dwc-archive 0.2.0 → 0.2.1
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/VERSION +1 -1
- data/lib/dwc-archive.rb +0 -11
- data/lib/dwc-archive/ingester.rb +1 -0
- data/lib/dwc-archive/utf_regex_ruby18.rb +10 -0
- metadata +13 -3
data/VERSION
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
0.2.
|
|
1
|
+
0.2.1
|
data/lib/dwc-archive.rb
CHANGED
|
@@ -22,17 +22,6 @@ class DarwinCore
|
|
|
22
22
|
alias :eml :metadata
|
|
23
23
|
|
|
24
24
|
DEFAULT_TMP_DIR = "/tmp"
|
|
25
|
-
UTF8RGX = /\A(
|
|
26
|
-
[\x09\x0A\x0D\x20-\x7E] # ASCII
|
|
27
|
-
| [\xC2-\xDF][\x80-\xBF] # non-overlong 2-byte
|
|
28
|
-
| \xE0[\xA0-\xBF][\x80-\xBF] # excluding overlongs
|
|
29
|
-
| [\xE1-\xEC\xEE\xEF][\x80-\xBF]{2} # straight 3-byte
|
|
30
|
-
| \xED[\x80-\x9F][\x80-\xBF] # excluding surrogates
|
|
31
|
-
| \xF0[\x90-\xBF][\x80-\xBF]{2} # planes 1-3
|
|
32
|
-
| [\xF1-\xF3][\x80-\xBF]{3} # planes 4-15
|
|
33
|
-
| \xF4[\x80-\x8F][\x80-\xBF]{2} # plane 16
|
|
34
|
-
)*\z/x unless defined? UTF8RGX
|
|
35
|
-
|
|
36
25
|
|
|
37
26
|
def initialize(dwc_path, tmp_dir = DEFAULT_TMP_DIR)
|
|
38
27
|
@archive = DarwinCore::Archive.new(dwc_path, tmp_dir)
|
data/lib/dwc-archive/ingester.rb
CHANGED
|
@@ -12,6 +12,7 @@ class DarwinCore
|
|
|
12
12
|
index_fix = 0; next if @ignore_headers && i == 0
|
|
13
13
|
str = r.join('')
|
|
14
14
|
if defined? FasterCSV
|
|
15
|
+
require File.join(File.dirname(__FILE__), 'utf_regex_ruby18')
|
|
15
16
|
UTF8RGX === str ? res << r : errors << r
|
|
16
17
|
else
|
|
17
18
|
str = str.force_encoding('utf-8')
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
UTF8RGX = /\A(
|
|
2
|
+
[\x09\x0A\x0D\x20-\x7E] # ASCII
|
|
3
|
+
| [\xC2-\xDF][\x80-\xBF] # non-overlong 2-byte
|
|
4
|
+
| \xE0[\xA0-\xBF][\x80-\xBF] # excluding overlongs
|
|
5
|
+
| [\xE1-\xEC\xEE\xEF][\x80-\xBF]{2} # straight 3-byte
|
|
6
|
+
| \xED[\x80-\x9F][\x80-\xBF] # excluding surrogates
|
|
7
|
+
| \xF0[\x90-\xBF][\x80-\xBF]{2} # planes 1-3
|
|
8
|
+
| [\xF1-\xF3][\x80-\xBF]{3} # planes 4-15
|
|
9
|
+
| \xF4[\x80-\x8F][\x80-\xBF]{2} # plane 16
|
|
10
|
+
)*\z/x
|
metadata
CHANGED
|
@@ -1,12 +1,13 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: dwc-archive
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
+
hash: 21
|
|
4
5
|
prerelease: false
|
|
5
6
|
segments:
|
|
6
7
|
- 0
|
|
7
8
|
- 2
|
|
8
|
-
-
|
|
9
|
-
version: 0.2.
|
|
9
|
+
- 1
|
|
10
|
+
version: 0.2.1
|
|
10
11
|
platform: ruby
|
|
11
12
|
authors:
|
|
12
13
|
- Dmitry Mozzherin
|
|
@@ -21,9 +22,11 @@ dependencies:
|
|
|
21
22
|
name: rspec
|
|
22
23
|
prerelease: false
|
|
23
24
|
requirement: &id001 !ruby/object:Gem::Requirement
|
|
25
|
+
none: false
|
|
24
26
|
requirements:
|
|
25
27
|
- - ">="
|
|
26
28
|
- !ruby/object:Gem::Version
|
|
29
|
+
hash: 13
|
|
27
30
|
segments:
|
|
28
31
|
- 1
|
|
29
32
|
- 2
|
|
@@ -35,9 +38,11 @@ dependencies:
|
|
|
35
38
|
name: cucumber
|
|
36
39
|
prerelease: false
|
|
37
40
|
requirement: &id002 !ruby/object:Gem::Requirement
|
|
41
|
+
none: false
|
|
38
42
|
requirements:
|
|
39
43
|
- - ">="
|
|
40
44
|
- !ruby/object:Gem::Version
|
|
45
|
+
hash: 3
|
|
41
46
|
segments:
|
|
42
47
|
- 0
|
|
43
48
|
version: "0"
|
|
@@ -71,6 +76,7 @@ files:
|
|
|
71
76
|
- lib/dwc-archive/extension.rb
|
|
72
77
|
- lib/dwc-archive/ingester.rb
|
|
73
78
|
- lib/dwc-archive/metadata.rb
|
|
79
|
+
- lib/dwc-archive/utf_regex_ruby18.rb
|
|
74
80
|
- lib/ruby_extensions.rb
|
|
75
81
|
- spec/files/broken.tar.gz
|
|
76
82
|
- spec/files/data.tar.gz
|
|
@@ -96,23 +102,27 @@ rdoc_options:
|
|
|
96
102
|
require_paths:
|
|
97
103
|
- lib
|
|
98
104
|
required_ruby_version: !ruby/object:Gem::Requirement
|
|
105
|
+
none: false
|
|
99
106
|
requirements:
|
|
100
107
|
- - ">="
|
|
101
108
|
- !ruby/object:Gem::Version
|
|
109
|
+
hash: 3
|
|
102
110
|
segments:
|
|
103
111
|
- 0
|
|
104
112
|
version: "0"
|
|
105
113
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
114
|
+
none: false
|
|
106
115
|
requirements:
|
|
107
116
|
- - ">="
|
|
108
117
|
- !ruby/object:Gem::Version
|
|
118
|
+
hash: 3
|
|
109
119
|
segments:
|
|
110
120
|
- 0
|
|
111
121
|
version: "0"
|
|
112
122
|
requirements: []
|
|
113
123
|
|
|
114
124
|
rubyforge_project:
|
|
115
|
-
rubygems_version: 1.3.
|
|
125
|
+
rubygems_version: 1.3.7
|
|
116
126
|
signing_key:
|
|
117
127
|
specification_version: 3
|
|
118
128
|
summary: Handler of Darwin Core Archive files
|