sanzang 1.0.4 → 1.0.5
Sign up to get free protection for your applications and to get access to all the features.
- data/bin/sanzang +2 -2
- data/lib/sanzang.rb +1 -2
- data/lib/sanzang/batch_translator.rb +9 -7
- data/lib/sanzang/command/batch.rb +1 -2
- data/lib/sanzang/command/reflow.rb +1 -2
- data/lib/sanzang/command/sanzang_cmd.rb +1 -2
- data/lib/sanzang/command/translate.rb +1 -2
- data/lib/sanzang/platform.rb +1 -2
- data/lib/sanzang/text_formatter.rb +1 -2
- data/lib/sanzang/translation_table.rb +16 -19
- data/lib/sanzang/translator.rb +1 -2
- data/lib/sanzang/version.rb +2 -3
- data/test/tc_reflow_encodings.rb +1 -2
- data/test/tc_simple_translation.rb +1 -2
- metadata +10 -10
data/bin/sanzang
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
#!/usr/bin/env ruby
|
2
|
-
#
|
2
|
+
# coding: UTF-8
|
3
3
|
#--
|
4
|
-
# Copyright (C) 2012 Lapis Lazuli Texts
|
4
|
+
# Copyright (C) 2012-2013 Lapis Lazuli Texts
|
5
5
|
#
|
6
6
|
# This program is free software: you can redistribute it and/or modify it under
|
7
7
|
# the terms of the GNU General Public License as published by the Free Software
|
data/lib/sanzang.rb
CHANGED
@@ -1,5 +1,4 @@
|
|
1
|
-
|
2
|
-
# -*- encoding: UTF-8 -*-
|
1
|
+
# coding: UTF-8
|
3
2
|
#--
|
4
3
|
# Copyright (C) 2012-2013 Lapis Lazuli Texts
|
5
4
|
#
|
@@ -36,12 +35,15 @@ module Sanzang
|
|
36
35
|
# return value is an array containing all the output file paths.
|
37
36
|
#
|
38
37
|
def translate_batch(fpath_pairs, verbose = true, jobs = nil)
|
39
|
-
|
40
|
-
|
41
|
-
|
42
|
-
|
38
|
+
options = {}
|
39
|
+
if RUBY_PLATFORM =~ /java/
|
40
|
+
options[:in_threads] = jobs || Sanzang::Platform.processor_count
|
41
|
+
elsif Sanzang::Platform.unix_processes?
|
42
|
+
options[:in_processes] = jobs || Sanzang::Platform.processor_count
|
43
|
+
else
|
44
|
+
options[:in_processes] = 0
|
43
45
|
end
|
44
|
-
Parallel.map(fpath_pairs,
|
46
|
+
Parallel.map(fpath_pairs, options) do |f1,f2|
|
45
47
|
translate_io(f1, f2)
|
46
48
|
if verbose
|
47
49
|
$stderr.write "[#{Process.pid}] #{File.expand_path(f2)} \n"
|
data/lib/sanzang/platform.rb
CHANGED
@@ -1,5 +1,4 @@
|
|
1
|
-
|
2
|
-
# -*- encoding: UTF-8 -*-
|
1
|
+
# coding: UTF-8
|
3
2
|
#--
|
4
3
|
# Copyright (C) 2012-2013 Lapis Lazuli Texts
|
5
4
|
#
|
@@ -29,7 +28,7 @@ module Sanzang
|
|
29
28
|
# summarized as follows:
|
30
29
|
#
|
31
30
|
# - Each line of text is a record for a translation rule.
|
32
|
-
# - Each record
|
31
|
+
# - Each record may begin with "~|" and end with "|~".
|
33
32
|
# - Fields in the record are separated by the "|" character.
|
34
33
|
# - The first field contains the term in the source language.
|
35
34
|
# - Subsequent fields are equivalent terms in destination languages.
|
@@ -43,15 +42,11 @@ module Sanzang
|
|
43
42
|
def initialize(rules)
|
44
43
|
contents = rules.kind_of?(String) ? rules : rules.read
|
45
44
|
contents.encode!(Encoding::UTF_8)
|
46
|
-
|
47
|
-
|
48
|
-
|
49
|
-
|
50
|
-
|
51
|
-
|
52
|
-
@records = contents.gsub("\r", "").split("\n").collect do |rec|
|
53
|
-
rec.strip.gsub(left, "").gsub(right, "").split(separator)
|
54
|
-
end
|
45
|
+
contents.strip!
|
46
|
+
contents.gsub!(/^\s*|\s*$|\r/, "")
|
47
|
+
contents.gsub!("~|", "")
|
48
|
+
contents.gsub!("|~", "")
|
49
|
+
@records = contents.split("\n").collect {|r| r.split("|") }
|
55
50
|
|
56
51
|
if @records.length < 1
|
57
52
|
raise "Table must have at least 1 row"
|
@@ -59,9 +54,9 @@ module Sanzang
|
|
59
54
|
raise "Table must have at least 2 columns"
|
60
55
|
end
|
61
56
|
|
62
|
-
|
63
|
-
|
64
|
-
if
|
57
|
+
width = records[0].length
|
58
|
+
@records.each do |r|
|
59
|
+
if r.length != width
|
65
60
|
raise "Column mismatch: Line #{i + 1}"
|
66
61
|
end
|
67
62
|
end
|
@@ -75,6 +70,12 @@ module Sanzang
|
|
75
70
|
@records[index]
|
76
71
|
end
|
77
72
|
|
73
|
+
# The text encoding used for all translation table data
|
74
|
+
#
|
75
|
+
def encoding
|
76
|
+
Encoding::UTF_8
|
77
|
+
end
|
78
|
+
|
78
79
|
# Find a record by the source language term (first column).
|
79
80
|
#
|
80
81
|
def find(term)
|
@@ -97,9 +98,5 @@ module Sanzang
|
|
97
98
|
#
|
98
99
|
attr_reader :records
|
99
100
|
|
100
|
-
# The text encoding used for all translation table data
|
101
|
-
#
|
102
|
-
attr_reader :encoding
|
103
|
-
|
104
101
|
end
|
105
102
|
end
|
data/lib/sanzang/translator.rb
CHANGED
data/lib/sanzang/version.rb
CHANGED
@@ -1,5 +1,4 @@
|
|
1
|
-
|
2
|
-
# -*- encoding: UTF-8 -*-
|
1
|
+
# coding: UTF-8
|
3
2
|
#--
|
4
3
|
# Copyright (C) 2012-2013 Lapis Lazuli Texts
|
5
4
|
#
|
@@ -20,6 +19,6 @@ module Sanzang
|
|
20
19
|
|
21
20
|
# Current version number of Sanzang
|
22
21
|
#
|
23
|
-
VERSION = "1.0.
|
22
|
+
VERSION = "1.0.5"
|
24
23
|
|
25
24
|
end
|
data/test/tc_reflow_encodings.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: sanzang
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.0.
|
4
|
+
version: 1.0.5
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,32 +9,31 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2013-
|
12
|
+
date: 2013-10-13 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: parallel
|
16
16
|
requirement: !ruby/object:Gem::Requirement
|
17
17
|
none: false
|
18
18
|
requirements:
|
19
|
-
- -
|
19
|
+
- - ~>
|
20
20
|
- !ruby/object:Gem::Version
|
21
|
-
version: 0.
|
21
|
+
version: '0.8'
|
22
22
|
type: :runtime
|
23
23
|
prerelease: false
|
24
24
|
version_requirements: !ruby/object:Gem::Requirement
|
25
25
|
none: false
|
26
26
|
requirements:
|
27
|
-
- -
|
27
|
+
- - ~>
|
28
28
|
- !ruby/object:Gem::Version
|
29
|
-
version: 0.
|
29
|
+
version: '0.8'
|
30
30
|
description: Sanzang is a program built for machine translation of natural languages.
|
31
31
|
This application is particularly suitable as a translation aid for CJK languages
|
32
32
|
including ancient texts. The translation method is rule-based, and translation rules
|
33
33
|
are stored in flat files as delimited text. This program can also utilize multiprocessing
|
34
34
|
to naturally scale to multiple processors and processor cores. Sanzang is available
|
35
35
|
under the GNU GPL, version 3.
|
36
|
-
email:
|
37
|
-
- lapislazulitexts@gmail.com
|
36
|
+
email: lapislazulitexts@gmail.com
|
38
37
|
executables:
|
39
38
|
- sanzang
|
40
39
|
extensions: []
|
@@ -90,12 +89,13 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
90
89
|
- - ! '>='
|
91
90
|
- !ruby/object:Gem::Version
|
92
91
|
version: '0'
|
93
|
-
requirements:
|
92
|
+
requirements:
|
93
|
+
- parallel ~> 0.8
|
94
94
|
rubyforge_project:
|
95
95
|
rubygems_version: 1.8.23
|
96
96
|
signing_key:
|
97
97
|
specification_version: 3
|
98
|
-
summary:
|
98
|
+
summary: Machine translation for CJK languages
|
99
99
|
test_files:
|
100
100
|
- test/tc_reflow_encodings.rb
|
101
101
|
- test/tc_simple_translation.rb
|