datacatalog-importer 0.1.5 → 0.1.6
Sign up to get free protection for your applications and to get access to all the features.
- data/VERSION +1 -1
- data/datacatalog-importer.gemspec +3 -3
- data/lib/pusher.rb +44 -3
- data/lib/utility.rb +1 -1
- metadata +28 -13
data/VERSION
CHANGED
@@ -1 +1 @@
|
|
1
|
-
0.1.
|
1
|
+
0.1.6
|
@@ -5,11 +5,11 @@
|
|
5
5
|
|
6
6
|
Gem::Specification.new do |s|
|
7
7
|
s.name = %q{datacatalog-importer}
|
8
|
-
s.version = "0.1.
|
8
|
+
s.version = "0.1.6"
|
9
9
|
|
10
10
|
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
11
|
s.authors = ["David James"]
|
12
|
-
s.date = %q{2010-03-
|
12
|
+
s.date = %q{2010-03-05}
|
13
13
|
s.description = %q{This framework makes it easier to write importers for the National Data Catalog.}
|
14
14
|
s.email = %q{djames@sunlightfoundation.com}
|
15
15
|
s.extra_rdoc_files = [
|
@@ -39,7 +39,7 @@ Gem::Specification.new do |s|
|
|
39
39
|
s.homepage = %q{http://github.com/djsun/datacatalog-importer}
|
40
40
|
s.rdoc_options = ["--charset=UTF-8"]
|
41
41
|
s.require_paths = ["lib"]
|
42
|
-
s.rubygems_version = %q{1.3.
|
42
|
+
s.rubygems_version = %q{1.3.6}
|
43
43
|
s.summary = %q{A framework to write National Data Catalog importers}
|
44
44
|
s.test_files = [
|
45
45
|
"spec/datacatalog-importer_spec.rb",
|
data/lib/pusher.rb
CHANGED
@@ -34,12 +34,14 @@ module DataCatalog
|
|
34
34
|
|
35
35
|
def push_organizations
|
36
36
|
read_data(:organization) do |data|
|
37
|
+
link_to_existing_organization!(data, :parent_id)
|
37
38
|
create_or_update_organization(data)
|
38
39
|
end
|
39
40
|
end
|
40
41
|
|
41
42
|
def push_sources
|
42
43
|
read_data(:source) do |data|
|
44
|
+
link_to_existing_organization!(data, :organization_id)
|
43
45
|
source = create_or_update_source(data)
|
44
46
|
data[:downloads].each do |download_data|
|
45
47
|
create_or_update_download(source, download_data)
|
@@ -68,25 +70,42 @@ module DataCatalog
|
|
68
70
|
# ---
|
69
71
|
|
70
72
|
def create_or_update_organization(data)
|
71
|
-
|
73
|
+
url, name = data[:url], data[:name]
|
74
|
+
raise "#{name} has blank URL" if url.blank?
|
75
|
+
docs = DataCatalog::Organization.all(:url => url)
|
72
76
|
n = docs.length
|
73
77
|
case n
|
74
78
|
when 0
|
75
|
-
puts "Creating Organization: #{
|
79
|
+
puts "Creating Organization: #{name}"
|
76
80
|
DataCatalog::Organization.create(data)
|
77
81
|
when 1
|
78
|
-
puts "Updating Organization: #{
|
82
|
+
puts "Updating Organization: #{name}"
|
79
83
|
DataCatalog::Organization.update(docs[0].id, data)
|
80
84
|
else
|
81
85
|
multiple_matches("Organization", n, data)
|
82
86
|
end
|
83
87
|
end
|
84
88
|
|
89
|
+
def find_organization_by(field, name)
|
90
|
+
docs = DataCatalog::Organization.all(field => name)
|
91
|
+
n = docs.length
|
92
|
+
case n
|
93
|
+
when 0
|
94
|
+
nil
|
95
|
+
when 1
|
96
|
+
docs[0]
|
97
|
+
else
|
98
|
+
multiple_matches("Organization", n, { field => name })
|
99
|
+
nil
|
100
|
+
end
|
101
|
+
end
|
102
|
+
|
85
103
|
# Important: do not modify data
|
86
104
|
def create_or_update_source(data)
|
87
105
|
data = data.reject do |key, value|
|
88
106
|
[:organization, :downloads].include?(key)
|
89
107
|
end
|
108
|
+
puts "data[:url] : #{data[:url]}"
|
90
109
|
docs = DataCatalog::Source.all(:url => data[:url])
|
91
110
|
n = docs.length
|
92
111
|
case n
|
@@ -121,6 +140,28 @@ module DataCatalog
|
|
121
140
|
end
|
122
141
|
end
|
123
142
|
|
143
|
+
# Try to link to an existing organization, first by
|
144
|
+
# using an URL, then by name.
|
145
|
+
#
|
146
|
+
# Note: modifies data (that is why I use the !)
|
147
|
+
def link_to_existing_organization!(data, organization_id_key)
|
148
|
+
organization_hash = data.delete(:organization)
|
149
|
+
raise "Could not find :organization key" unless organization_hash
|
150
|
+
name = organization_hash[:name]
|
151
|
+
url = organization_hash[:url]
|
152
|
+
organization = if url
|
153
|
+
find_organization_by(:url, url)
|
154
|
+
elsif name
|
155
|
+
find_organization_by(:name, name)
|
156
|
+
end
|
157
|
+
|
158
|
+
if organization
|
159
|
+
data[organization_id_key] = organization.id
|
160
|
+
else
|
161
|
+
puts " - Could not find organization named: #{name}"
|
162
|
+
end
|
163
|
+
end
|
164
|
+
|
124
165
|
# ---
|
125
166
|
|
126
167
|
def multiple_matches(model, n, data)
|
data/lib/utility.rb
CHANGED
metadata
CHANGED
@@ -1,7 +1,12 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: datacatalog-importer
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
|
4
|
+
prerelease: false
|
5
|
+
segments:
|
6
|
+
- 0
|
7
|
+
- 1
|
8
|
+
- 6
|
9
|
+
version: 0.1.6
|
5
10
|
platform: ruby
|
6
11
|
authors:
|
7
12
|
- David James
|
@@ -9,29 +14,37 @@ autorequire:
|
|
9
14
|
bindir: bin
|
10
15
|
cert_chain: []
|
11
16
|
|
12
|
-
date: 2010-03-
|
17
|
+
date: 2010-03-05 00:00:00 -05:00
|
13
18
|
default_executable:
|
14
19
|
dependencies:
|
15
20
|
- !ruby/object:Gem::Dependency
|
16
21
|
name: nokogiri
|
17
|
-
|
18
|
-
|
19
|
-
version_requirements: !ruby/object:Gem::Requirement
|
22
|
+
prerelease: false
|
23
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
20
24
|
requirements:
|
21
25
|
- - ">="
|
22
26
|
- !ruby/object:Gem::Version
|
27
|
+
segments:
|
28
|
+
- 1
|
29
|
+
- 4
|
30
|
+
- 1
|
23
31
|
version: 1.4.1
|
24
|
-
|
32
|
+
type: :runtime
|
33
|
+
version_requirements: *id001
|
25
34
|
- !ruby/object:Gem::Dependency
|
26
35
|
name: rspec
|
27
|
-
|
28
|
-
|
29
|
-
version_requirements: !ruby/object:Gem::Requirement
|
36
|
+
prerelease: false
|
37
|
+
requirement: &id002 !ruby/object:Gem::Requirement
|
30
38
|
requirements:
|
31
39
|
- - ">="
|
32
40
|
- !ruby/object:Gem::Version
|
41
|
+
segments:
|
42
|
+
- 1
|
43
|
+
- 2
|
44
|
+
- 9
|
33
45
|
version: 1.2.9
|
34
|
-
|
46
|
+
type: :development
|
47
|
+
version_requirements: *id002
|
35
48
|
description: This framework makes it easier to write importers for the National Data Catalog.
|
36
49
|
email: djames@sunlightfoundation.com
|
37
50
|
executables: []
|
@@ -73,18 +86,20 @@ required_ruby_version: !ruby/object:Gem::Requirement
|
|
73
86
|
requirements:
|
74
87
|
- - ">="
|
75
88
|
- !ruby/object:Gem::Version
|
89
|
+
segments:
|
90
|
+
- 0
|
76
91
|
version: "0"
|
77
|
-
version:
|
78
92
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
79
93
|
requirements:
|
80
94
|
- - ">="
|
81
95
|
- !ruby/object:Gem::Version
|
96
|
+
segments:
|
97
|
+
- 0
|
82
98
|
version: "0"
|
83
|
-
version:
|
84
99
|
requirements: []
|
85
100
|
|
86
101
|
rubyforge_project:
|
87
|
-
rubygems_version: 1.3.
|
102
|
+
rubygems_version: 1.3.6
|
88
103
|
signing_key:
|
89
104
|
specification_version: 3
|
90
105
|
summary: A framework to write National Data Catalog importers
|