compress_option(extension)
|
31
|
+
return case extension
|
32
|
+
when '.gz' then ' | gzip --stdout --force'
|
33
|
+
when '.bz2' then ' | bzip2 --stdout --force'
|
34
|
+
else ''
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
|
30
39
|
def _execute_hivequery_internal(command, output_file)
|
31
|
-
|
40
|
+
res = execute_command(command) do |status, so, se|
|
32
41
|
err_size = File.stat(se).size
|
33
42
|
err_msg = ""
|
34
43
|
max_err_size = HIVE_MAX_ERROR_MSG_SIZE
|
@@ -41,7 +50,7 @@ module PatriotHadoop
|
|
41
50
|
end
|
42
51
|
raise HiveException, "#{command}\n#{err_msg}"
|
43
52
|
end
|
44
|
-
File.rename(
|
53
|
+
File.rename(res, output_file) unless output_file.nil?
|
45
54
|
end
|
46
55
|
|
47
56
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: patriot-hadoop
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.1.
|
4
|
+
version: 0.1.2
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Hitoshi Tsuda
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2018-05-23 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: patriot-workflow-scheduler
|
@@ -38,9 +38,9 @@ files:
|
|
38
38
|
- lib/patriot_hadoop/ext.rb
|
39
39
|
- lib/patriot_hadoop.rb
|
40
40
|
- init.rb
|
41
|
-
homepage: https://github.com/CyberAgent/patriot-workflow-scheduler
|
41
|
+
homepage: https://github.com/CyberAgent/patriot-workflow-scheduler/tree/master/plugins/patriot-hadoop
|
42
42
|
licenses:
|
43
|
-
- Apache
|
43
|
+
- Apache-2.0
|
44
44
|
metadata: {}
|
45
45
|
post_install_message:
|
46
46
|
rdoc_options: []
|
@@ -58,9 +58,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
58
58
|
version: '0'
|
59
59
|
requirements: []
|
60
60
|
rubyforge_project: patriot-hadoop
|
61
|
-
rubygems_version: 2.0.14
|
61
|
+
rubygems_version: 2.0.14.1
|
62
62
|
signing_key:
|
63
63
|
specification_version: 4
|
64
64
|
summary: Hadoop plugin for Patriot Workflow Scheduler
|
65
65
|
test_files: []
|
66
|
-
has_rdoc:
|