dwc-archive 0.1.2
Sign up to get free protection for your applications and to get access to all the features.
- data/.document +5 -0
- data/.gitignore +21 -0
- data/LICENSE +20 -0
- data/README.rdoc +38 -0
- data/Rakefile +57 -0
- data/VERSION +1 -0
- data/dwc-archive.gemspec +74 -0
- data/features/dwc-archive.feature +34 -0
- data/features/step_definitions/dwc-archive_steps.rb +95 -0
- data/features/support/env.rb +4 -0
- data/lib/dwc-archive.rb +29 -0
- data/lib/dwc-archive/.expander.rb.swo +0 -0
- data/lib/dwc-archive/archive.rb +37 -0
- data/lib/dwc-archive/core.rb +34 -0
- data/lib/dwc-archive/expander.rb +71 -0
- data/lib/dwc-archive/extension.rb +31 -0
- data/lib/dwc-archive/metadata.rb +38 -0
- data/lib/ruby_extensions.rb +64 -0
- data/spec/dwc-archive_spec.rb +7 -0
- data/spec/files/data.tar.gz +0 -0
- data/spec/files/data.zip +0 -0
- data/spec/files/eml.xml +46 -0
- data/spec/files/meta.xml +22 -0
- data/spec/lib/ruby_extenstions_spec.rb +15 -0
- data/spec/spec.opts +1 -0
- data/spec/spec_helper.rb +9 -0
- metadata +115 -0
data/.document
ADDED
data/.gitignore
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
Copyright (c) 2010 Dmitry Mozzherin
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
4
|
+
a copy of this software and associated documentation files (the
|
5
|
+
"Software"), to deal in the Software without restriction, including
|
6
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
7
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
8
|
+
permit persons to whom the Software is furnished to do so, subject to
|
9
|
+
the following conditions:
|
10
|
+
|
11
|
+
The above copyright notice and this permission notice shall be
|
12
|
+
included in all copies or substantial portions of the Software.
|
13
|
+
|
14
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
15
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
16
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
17
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
18
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
19
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
20
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.rdoc
ADDED
@@ -0,0 +1,38 @@
|
|
1
|
+
= Darwin Core Archive
|
2
|
+
|
3
|
+
Darwin Core Archive format is a current standard for information exchange between Global Names Architecture modules. This gem allows to work with Darwin Core Archive data compressed to either zip or tar.gz files. More information about Darwing Core Archive can be found on a GBIF page[http://www.gbif.org/informatics/standards-and-tools/publishing-data/data-standards/darwin-core-archives/]:
|
4
|
+
|
5
|
+
== Installation
|
6
|
+
|
7
|
+
Update to latest rubygems (v >= 1.3.6) which adds gemcutter sources by default.
|
8
|
+
|
9
|
+
sudo gem install dwc-archive
|
10
|
+
|
11
|
+
== Usage
|
12
|
+
|
13
|
+
require 'rubygems'
|
14
|
+
require 'dwc-archive'
|
15
|
+
|
16
|
+
dwc = DWCA::DarwinCore.new('/path_to_file/archive_file.tar.gz')
|
17
|
+
dwc.archive.files # the archive file list
|
18
|
+
dwc.metadata.data # summary of metadata from eml.xml if it exists
|
19
|
+
dwc.metadata.authors # authors of the archive
|
20
|
+
dwc.core.data # summary of DarwinCore main file
|
21
|
+
dwc.core.file_path # path to the DarwinCore main file
|
22
|
+
dwc.core.extensions # array of DarwinCore Star extensions
|
23
|
+
dwc.core.extensions[0].data # summary for an extension
|
24
|
+
|
25
|
+
== Note on Patches/Pull Requests
|
26
|
+
|
27
|
+
* Fork the project.
|
28
|
+
* Make your feature addition or bug fix.
|
29
|
+
* Add tests for it. This is important so I don't break it in a
|
30
|
+
future version unintentionally.
|
31
|
+
* Commit, do not mess with rakefile, version, or history.
|
32
|
+
(if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
|
33
|
+
* Send me a pull request. Bonus points for topic branches.
|
34
|
+
|
35
|
+
|
36
|
+
== Copyright
|
37
|
+
|
38
|
+
Copyright (c) 2010 Dmitry Mozzherin. See LICENSE for details.
|
data/Rakefile
ADDED
@@ -0,0 +1,57 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'rake'
|
3
|
+
|
4
|
+
begin
|
5
|
+
require 'jeweler'
|
6
|
+
Jeweler::Tasks.new do |gem|
|
7
|
+
gem.name = "dwc-archive"
|
8
|
+
gem.summary = %Q{Handler of Darwin Core Archive files}
|
9
|
+
gem.description = %Q{Darwin Core Archive Files are current standard exchange format for GLobal Names Architecture modules. This project creates ways to work with such files.}
|
10
|
+
gem.email = "dmozzherin at gmail dot com"
|
11
|
+
gem.homepage = "http://github.com/dimus/dwc-archive"
|
12
|
+
gem.authors = ["Dmitry Mozzherin"]
|
13
|
+
gem.add_development_dependency "rspec", ">= 1.2.9"
|
14
|
+
gem.add_development_dependency "cucumber", ">= 0"
|
15
|
+
# gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
|
16
|
+
end
|
17
|
+
Jeweler::GemcutterTasks.new
|
18
|
+
rescue LoadError
|
19
|
+
puts "Jeweler (or a dependency) not available. Install it with: gem install jeweler"
|
20
|
+
end
|
21
|
+
|
22
|
+
require 'spec/rake/spectask'
|
23
|
+
Spec::Rake::SpecTask.new(:spec) do |spec|
|
24
|
+
spec.libs << 'lib' << 'spec'
|
25
|
+
spec.spec_files = FileList['spec/**/*_spec.rb']
|
26
|
+
end
|
27
|
+
|
28
|
+
Spec::Rake::SpecTask.new(:rcov) do |spec|
|
29
|
+
spec.libs << 'lib' << 'spec'
|
30
|
+
spec.pattern = 'spec/**/*_spec.rb'
|
31
|
+
spec.rcov = true
|
32
|
+
end
|
33
|
+
|
34
|
+
task :spec => :check_dependencies
|
35
|
+
|
36
|
+
begin
|
37
|
+
require 'cucumber/rake/task'
|
38
|
+
Cucumber::Rake::Task.new(:features)
|
39
|
+
|
40
|
+
task :features => :check_dependencies
|
41
|
+
rescue LoadError
|
42
|
+
task :features do
|
43
|
+
abort "Cucumber is not available. In order to run features, you must: sudo gem install cucumber"
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
task :default => :spec
|
48
|
+
|
49
|
+
require 'rake/rdoctask'
|
50
|
+
Rake::RDocTask.new do |rdoc|
|
51
|
+
version = File.exist?('VERSION') ? File.read('VERSION') : ""
|
52
|
+
|
53
|
+
rdoc.rdoc_dir = 'rdoc'
|
54
|
+
rdoc.title = "dwc-archive #{version}"
|
55
|
+
rdoc.rdoc_files.include('README*')
|
56
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
57
|
+
end
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.1.2
|
data/dwc-archive.gemspec
ADDED
@@ -0,0 +1,74 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE DIRECTLY
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run the gemspec command
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{dwc-archive}
|
8
|
+
s.version = "0.1.2"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Dmitry Mozzherin"]
|
12
|
+
s.date = %q{2010-03-18}
|
13
|
+
s.description = %q{Darwin Core Archive Files are current standard exchange format for GLobal Names Architecture modules. This project creates ways to work with such files.}
|
14
|
+
s.email = %q{dmozzherin at gmail dot com}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
".gitignore",
|
22
|
+
"LICENSE",
|
23
|
+
"README.rdoc",
|
24
|
+
"Rakefile",
|
25
|
+
"VERSION",
|
26
|
+
"dwc-archive.gemspec",
|
27
|
+
"features/dwc-archive.feature",
|
28
|
+
"features/step_definitions/dwc-archive_steps.rb",
|
29
|
+
"features/support/env.rb",
|
30
|
+
"lib/dwc-archive.rb",
|
31
|
+
"lib/dwc-archive/.expander.rb.swo",
|
32
|
+
"lib/dwc-archive/archive.rb",
|
33
|
+
"lib/dwc-archive/core.rb",
|
34
|
+
"lib/dwc-archive/expander.rb",
|
35
|
+
"lib/dwc-archive/extension.rb",
|
36
|
+
"lib/dwc-archive/metadata.rb",
|
37
|
+
"lib/ruby_extensions.rb",
|
38
|
+
"spec/dwc-archive_spec.rb",
|
39
|
+
"spec/files/data.tar.gz",
|
40
|
+
"spec/files/data.zip",
|
41
|
+
"spec/files/eml.xml",
|
42
|
+
"spec/files/meta.xml",
|
43
|
+
"spec/lib/ruby_extenstions_spec.rb",
|
44
|
+
"spec/spec.opts",
|
45
|
+
"spec/spec_helper.rb"
|
46
|
+
]
|
47
|
+
s.homepage = %q{http://github.com/dimus/dwc-archive}
|
48
|
+
s.rdoc_options = ["--charset=UTF-8"]
|
49
|
+
s.require_paths = ["lib"]
|
50
|
+
s.rubygems_version = %q{1.3.6}
|
51
|
+
s.summary = %q{Handler of Darwin Core Archive files}
|
52
|
+
s.test_files = [
|
53
|
+
"spec/dwc-archive_spec.rb",
|
54
|
+
"spec/lib/ruby_extenstions_spec.rb",
|
55
|
+
"spec/spec_helper.rb"
|
56
|
+
]
|
57
|
+
|
58
|
+
if s.respond_to? :specification_version then
|
59
|
+
current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
|
60
|
+
s.specification_version = 3
|
61
|
+
|
62
|
+
if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
|
63
|
+
s.add_development_dependency(%q<rspec>, [">= 1.2.9"])
|
64
|
+
s.add_development_dependency(%q<cucumber>, [">= 0"])
|
65
|
+
else
|
66
|
+
s.add_dependency(%q<rspec>, [">= 1.2.9"])
|
67
|
+
s.add_dependency(%q<cucumber>, [">= 0"])
|
68
|
+
end
|
69
|
+
else
|
70
|
+
s.add_dependency(%q<rspec>, [">= 1.2.9"])
|
71
|
+
s.add_dependency(%q<cucumber>, [">= 0"])
|
72
|
+
end
|
73
|
+
end
|
74
|
+
|
@@ -0,0 +1,34 @@
|
|
1
|
+
Feature: Creation of a Darwing Core Archive
|
2
|
+
In order to start working with Darwin Core Archive file
|
3
|
+
A user should be able initiate dwc object from a file
|
4
|
+
So I want to implement handling of dwc object creation
|
5
|
+
|
6
|
+
Scenario: Creating Darwin Core Archive object
|
7
|
+
Given path to a dwc file "data.tar.gz"
|
8
|
+
When I create a new DarwinCore::Archive instance
|
9
|
+
Then I should find that the archive is valid
|
10
|
+
Then I should see what files the archive has
|
11
|
+
|
12
|
+
When I delete expanded files
|
13
|
+
Then they should disappear
|
14
|
+
|
15
|
+
Scenario: Instantiating DarwinCore with tar.gz file
|
16
|
+
Given path to a dwc file "data.tar.gz"
|
17
|
+
When I create a new DarwinCore instance
|
18
|
+
Then instance should have a valid archive
|
19
|
+
And instance should have a core
|
20
|
+
When I check core data
|
21
|
+
Then I should find core.properties
|
22
|
+
And core.file_path
|
23
|
+
And core.id
|
24
|
+
And core.fields
|
25
|
+
Then DarwinCore instance should have an extensions array
|
26
|
+
And every extension in array should be an instance of DarwinCore::Extension
|
27
|
+
And extension should have properties, data, file_path, coreid, fields
|
28
|
+
Then DarwinCore instance should have dwc.metadata object
|
29
|
+
And I should find id, title, creators, metadata provider
|
30
|
+
|
31
|
+
Scenario: Instantiating DawinCore with zip file
|
32
|
+
Given path to a dwc file "data.zip"
|
33
|
+
When I create a new DarwinCore instance
|
34
|
+
Then instance should have a valid archive
|
@@ -0,0 +1,95 @@
|
|
1
|
+
require 'ruby-debug'
|
2
|
+
|
3
|
+
Given /^path to a dwc file "([^\"]*)"$/ do |arg1|
|
4
|
+
@dwca_file = File.expand_path(File.dirname(__FILE__) + "../../../spec/files/" + arg1)
|
5
|
+
@tmp_dir = "/tmp"
|
6
|
+
end
|
7
|
+
|
8
|
+
When /^I create a new DarwinCore::Archive instance$/ do
|
9
|
+
@dwca = DarwinCore::Archive.new(@dwca_file, @tmp_dir)
|
10
|
+
end
|
11
|
+
|
12
|
+
Then /^I should find that the archive is valid$/ do
|
13
|
+
@dwca.valid?.should be_true
|
14
|
+
end
|
15
|
+
|
16
|
+
Then /^I should see what files the archive has$/ do
|
17
|
+
@dwca.files.should == ["DarwinCore.txt", "VernacularName.txt", "eml.xml", "leptogastrinae.xlsx", "meta.xml", "metadata.txt"]
|
18
|
+
end
|
19
|
+
|
20
|
+
When /^I delete expanded files$/ do
|
21
|
+
@dwca.clean
|
22
|
+
end
|
23
|
+
|
24
|
+
Then /^they should disappear$/ do
|
25
|
+
@dwca.files.should be_nil
|
26
|
+
end
|
27
|
+
|
28
|
+
When /^I create a new DarwinCore instance$/ do
|
29
|
+
@dwc = DarwinCore.new(@dwca_file)
|
30
|
+
end
|
31
|
+
|
32
|
+
Then /^instance should have a valid archive$/ do
|
33
|
+
@dwc.archive.valid?.should be_true
|
34
|
+
end
|
35
|
+
|
36
|
+
Then /^instance should have a core$/ do
|
37
|
+
@dwc.core.class.should == DarwinCore::Core
|
38
|
+
end
|
39
|
+
|
40
|
+
When /^I check core data$/ do
|
41
|
+
@core = @dwc.core
|
42
|
+
end
|
43
|
+
|
44
|
+
Then /^I should find core.properties$/ do
|
45
|
+
@core.properties.class.should == Hash
|
46
|
+
@core.properties[:encoding].should == "UTF-8"
|
47
|
+
@core.properties[:fieldsTerminatedBy].should == "\\t"
|
48
|
+
@core.properties[:linesTerminatedBy].should == "\\n"
|
49
|
+
end
|
50
|
+
|
51
|
+
And /^core\.file_path$/ do
|
52
|
+
@core.file_path.should match(/\/tmp\/dwc_[\d]+\/DarwinCore.txt/)
|
53
|
+
end
|
54
|
+
|
55
|
+
And /^core\.id$/ do
|
56
|
+
@core.id.should == {:index => 0, :term => 'http://rs.tdwg.org/dwc/terms/TaxonID'}
|
57
|
+
end
|
58
|
+
|
59
|
+
And /^core\.fields$/ do
|
60
|
+
@core.fields.size.should == 5
|
61
|
+
end
|
62
|
+
Then /^DarwinCore instance should have dwc\.metadata object$/ do
|
63
|
+
@dwc.metadata.class.should == DarwinCore::Metadata
|
64
|
+
end
|
65
|
+
|
66
|
+
And /^I should find id, title, creators, metadata provider$/ do
|
67
|
+
@dwc.metadata.id.should == 'leptogastrinae:version:2.5'
|
68
|
+
@dwc.metadata.title.should == 'Leptogastrinae (Diptera: Asilidae) Classification'
|
69
|
+
@dwc.metadata.authors.should == [
|
70
|
+
{:last_name=>"Bayless", :email=>"keith.bayless@gmail.com", :first_name=>"Keith"},
|
71
|
+
{:last_name=>"Dikow", :email=>"dshorthouse@eol.org", :first_name=>"Torsten"}]
|
72
|
+
@dwc.metadata.abstract.should == 'These are all the names in the Leptogastrinae classification.'
|
73
|
+
@dwc.metadata.citation.should == 'Dikow, Torsten. 2010. The Leptogastrinae classification.'
|
74
|
+
@dwc.metadata.url.should == 'http://leptogastrinae.lifedesks.org/files/leptogastrinae/classification_export/shared/leptogastrinae.tar.gz'
|
75
|
+
end
|
76
|
+
|
77
|
+
Then /^DarwinCore instance should have an extensions array$/ do
|
78
|
+
@dwc.extensions.class.should == Array
|
79
|
+
@dwc.extensions.size.should == 1
|
80
|
+
end
|
81
|
+
|
82
|
+
And /^every extension in array should be an instance of DarwinCore::Extension$/ do
|
83
|
+
classes = @dwc.extensions.map {|e| e.class}.uniq
|
84
|
+
classes.size.should == 1
|
85
|
+
classes[0].should == DarwinCore::Extension
|
86
|
+
end
|
87
|
+
|
88
|
+
Then /^extension should have properties, data, file_path, coreid, fields$/ do
|
89
|
+
ext = @dwc.extensions[0]
|
90
|
+
ext.properties.should == {:ignoreHeaderLines=>1, :encoding=>"UTF-8", :rowType=>"http://rs.gbif.org/ipt/terms/1.0/VernacularName", :fieldsEnclosedBy=>"", :fieldsTerminatedBy=>"\\t", :linesTerminatedBy=>"\\n"}
|
91
|
+
ext.data.class.should == Hash
|
92
|
+
ext.file_path.should match(/\/tmp\/dwc_[\d]+\/VernacularName.txt/)
|
93
|
+
ext.coreid.should == {:index=>0}
|
94
|
+
ext.fields.should == [{:term=>"http://rs.gbif.org/ecat/terms/vernacularName", :index=>1}, {:term=>"http://rs.gbif.org/thesaurus/languageCode", :index=>2}]
|
95
|
+
end
|
data/lib/dwc-archive.rb
ADDED
@@ -0,0 +1,29 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
$:.unshift(File.dirname(__FILE__)) unless
|
3
|
+
$:.include?(File.dirname(__FILE__)) || $:.include?(File.expand_path(File.dirname(__FILE__)))
|
4
|
+
require 'ruby_extensions'
|
5
|
+
require 'dwc-archive/expander'
|
6
|
+
require 'dwc-archive/archive'
|
7
|
+
require 'dwc-archive/core'
|
8
|
+
require 'dwc-archive/extension'
|
9
|
+
require 'dwc-archive/metadata'
|
10
|
+
|
11
|
+
class DarwinCore
|
12
|
+
attr_reader :archive, :core, :metadata, :extensions
|
13
|
+
alias :eml :metadata
|
14
|
+
def initialize(dwc_path, tmp_dir = "/tmp")
|
15
|
+
@archive = DarwinCore::Archive.new(dwc_path, tmp_dir)
|
16
|
+
@core = DarwinCore::Core.new(@archive)
|
17
|
+
@metadata = DarwinCore::Metadata.new(@archive)
|
18
|
+
@extensions = get_extensions
|
19
|
+
end
|
20
|
+
private
|
21
|
+
def get_extensions
|
22
|
+
res = []
|
23
|
+
root_key = @archive.meta.keys[0]
|
24
|
+
ext = @archive.meta[root_key][:extension]
|
25
|
+
return [] unless ext
|
26
|
+
ext = [ext] unless ext.class == Array
|
27
|
+
ext.map {|e| DarwinCore::Extension.new(@archive, e)}
|
28
|
+
end
|
29
|
+
end
|
Binary file
|
@@ -0,0 +1,37 @@
|
|
1
|
+
require 'nokogiri'
|
2
|
+
class DarwinCore
|
3
|
+
class Archive
|
4
|
+
attr_reader :meta, :eml
|
5
|
+
def initialize(archive_path, tmp_dir)
|
6
|
+
@archive_path = archive_path
|
7
|
+
@tmp_dir = tmp_dir
|
8
|
+
@expander = DarwinCore::Expander.new(@archive_path, @tmp_dir)
|
9
|
+
@expander.unpack
|
10
|
+
if valid?
|
11
|
+
@meta = Hash.from_xml(open(File.join(@expander.path, 'meta.xml')))
|
12
|
+
@eml = files.include?("eml.xml") ? Hash.from_xml(open(File.join(@expander.path, 'eml.xml'))) : nil
|
13
|
+
else
|
14
|
+
clean
|
15
|
+
raise 'not a valid Darwin Core Archive File'
|
16
|
+
end
|
17
|
+
end
|
18
|
+
|
19
|
+
def valid?
|
20
|
+
valid = true
|
21
|
+
valid = valid && FileTest.exists?(@archive_path)
|
22
|
+
valid = valid && files && files.include?('meta.xml')
|
23
|
+
end
|
24
|
+
|
25
|
+
def files
|
26
|
+
@expander.files
|
27
|
+
end
|
28
|
+
|
29
|
+
def files_path
|
30
|
+
@expander.path
|
31
|
+
end
|
32
|
+
|
33
|
+
def clean
|
34
|
+
@expander.clean
|
35
|
+
end
|
36
|
+
end
|
37
|
+
end
|
@@ -0,0 +1,34 @@
|
|
1
|
+
class DarwinCore
|
2
|
+
class Core
|
3
|
+
def initialize(archive)
|
4
|
+
@archive = archive
|
5
|
+
@path = @archive.files_path
|
6
|
+
root_key = @archive.meta.keys[0]
|
7
|
+
@core = @archive.meta[root_key][:core]
|
8
|
+
raise "Cannot found core in meta.xml, is meta.xml valid?" unless @core
|
9
|
+
end
|
10
|
+
|
11
|
+
def data
|
12
|
+
@core
|
13
|
+
end
|
14
|
+
|
15
|
+
def properties
|
16
|
+
@core[:attributes]
|
17
|
+
end
|
18
|
+
|
19
|
+
def file_path
|
20
|
+
file = @core[:files][:location] || @core[:location]
|
21
|
+
File.join(@path, file)
|
22
|
+
end
|
23
|
+
|
24
|
+
def id
|
25
|
+
@core[:id][:attributes]
|
26
|
+
end
|
27
|
+
|
28
|
+
def fields
|
29
|
+
@core[:field] = [@core[:field]] unless @core[:field].class == Array
|
30
|
+
@core[:field].map {|f| f[:attributes]}
|
31
|
+
end
|
32
|
+
|
33
|
+
end
|
34
|
+
end
|
@@ -0,0 +1,71 @@
|
|
1
|
+
class DarwinCore
|
2
|
+
class Expander
|
3
|
+
def initialize(archive_path, tmp_dir)
|
4
|
+
@archive_path = archive_path
|
5
|
+
@tmp_dir = tmp_dir
|
6
|
+
@path = File.join(tmp_dir, 'dwc_' + rand(10000000000).to_s)
|
7
|
+
@unpacker = get_unpacker
|
8
|
+
end
|
9
|
+
|
10
|
+
def unpack
|
11
|
+
clean
|
12
|
+
@unpacker.call(@path, @archive_path) if @unpacker
|
13
|
+
end
|
14
|
+
|
15
|
+
def path
|
16
|
+
@files_path ||= files_path
|
17
|
+
end
|
18
|
+
|
19
|
+
def clean
|
20
|
+
FileUtils.rm_rf(@path) if FileTest.exists?(@path)
|
21
|
+
end
|
22
|
+
|
23
|
+
def files
|
24
|
+
return nil unless path && FileTest.exists?(path)
|
25
|
+
Dir.entries(path).select {|e| e !~ /[\.]{1,2}$/}.sort
|
26
|
+
end
|
27
|
+
|
28
|
+
private
|
29
|
+
def get_unpacker
|
30
|
+
file_type = IO.popen("file -z " + @archive_path).read
|
31
|
+
|
32
|
+
if file_type.match(/tar.*gzip/i)
|
33
|
+
return proc do |tmp_path, archive_path|
|
34
|
+
FileUtils.mkdir tmp_path
|
35
|
+
system "tar -zxvf #{archive_path} -C #{tmp_path}"
|
36
|
+
end
|
37
|
+
end
|
38
|
+
|
39
|
+
if file_type.match(/Zip/)
|
40
|
+
return proc { |tmp_path, archive_path| system "unzip -qq -d #{tmp_path} #{archive_path}" }
|
41
|
+
end
|
42
|
+
|
43
|
+
return nil
|
44
|
+
end
|
45
|
+
|
46
|
+
def path_entries(dir)
|
47
|
+
Dir.entries(dir).select {|e| e !~ /[\.]{1,2}$/}.sort
|
48
|
+
end
|
49
|
+
|
50
|
+
def files_path
|
51
|
+
res = nil
|
52
|
+
entries = path_entries(@path)
|
53
|
+
if entries.include?('meta.xml')
|
54
|
+
res = @path
|
55
|
+
else
|
56
|
+
entries.each do |e|
|
57
|
+
check_path = File.join(@path, e)
|
58
|
+
if FileTest.directory?(check_path)
|
59
|
+
if path_entries(check_path).include?('meta.xml')
|
60
|
+
res = check_path
|
61
|
+
break
|
62
|
+
end
|
63
|
+
end
|
64
|
+
end
|
65
|
+
end
|
66
|
+
res
|
67
|
+
end
|
68
|
+
end
|
69
|
+
|
70
|
+
|
71
|
+
end
|
@@ -0,0 +1,31 @@
|
|
1
|
+
class DarwinCore
|
2
|
+
class Extension
|
3
|
+
def initialize(archive, extension)
|
4
|
+
@archive = archive
|
5
|
+
@path = @archive.files_path
|
6
|
+
@extension = extension
|
7
|
+
end
|
8
|
+
|
9
|
+
def data
|
10
|
+
@extension
|
11
|
+
end
|
12
|
+
|
13
|
+
def properties
|
14
|
+
@extension[:attributes]
|
15
|
+
end
|
16
|
+
|
17
|
+
def file_path
|
18
|
+
file = @extension[:files][:location]
|
19
|
+
File.join(@path, file)
|
20
|
+
end
|
21
|
+
|
22
|
+
def coreid
|
23
|
+
@extension[:coreid][:attributes]
|
24
|
+
end
|
25
|
+
|
26
|
+
def fields
|
27
|
+
@extension[:field] = [@extension[:field]] unless @extension[:field].class == Array
|
28
|
+
@extension[:field].map {|f| f[:attributes]}
|
29
|
+
end
|
30
|
+
end
|
31
|
+
end
|
@@ -0,0 +1,38 @@
|
|
1
|
+
class DarwinCore
|
2
|
+
class Metadata
|
3
|
+
def initialize(archive)
|
4
|
+
@archive = archive
|
5
|
+
@metadata = @archive.eml
|
6
|
+
end
|
7
|
+
|
8
|
+
def data
|
9
|
+
@metadata
|
10
|
+
end
|
11
|
+
|
12
|
+
def id
|
13
|
+
@metadata[:eml][:dataset][:attributes][:id]
|
14
|
+
end
|
15
|
+
|
16
|
+
def title
|
17
|
+
@metadata[:eml][:dataset][:title]
|
18
|
+
end
|
19
|
+
|
20
|
+
def authors
|
21
|
+
return nil unless defined?(@metadata[:eml][:dataset][:creator])
|
22
|
+
@metadata[:eml][:dataset][:creator] = [@metadata[:eml][:dataset][:creator]] unless @metadata[:eml][:dataset][:creator].class == Array
|
23
|
+
@metadata[:eml][:dataset][:creator].map {|c| {:first_name => c[:individualName][:givenName], :last_name => c[:individualName][:surName], :email => c[:electronicMailAddress]}}
|
24
|
+
end
|
25
|
+
|
26
|
+
def abstract
|
27
|
+
@metadata[:eml][:dataset][:abstract] rescue nil
|
28
|
+
end
|
29
|
+
|
30
|
+
def citation
|
31
|
+
@metadata[:eml][:additionalMetadata][:metadata][:citation] rescue nil
|
32
|
+
end
|
33
|
+
|
34
|
+
def url
|
35
|
+
@metadata[:eml][:dataset][:distribution][:online][:url] rescue nil
|
36
|
+
end
|
37
|
+
end
|
38
|
+
end
|
@@ -0,0 +1,64 @@
|
|
1
|
+
# USAGE: Hash.from_xml:(YOUR_XML_STRING)
|
2
|
+
require 'nokogiri'
|
3
|
+
# modified from http://stackoverflow.com/questions/1230741/convert-a-nokogiri-document-to-a-ruby-hash/1231297#1231297
|
4
|
+
|
5
|
+
class Hash
|
6
|
+
class << self
|
7
|
+
def from_xml(xml_io)
|
8
|
+
begin
|
9
|
+
result = Nokogiri::XML(xml_io)
|
10
|
+
return { result.root.name.to_sym => xml_node_to_hash(result.root)}
|
11
|
+
rescue Exception => e
|
12
|
+
# raise your custom exception here
|
13
|
+
end
|
14
|
+
end
|
15
|
+
|
16
|
+
def xml_node_to_hash(node)
|
17
|
+
# If we are at the root of the document, start the hash
|
18
|
+
if node.element?
|
19
|
+
result_hash = {}
|
20
|
+
if node.attributes != {}
|
21
|
+
result_hash[:attributes] = {}
|
22
|
+
node.attributes.keys.each do |key|
|
23
|
+
result_hash[:attributes][node.attributes[key].name.to_sym] = prepare(node.attributes[key].value)
|
24
|
+
end
|
25
|
+
end
|
26
|
+
if node.children.size > 0
|
27
|
+
node.children.each do |child|
|
28
|
+
result = xml_node_to_hash(child)
|
29
|
+
|
30
|
+
if child.name == "text"
|
31
|
+
unless child.next_sibling || child.previous_sibling
|
32
|
+
return prepare(result)
|
33
|
+
end
|
34
|
+
elsif result_hash[child.name.to_sym]
|
35
|
+
if result_hash[child.name.to_sym].is_a?(Object::Array)
|
36
|
+
result_hash[child.name.to_sym] << prepare(result)
|
37
|
+
else
|
38
|
+
result_hash[child.name.to_sym] = [result_hash[child.name.to_sym]] << prepare(result)
|
39
|
+
end
|
40
|
+
else
|
41
|
+
result_hash[child.name.to_sym] = prepare(result)
|
42
|
+
end
|
43
|
+
end
|
44
|
+
|
45
|
+
return result_hash
|
46
|
+
else
|
47
|
+
return result_hash
|
48
|
+
end
|
49
|
+
else
|
50
|
+
return prepare(node.content.to_s)
|
51
|
+
end
|
52
|
+
end
|
53
|
+
|
54
|
+
def prepare(data)
|
55
|
+
(data.class == String && data.to_i.to_s == data) ? data.to_i : data
|
56
|
+
end
|
57
|
+
end
|
58
|
+
|
59
|
+
def to_struct(struct_name)
|
60
|
+
Struct.new(struct_name,*keys).new(*values)
|
61
|
+
end
|
62
|
+
end
|
63
|
+
|
64
|
+
|
Binary file
|
data/spec/files/data.zip
ADDED
Binary file
|
data/spec/files/eml.xml
ADDED
@@ -0,0 +1,46 @@
|
|
1
|
+
<?xml version="1.0" encoding="UTF-8"?>
|
2
|
+
<eml:eml
|
3
|
+
packageId="eml.1.1" system="knb"
|
4
|
+
xmlns:eml="eml://ecoinformatics.org/eml-2.1.0"
|
5
|
+
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
|
6
|
+
xsi:schemaLocation="eml://ecoinformatics.org/eml-2.1.0 eml.xsd">
|
7
|
+
<dataset id="leptogastrinae:version:2.5">
|
8
|
+
<title>Leptogastrinae (Diptera: Asilidae) Classification</title>
|
9
|
+
<creator id="10" scope="document">
|
10
|
+
<individualName>
|
11
|
+
<givenName>Keith</givenName>
|
12
|
+
<surName>Bayless</surName>
|
13
|
+
</individualName>
|
14
|
+
<electronicMailAddress>keith.bayless@gmail.com</electronicMailAddress>
|
15
|
+
</creator>
|
16
|
+
<creator id="5" scope="document">
|
17
|
+
<individualName>
|
18
|
+
<givenName>Torsten</givenName>
|
19
|
+
<surName>Dikow</surName>
|
20
|
+
</individualName>
|
21
|
+
<electronicMailAddress>dshorthouse@eol.org</electronicMailAddress>
|
22
|
+
</creator>
|
23
|
+
<metadataProvider>
|
24
|
+
<organizationName>Encyclopedia of Life: LifeDesks (http://www.lifedesks.org)</organizationName>
|
25
|
+
</metadataProvider>
|
26
|
+
<pubDate>2010-02-03T01:09:41-05:00</pubDate>
|
27
|
+
<abstract>These are all the names in the Leptogastrinae classification.</abstract>
|
28
|
+
<intellectualRights>Creative Commons: publicdomain</intellectualRights>
|
29
|
+
<distribution>
|
30
|
+
<online>
|
31
|
+
<url function="download">http://leptogastrinae.lifedesks.org/files/leptogastrinae/classification_export/shared/leptogastrinae.tar.gz</url>
|
32
|
+
</online>
|
33
|
+
</distribution>
|
34
|
+
<contact>
|
35
|
+
<references>5</references>
|
36
|
+
</contact>
|
37
|
+
<publisher>
|
38
|
+
<organizationName>The Marine Biological Laboratory</organizationName>
|
39
|
+
</publisher>
|
40
|
+
</dataset>
|
41
|
+
<additionalMetadata>
|
42
|
+
<metadata>
|
43
|
+
<citation>Dikow, Torsten. 2010. The Leptogastrinae classification.</citation>
|
44
|
+
</metadata>
|
45
|
+
</additionalMetadata>
|
46
|
+
</eml:eml>
|
data/spec/files/meta.xml
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
<?xml version="1.0" encoding="UTF-8"?>
|
2
|
+
<archive xmlns="http://rs.tdwg.org/dwc/text/">
|
3
|
+
<core encoding="UTF-8" fieldsTerminatedBy="\t" linesTerminatedBy="\n" fieldsEnclosedBy="" ignoreHeaderLines="1" rowType="http://rs.tdwg.org/dwc/terms/DarwinCore">
|
4
|
+
<files>
|
5
|
+
<location>DarwinCore.txt</location>
|
6
|
+
</files>
|
7
|
+
<id index="0" term="http://rs.tdwg.org/dwc/terms/TaxonID"/>
|
8
|
+
<field index="1" term="http://purl.org/dc/terms/source"/>
|
9
|
+
<field index="2" term="http://rs.tdwg.org/dwc/terms/ScientificName"/>
|
10
|
+
<field index="3" term="http://rs.tdwg.org/dwc/terms/HigherTaxonID"/>
|
11
|
+
<field index="4" term="http://rs.tdwg.org/dwc/terms/TaxonRank"/>
|
12
|
+
<field index="5" term="http://rs.tdwg.org/dwc/terms/TaxonomicStatus"/>
|
13
|
+
</core>
|
14
|
+
<extension encoding="UTF-8" fieldsTerminatedBy="\t" linesTerminatedBy="\n" fieldsEnclosedBy="" ignoreHeaderLines="1" rowType="http://rs.gbif.org/ipt/terms/1.0/VernacularName">
|
15
|
+
<files>
|
16
|
+
<location>VernacularName.txt</location>
|
17
|
+
</files>
|
18
|
+
<coreid index="0"/>
|
19
|
+
<field index="1" term="http://rs.gbif.org/ecat/terms/vernacularName"/>
|
20
|
+
<field index="2" term="http://rs.gbif.org/thesaurus/languageCode"/>
|
21
|
+
</extension>
|
22
|
+
</archive>
|
@@ -0,0 +1,15 @@
|
|
1
|
+
require File.dirname(__FILE__) + "/../spec_helper"
|
2
|
+
|
3
|
+
describe "Hash" do
|
4
|
+
it "should parse xml to hash" do
|
5
|
+
Hash.public_methods.include?("from_xml").should be_true
|
6
|
+
end
|
7
|
+
|
8
|
+
it "should parse xml" do
|
9
|
+
xml_string = open(File.dirname(__FILE__) + "/../files/meta.xml").read
|
10
|
+
meta = Hash.from_xml(xml_string)
|
11
|
+
meta[:archive].keys.map {|k| k.to_s}.sort.should == %w(core extension)
|
12
|
+
meta[:archive][:core].keys.map{|k| k.to_s}.sort.should == ["attributes", "field", "files", "id"]
|
13
|
+
end
|
14
|
+
|
15
|
+
end
|
data/spec/spec.opts
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
--color
|
data/spec/spec_helper.rb
ADDED
metadata
ADDED
@@ -0,0 +1,115 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: dwc-archive
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
prerelease: false
|
5
|
+
segments:
|
6
|
+
- 0
|
7
|
+
- 1
|
8
|
+
- 2
|
9
|
+
version: 0.1.2
|
10
|
+
platform: ruby
|
11
|
+
authors:
|
12
|
+
- Dmitry Mozzherin
|
13
|
+
autorequire:
|
14
|
+
bindir: bin
|
15
|
+
cert_chain: []
|
16
|
+
|
17
|
+
date: 2010-03-18 00:00:00 -04:00
|
18
|
+
default_executable:
|
19
|
+
dependencies:
|
20
|
+
- !ruby/object:Gem::Dependency
|
21
|
+
name: rspec
|
22
|
+
prerelease: false
|
23
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
24
|
+
requirements:
|
25
|
+
- - ">="
|
26
|
+
- !ruby/object:Gem::Version
|
27
|
+
segments:
|
28
|
+
- 1
|
29
|
+
- 2
|
30
|
+
- 9
|
31
|
+
version: 1.2.9
|
32
|
+
type: :development
|
33
|
+
version_requirements: *id001
|
34
|
+
- !ruby/object:Gem::Dependency
|
35
|
+
name: cucumber
|
36
|
+
prerelease: false
|
37
|
+
requirement: &id002 !ruby/object:Gem::Requirement
|
38
|
+
requirements:
|
39
|
+
- - ">="
|
40
|
+
- !ruby/object:Gem::Version
|
41
|
+
segments:
|
42
|
+
- 0
|
43
|
+
version: "0"
|
44
|
+
type: :development
|
45
|
+
version_requirements: *id002
|
46
|
+
description: Darwin Core Archive Files are current standard exchange format for GLobal Names Architecture modules. This project creates ways to work with such files.
|
47
|
+
email: dmozzherin at gmail dot com
|
48
|
+
executables: []
|
49
|
+
|
50
|
+
extensions: []
|
51
|
+
|
52
|
+
extra_rdoc_files:
|
53
|
+
- LICENSE
|
54
|
+
- README.rdoc
|
55
|
+
files:
|
56
|
+
- .document
|
57
|
+
- .gitignore
|
58
|
+
- LICENSE
|
59
|
+
- README.rdoc
|
60
|
+
- Rakefile
|
61
|
+
- VERSION
|
62
|
+
- dwc-archive.gemspec
|
63
|
+
- features/dwc-archive.feature
|
64
|
+
- features/step_definitions/dwc-archive_steps.rb
|
65
|
+
- features/support/env.rb
|
66
|
+
- lib/dwc-archive.rb
|
67
|
+
- lib/dwc-archive/.expander.rb.swo
|
68
|
+
- lib/dwc-archive/archive.rb
|
69
|
+
- lib/dwc-archive/core.rb
|
70
|
+
- lib/dwc-archive/expander.rb
|
71
|
+
- lib/dwc-archive/extension.rb
|
72
|
+
- lib/dwc-archive/metadata.rb
|
73
|
+
- lib/ruby_extensions.rb
|
74
|
+
- spec/dwc-archive_spec.rb
|
75
|
+
- spec/files/data.tar.gz
|
76
|
+
- spec/files/data.zip
|
77
|
+
- spec/files/eml.xml
|
78
|
+
- spec/files/meta.xml
|
79
|
+
- spec/lib/ruby_extenstions_spec.rb
|
80
|
+
- spec/spec.opts
|
81
|
+
- spec/spec_helper.rb
|
82
|
+
has_rdoc: true
|
83
|
+
homepage: http://github.com/dimus/dwc-archive
|
84
|
+
licenses: []
|
85
|
+
|
86
|
+
post_install_message:
|
87
|
+
rdoc_options:
|
88
|
+
- --charset=UTF-8
|
89
|
+
require_paths:
|
90
|
+
- lib
|
91
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
92
|
+
requirements:
|
93
|
+
- - ">="
|
94
|
+
- !ruby/object:Gem::Version
|
95
|
+
segments:
|
96
|
+
- 0
|
97
|
+
version: "0"
|
98
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
99
|
+
requirements:
|
100
|
+
- - ">="
|
101
|
+
- !ruby/object:Gem::Version
|
102
|
+
segments:
|
103
|
+
- 0
|
104
|
+
version: "0"
|
105
|
+
requirements: []
|
106
|
+
|
107
|
+
rubyforge_project:
|
108
|
+
rubygems_version: 1.3.6
|
109
|
+
signing_key:
|
110
|
+
specification_version: 3
|
111
|
+
summary: Handler of Darwin Core Archive files
|
112
|
+
test_files:
|
113
|
+
- spec/dwc-archive_spec.rb
|
114
|
+
- spec/lib/ruby_extenstions_spec.rb
|
115
|
+
- spec/spec_helper.rb
|