turbot-runner 0.2.7 → 0.2.8
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.
- checksums.yaml +7 -0
- data/README.md +18 -0
- data/lib/turbot_runner/version.rb +1 -1
- data/schema/schemas/gazette-notice-schema.json +331 -0
- metadata +43 -47
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: e2229c474e1e419b9bc6f665f23a2fcc08ffc85c
|
4
|
+
data.tar.gz: 0dea65843d3cbb38736dac5de99ddf0f360cde4f
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 30f4b3241b3d21780add84ed119d904d4ce764433dfe3d4493d1c2771bc196deb04d2a3a4a19d4db5fa8839de158e1689cc0690151b1ad210a19bf44c248ffc3
|
7
|
+
data.tar.gz: a9b9509032f3091ddfb01b53f08081f90cb2882c2498fe2525dae2ca7f46454fb12a83b7e47e9661be8778809b069dfedb345aeb5cb2b77f76e2c8d89e205946
|
data/README.md
ADDED
@@ -0,0 +1,18 @@
|
|
1
|
+
# turbot-runner
|
2
|
+
|
3
|
+
## Getting started
|
4
|
+
|
5
|
+
git submodule update --init
|
6
|
+
cd schema && git checkout master && cd ..
|
7
|
+
|
8
|
+
## Updating the schema
|
9
|
+
|
10
|
+
cd schema && git pull --rebase && cd ..
|
11
|
+
git commit schema -m 'Pull in new schema'
|
12
|
+
|
13
|
+
## Releasing a new version
|
14
|
+
|
15
|
+
Bump the version in `lib/turbot_runner/version.rb`, then:
|
16
|
+
|
17
|
+
git commit lib/turbot_runner/version.rb -m 'Release new version'
|
18
|
+
bundle exec rake release # requires Rubygems credentials
|
@@ -0,0 +1,331 @@
|
|
1
|
+
{
|
2
|
+
"$schema": "http://json-schema.org/draft-04/schema#",
|
3
|
+
"title": "Notice",
|
4
|
+
"description": "A gazette notice",
|
5
|
+
"type": "object",
|
6
|
+
"properties": {
|
7
|
+
"issue": {
|
8
|
+
"description": "The issue in which the notice is published",
|
9
|
+
"$ref": "#/definitions/issue"
|
10
|
+
},
|
11
|
+
"document": {
|
12
|
+
"$ref": "#/definitions/document"
|
13
|
+
},
|
14
|
+
"uid": {
|
15
|
+
"description": "An internal, constructed, unique identifier",
|
16
|
+
"type": "string"
|
17
|
+
},
|
18
|
+
"identifier": {
|
19
|
+
"description": "The publisher's identifier for the notice",
|
20
|
+
"type": [
|
21
|
+
"integer",
|
22
|
+
"string"
|
23
|
+
]
|
24
|
+
},
|
25
|
+
"title": {
|
26
|
+
"description": "The title of the notice",
|
27
|
+
"type": "string"
|
28
|
+
},
|
29
|
+
"description": {
|
30
|
+
"description": "A description of the notice",
|
31
|
+
"type": "string"
|
32
|
+
},
|
33
|
+
"body": {
|
34
|
+
"description": "The content of the notice",
|
35
|
+
"$ref": "#/definitions/body"
|
36
|
+
},
|
37
|
+
"date_published": {
|
38
|
+
"description": "The date of publication of the notice",
|
39
|
+
"type": "string",
|
40
|
+
"format": "date"
|
41
|
+
},
|
42
|
+
"classification": {
|
43
|
+
"description": "The type of the notice",
|
44
|
+
"type": "array",
|
45
|
+
"items": {
|
46
|
+
"$ref": "#/definitions/classification"
|
47
|
+
}
|
48
|
+
},
|
49
|
+
"subjects": {
|
50
|
+
"description": "The entities that the notice is about",
|
51
|
+
"type": "array",
|
52
|
+
"items": {
|
53
|
+
"$ref": "includes/entity.json"
|
54
|
+
}
|
55
|
+
},
|
56
|
+
"about": {
|
57
|
+
"description": "The subject matter of the notice",
|
58
|
+
"type": "object",
|
59
|
+
"properties": {
|
60
|
+
"kind": {
|
61
|
+
"description": "The class of subject matter",
|
62
|
+
"type": "string"
|
63
|
+
},
|
64
|
+
"classification": {
|
65
|
+
"description": "The type of subject matter",
|
66
|
+
"type": "array",
|
67
|
+
"items": {
|
68
|
+
"$ref": "#/definitions/classification"
|
69
|
+
}
|
70
|
+
},
|
71
|
+
"body": {
|
72
|
+
"description": "A description of the item",
|
73
|
+
"$ref": "#/definitions/body"
|
74
|
+
}
|
75
|
+
}
|
76
|
+
},
|
77
|
+
"url": {
|
78
|
+
"description": "The URL of the notice",
|
79
|
+
"type": "string",
|
80
|
+
"format": "uri"
|
81
|
+
},
|
82
|
+
"media_type": {
|
83
|
+
"description": "The media type of the URL resource",
|
84
|
+
"type": "string"
|
85
|
+
},
|
86
|
+
"creator": {
|
87
|
+
"description": "The entity placing the notice",
|
88
|
+
"$ref": "includes/entity.json"
|
89
|
+
},
|
90
|
+
"publisher": {
|
91
|
+
"description": "The publisher of the notice",
|
92
|
+
"$ref": "#/definitions/publisher"
|
93
|
+
},
|
94
|
+
"update_action": {
|
95
|
+
"description": "Whether this notice corrects or cancels an earlier notice",
|
96
|
+
"$ref": "#/definitions/update_action"
|
97
|
+
},
|
98
|
+
"source_url": {
|
99
|
+
"description": "A URL where this fact may be verified",
|
100
|
+
"type": "string",
|
101
|
+
"format": "uri"
|
102
|
+
},
|
103
|
+
"sample_date": {
|
104
|
+
"description": "A date on which this fact is known to be true",
|
105
|
+
"type": "string",
|
106
|
+
"format": "date"
|
107
|
+
},
|
108
|
+
"retrieved_at": {
|
109
|
+
"description": "The time at which the notice was retrieved",
|
110
|
+
"type": "string",
|
111
|
+
"format": "date-time"
|
112
|
+
},
|
113
|
+
"confidence": {
|
114
|
+
"description": "The aggregator's confidence in the accuracy of the data",
|
115
|
+
"type": "string",
|
116
|
+
"enum": [
|
117
|
+
"HIGH",
|
118
|
+
"MEDIUM",
|
119
|
+
"LOW"
|
120
|
+
]
|
121
|
+
},
|
122
|
+
"other_attributes": {
|
123
|
+
"description": "Any other structured data from the notice ",
|
124
|
+
"type": "object"
|
125
|
+
}
|
126
|
+
},
|
127
|
+
"additionalProperties": false,
|
128
|
+
"required": [
|
129
|
+
"issue",
|
130
|
+
"uid",
|
131
|
+
"identifier",
|
132
|
+
"source_url",
|
133
|
+
"sample_date",
|
134
|
+
"retrieved_at",
|
135
|
+
"confidence"
|
136
|
+
],
|
137
|
+
"definitions": {
|
138
|
+
"publisher": {
|
139
|
+
"title": "Publisher",
|
140
|
+
"description": "A gazette publisher",
|
141
|
+
"type": "object",
|
142
|
+
"properties": {
|
143
|
+
"name": {
|
144
|
+
"description": "The name of the publisher",
|
145
|
+
"type": "string"
|
146
|
+
},
|
147
|
+
"identifier": {
|
148
|
+
"description": "The identifier of the publisher",
|
149
|
+
"type": [
|
150
|
+
"string",
|
151
|
+
"integer"
|
152
|
+
]
|
153
|
+
},
|
154
|
+
"url": {
|
155
|
+
"description": "The URL of the publisher",
|
156
|
+
"type": "string",
|
157
|
+
"format": "uri"
|
158
|
+
}
|
159
|
+
},
|
160
|
+
"additionalProperties": false,
|
161
|
+
"required": [
|
162
|
+
"name"
|
163
|
+
]
|
164
|
+
},
|
165
|
+
"publication": {
|
166
|
+
"title": "Publication",
|
167
|
+
"description": "A gazette",
|
168
|
+
"type": "object",
|
169
|
+
"properties": {
|
170
|
+
"publisher": {
|
171
|
+
"description": "The publisher of the gazette",
|
172
|
+
"$ref": "#/definitions/publisher"
|
173
|
+
},
|
174
|
+
"jurisdiction_code": {
|
175
|
+
"description": "The jurisdiction covered by the gazette",
|
176
|
+
"type": "string"
|
177
|
+
},
|
178
|
+
"title": {
|
179
|
+
"description": "The title of the gazette",
|
180
|
+
"type": "string"
|
181
|
+
},
|
182
|
+
"url": {
|
183
|
+
"description": "The URL of the gazette",
|
184
|
+
"type": "string",
|
185
|
+
"format": "uri"
|
186
|
+
}
|
187
|
+
},
|
188
|
+
"additionalProperties": false,
|
189
|
+
"required": [
|
190
|
+
"publisher"
|
191
|
+
]
|
192
|
+
},
|
193
|
+
"issue": {
|
194
|
+
"title": "Issue",
|
195
|
+
"description": "An issue of a gazette",
|
196
|
+
"type": "object",
|
197
|
+
"properties": {
|
198
|
+
"publication": {
|
199
|
+
"description": "The gazette of which the issue is a part",
|
200
|
+
"$ref": "#/definitions/publication"
|
201
|
+
},
|
202
|
+
"identifier": {
|
203
|
+
"description": "The identifier of the issue",
|
204
|
+
"type": [
|
205
|
+
"string",
|
206
|
+
"integer"
|
207
|
+
]
|
208
|
+
},
|
209
|
+
"title": {
|
210
|
+
"description": "The title of the issue",
|
211
|
+
"type": "string"
|
212
|
+
},
|
213
|
+
"edition_id": {
|
214
|
+
"description": "The identifier of the edition of which the issue is a part",
|
215
|
+
"type": [
|
216
|
+
"string",
|
217
|
+
"integer"
|
218
|
+
]
|
219
|
+
},
|
220
|
+
"page_range": {
|
221
|
+
"description": "The page range of the issue",
|
222
|
+
"type": "string",
|
223
|
+
"pattern": "^[0-9]*(-[0-9]*)?$"
|
224
|
+
},
|
225
|
+
"url": {
|
226
|
+
"description": "The URL of the issue",
|
227
|
+
"type": "string",
|
228
|
+
"format": "uri"
|
229
|
+
},
|
230
|
+
"media_type": {
|
231
|
+
"description": "The media type of the URL resource",
|
232
|
+
"type": "string"
|
233
|
+
}
|
234
|
+
},
|
235
|
+
"additionalProperties": false,
|
236
|
+
"required": [
|
237
|
+
"publication"
|
238
|
+
]
|
239
|
+
},
|
240
|
+
"document": {
|
241
|
+
"title": "Document",
|
242
|
+
"description": "A section of an issue",
|
243
|
+
"type": "object",
|
244
|
+
"properties": {
|
245
|
+
"page_range": {
|
246
|
+
"description": "The page range of the document",
|
247
|
+
"type": "string",
|
248
|
+
"pattern": "^[0-9]*(-[0-9]*)?$"
|
249
|
+
},
|
250
|
+
"url": {
|
251
|
+
"description": "The URL of the document",
|
252
|
+
"type": "string",
|
253
|
+
"format": "uri"
|
254
|
+
},
|
255
|
+
"media_type": {
|
256
|
+
"description": "The media type of the URL resource",
|
257
|
+
"type": "string"
|
258
|
+
}
|
259
|
+
},
|
260
|
+
"additionalProperties": false,
|
261
|
+
"required": [
|
262
|
+
"url"
|
263
|
+
]
|
264
|
+
},
|
265
|
+
"classification": {
|
266
|
+
"title": "Classification",
|
267
|
+
"description": "A classification",
|
268
|
+
"type": "object",
|
269
|
+
"properties": {
|
270
|
+
"scheme": {
|
271
|
+
"description": "The scheme of the classification",
|
272
|
+
"type": "string"
|
273
|
+
},
|
274
|
+
"value": {
|
275
|
+
"description": "The classification",
|
276
|
+
"type": "string"
|
277
|
+
}
|
278
|
+
},
|
279
|
+
"additionalProperties": false,
|
280
|
+
"required": [
|
281
|
+
"value"
|
282
|
+
]
|
283
|
+
},
|
284
|
+
"body": {
|
285
|
+
"title": "Body",
|
286
|
+
"description": "Content",
|
287
|
+
"type": "object",
|
288
|
+
"properties": {
|
289
|
+
"value": {
|
290
|
+
"description": "The content",
|
291
|
+
"type": "string"
|
292
|
+
},
|
293
|
+
"media_type": {
|
294
|
+
"description": "The media type of the content",
|
295
|
+
"type": "string",
|
296
|
+
"enum": [
|
297
|
+
"text/plain",
|
298
|
+
"text/html"
|
299
|
+
]
|
300
|
+
}
|
301
|
+
},
|
302
|
+
"additionalProperties": false,
|
303
|
+
"required": [
|
304
|
+
"value"
|
305
|
+
]
|
306
|
+
},
|
307
|
+
"update_action": {
|
308
|
+
"title": "Update action",
|
309
|
+
"description": "An act of changing an object",
|
310
|
+
"type": "object",
|
311
|
+
"properties": {
|
312
|
+
"type": {
|
313
|
+
"description": "The type of change",
|
314
|
+
"type": "string",
|
315
|
+
"enum": [
|
316
|
+
"cancellation",
|
317
|
+
"correction"
|
318
|
+
]
|
319
|
+
},
|
320
|
+
"object": {
|
321
|
+
"description": "The object that was changed",
|
322
|
+
"type": "object"
|
323
|
+
}
|
324
|
+
},
|
325
|
+
"additionalProperties": false,
|
326
|
+
"required": [
|
327
|
+
"object"
|
328
|
+
]
|
329
|
+
}
|
330
|
+
}
|
331
|
+
}
|
metadata
CHANGED
@@ -1,20 +1,18 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: turbot-runner
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.2.
|
5
|
-
prerelease:
|
4
|
+
version: 0.2.8
|
6
5
|
platform: ruby
|
7
6
|
authors:
|
8
7
|
- OpenCorporates
|
9
8
|
autorequire:
|
10
9
|
bindir: bin
|
11
10
|
cert_chain: []
|
12
|
-
date: 2016-01-
|
11
|
+
date: 2016-01-15 00:00:00.000000000 Z
|
13
12
|
dependencies:
|
14
13
|
- !ruby/object:Gem::Dependency
|
15
14
|
name: activesupport
|
16
15
|
requirement: !ruby/object:Gem::Requirement
|
17
|
-
none: false
|
18
16
|
requirements:
|
19
17
|
- - '='
|
20
18
|
- !ruby/object:Gem::Version
|
@@ -22,7 +20,6 @@ dependencies:
|
|
22
20
|
type: :runtime
|
23
21
|
prerelease: false
|
24
22
|
version_requirements: !ruby/object:Gem::Requirement
|
25
|
-
none: false
|
26
23
|
requirements:
|
27
24
|
- - '='
|
28
25
|
- !ruby/object:Gem::Version
|
@@ -30,7 +27,6 @@ dependencies:
|
|
30
27
|
- !ruby/object:Gem::Dependency
|
31
28
|
name: openc-json_schema
|
32
29
|
requirement: !ruby/object:Gem::Requirement
|
33
|
-
none: false
|
34
30
|
requirements:
|
35
31
|
- - '='
|
36
32
|
- !ruby/object:Gem::Version
|
@@ -38,7 +34,6 @@ dependencies:
|
|
38
34
|
type: :runtime
|
39
35
|
prerelease: false
|
40
36
|
version_requirements: !ruby/object:Gem::Requirement
|
41
|
-
none: false
|
42
37
|
requirements:
|
43
38
|
- - '='
|
44
39
|
- !ruby/object:Gem::Version
|
@@ -49,6 +44,7 @@ executables: []
|
|
49
44
|
extensions: []
|
50
45
|
extra_rdoc_files: []
|
51
46
|
files:
|
47
|
+
- README.md
|
52
48
|
- bin/rspec
|
53
49
|
- lib/turbot_runner.rb
|
54
50
|
- lib/turbot_runner/base_handler.rb
|
@@ -60,6 +56,41 @@ files:
|
|
60
56
|
- lib/turbot_runner/utils.rb
|
61
57
|
- lib/turbot_runner/validator.rb
|
62
58
|
- lib/turbot_runner/version.rb
|
59
|
+
- schema/schemas/accounts-statement-schema.json
|
60
|
+
- schema/schemas/company-schema.json
|
61
|
+
- schema/schemas/filing-schema.json
|
62
|
+
- schema/schemas/financial-payment-schema.json
|
63
|
+
- schema/schemas/gazette-notice-schema.json
|
64
|
+
- schema/schemas/includes/accounts_element.json
|
65
|
+
- schema/schemas/includes/address.json
|
66
|
+
- schema/schemas/includes/alternative_name.json
|
67
|
+
- schema/schemas/includes/company-for-nesting.json
|
68
|
+
- schema/schemas/includes/company.json
|
69
|
+
- schema/schemas/includes/entity.json
|
70
|
+
- schema/schemas/includes/filing.json
|
71
|
+
- schema/schemas/includes/filing_document.json
|
72
|
+
- schema/schemas/includes/financial-payment-data-object.json
|
73
|
+
- schema/schemas/includes/identifier.json
|
74
|
+
- schema/schemas/includes/industry_code.json
|
75
|
+
- schema/schemas/includes/licence-data-object.json
|
76
|
+
- schema/schemas/includes/officer.json
|
77
|
+
- schema/schemas/includes/organisation.json
|
78
|
+
- schema/schemas/includes/permission.json
|
79
|
+
- schema/schemas/includes/person.json
|
80
|
+
- schema/schemas/includes/person_name.json
|
81
|
+
- schema/schemas/includes/previous_name.json
|
82
|
+
- schema/schemas/includes/share-parcel-data.json
|
83
|
+
- schema/schemas/includes/share-parcel.json
|
84
|
+
- schema/schemas/includes/subsidiary-relationship-data.json
|
85
|
+
- schema/schemas/includes/total-shares.json
|
86
|
+
- schema/schemas/includes/unknown_entity_type.json
|
87
|
+
- schema/schemas/licence-schema.json
|
88
|
+
- schema/schemas/primary-data-schema.json
|
89
|
+
- schema/schemas/share-parcel-schema.json
|
90
|
+
- schema/schemas/simple-financial-payment-schema.json
|
91
|
+
- schema/schemas/simple-licence-schema.json
|
92
|
+
- schema/schemas/simple-subsidiary-schema.json
|
93
|
+
- schema/schemas/subsidiary-relationship-schema.json
|
63
94
|
- spec/bots/bot-that-crashes-immediately/manifest.json
|
64
95
|
- spec/bots/bot-that-crashes-immediately/scraper.rb
|
65
96
|
- spec/bots/bot-that-crashes-immediately/transformer1.rb
|
@@ -115,63 +146,28 @@ files:
|
|
115
146
|
- spec/outputs/full-transformer.out
|
116
147
|
- spec/outputs/truncated-scraper.out
|
117
148
|
- spec/spec_helper.rb
|
118
|
-
- schema/schemas/accounts-statement-schema.json
|
119
|
-
- schema/schemas/company-schema.json
|
120
|
-
- schema/schemas/filing-schema.json
|
121
|
-
- schema/schemas/financial-payment-schema.json
|
122
|
-
- schema/schemas/includes/accounts_element.json
|
123
|
-
- schema/schemas/includes/address.json
|
124
|
-
- schema/schemas/includes/alternative_name.json
|
125
|
-
- schema/schemas/includes/company-for-nesting.json
|
126
|
-
- schema/schemas/includes/company.json
|
127
|
-
- schema/schemas/includes/entity.json
|
128
|
-
- schema/schemas/includes/filing.json
|
129
|
-
- schema/schemas/includes/filing_document.json
|
130
|
-
- schema/schemas/includes/financial-payment-data-object.json
|
131
|
-
- schema/schemas/includes/identifier.json
|
132
|
-
- schema/schemas/includes/industry_code.json
|
133
|
-
- schema/schemas/includes/licence-data-object.json
|
134
|
-
- schema/schemas/includes/officer.json
|
135
|
-
- schema/schemas/includes/organisation.json
|
136
|
-
- schema/schemas/includes/permission.json
|
137
|
-
- schema/schemas/includes/person.json
|
138
|
-
- schema/schemas/includes/person_name.json
|
139
|
-
- schema/schemas/includes/previous_name.json
|
140
|
-
- schema/schemas/includes/share-parcel-data.json
|
141
|
-
- schema/schemas/includes/share-parcel.json
|
142
|
-
- schema/schemas/includes/subsidiary-relationship-data.json
|
143
|
-
- schema/schemas/includes/total-shares.json
|
144
|
-
- schema/schemas/includes/unknown_entity_type.json
|
145
|
-
- schema/schemas/licence-schema.json
|
146
|
-
- schema/schemas/primary-data-schema.json
|
147
|
-
- schema/schemas/share-parcel-schema.json
|
148
|
-
- schema/schemas/simple-financial-payment-schema.json
|
149
|
-
- schema/schemas/simple-licence-schema.json
|
150
|
-
- schema/schemas/simple-subsidiary-schema.json
|
151
|
-
- schema/schemas/subsidiary-relationship-schema.json
|
152
149
|
homepage: http://turbot.opencorporates.com/
|
153
150
|
licenses:
|
154
151
|
- MIT
|
152
|
+
metadata: {}
|
155
153
|
post_install_message:
|
156
154
|
rdoc_options: []
|
157
155
|
require_paths:
|
158
156
|
- lib
|
159
157
|
required_ruby_version: !ruby/object:Gem::Requirement
|
160
|
-
none: false
|
161
158
|
requirements:
|
162
|
-
- -
|
159
|
+
- - ">="
|
163
160
|
- !ruby/object:Gem::Version
|
164
161
|
version: 1.9.2
|
165
162
|
required_rubygems_version: !ruby/object:Gem::Requirement
|
166
|
-
none: false
|
167
163
|
requirements:
|
168
|
-
- -
|
164
|
+
- - ">="
|
169
165
|
- !ruby/object:Gem::Version
|
170
166
|
version: '0'
|
171
167
|
requirements: []
|
172
168
|
rubyforge_project:
|
173
|
-
rubygems_version:
|
169
|
+
rubygems_version: 2.4.5
|
174
170
|
signing_key:
|
175
|
-
specification_version:
|
171
|
+
specification_version: 4
|
176
172
|
summary: Utilities for running bots with Turbot
|
177
173
|
test_files: []
|