mirador 0.0.4 → 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +8 -8
- data/README.md +231 -16
- data/lib/mirador.rb +277 -71
- data/lib/mirador/version.rb +1 -1
- data/test/test_mirador.rb +111 -56
- metadata +2 -2
checksums.yaml
CHANGED
@@ -1,15 +1,15 @@
|
|
1
1
|
---
|
2
2
|
!binary "U0hBMQ==":
|
3
3
|
metadata.gz: !binary |-
|
4
|
-
|
4
|
+
MGZmMTU1ODJkZDdiZjM3MmQ1NzA1MjJkNjkxMjUxMzM1MmMyOTQ1Yg==
|
5
5
|
data.tar.gz: !binary |-
|
6
|
-
|
6
|
+
OWYyNjc4N2EzMWU1Mjc4YWUxMWJmOGU4NjNhMmFkZjRiZWUzYWYwMA==
|
7
7
|
SHA512:
|
8
8
|
metadata.gz: !binary |-
|
9
|
-
|
10
|
-
|
11
|
-
|
9
|
+
Mjk2MWU1MDM2ZmNiZDUyYTMwMTRlNzIyMTMyZWY5ZDg5ZDdmNGZhM2FhNDlk
|
10
|
+
MDk5MmQ4Nzg3MTI1MjE2ODM1MDY4MjU5YzEzNjAyZmRhYTEwYThmMjI2MTVj
|
11
|
+
Mjk1N2IxMGIyNzdhMmQxY2IyN2U3ZGQxZGQxOGRlZmI2ZDAyNDU=
|
12
12
|
data.tar.gz: !binary |-
|
13
|
-
|
14
|
-
|
15
|
-
|
13
|
+
Yzg0NDBiNTgwODg2MjRlMGVmZTUzNzQ1ZDQ1NTE0ZDhlZTE5MzYxMjhjNTQ0
|
14
|
+
MDVkNzFhZGIwNjU5NGQyMmY1MTBhN2RiZmM1MjU0MTJkYTYwNTNmY2EyZjE0
|
15
|
+
MmZkNGJmMmY0YzhkZTkwZjgzMjY4ZmFmNjdlNzYzMjdmNDBiNmU=
|
data/README.md
CHANGED
@@ -18,42 +18,257 @@ Or install it yourself as:
|
|
18
18
|
|
19
19
|
## Usage
|
20
20
|
|
21
|
-
|
21
|
+
To get started, you need an API key, available from [mirador.im/join](http://mirador.im/). If you have problems with the API or this client, please contact support@mirador.im.
|
22
22
|
|
23
|
-
|
23
|
+
## Mirador::Result and Mirador::ResultList
|
24
24
|
|
25
|
-
|
25
|
+
All multiple-request methods (e.g., classify_files), return a [Mirador::ResultList](#resultlist), which is effectively a list of Mirador::Result objects; single-request methods (e.g., classify_url) return a [Mirador::Result](#result) object.
|
26
|
+
|
27
|
+
## Classifying Files & Buffers
|
28
|
+
|
29
|
+
You can classify 4 types of files/file-objects:
|
30
|
+
|
31
|
+
* file objects (e.g., `x` where `x = File.open('myfile.jpg')`); [classify_files](#classify_files)
|
32
|
+
* filenames `myfile.jpg` [classify_files](#classify_files)
|
33
|
+
* buffers `buffer = File.read('myfile.jpg')` [classify_buffers](#classify_buffers)
|
34
|
+
* base64-encoded buffers (e.g., from a data URI) [classify_encoded_strings](#classify_encoded_strings)
|
35
|
+
|
36
|
+
The methods for file-based classification are as follows:
|
37
|
+
|
38
|
+
|
39
|
+
### <a name="classify_files"></a> Mirador::Client#classify_files
|
26
40
|
|
27
41
|
```ruby
|
28
42
|
require 'mirador'
|
43
|
+
mc = Mirador::Client.new('your_api_key')
|
44
|
+
|
45
|
+
results = mc.classify_files('test.jpg', 'picture.jpg')
|
29
46
|
|
30
|
-
|
31
|
-
|
32
|
-
|
47
|
+
assert results['test.jpg']
|
48
|
+
assert_equal results.length, 2
|
49
|
+
|
50
|
+
results.each do |res|
|
51
|
+
puts "#{ res.id }, #{ res.value }"
|
52
|
+
end
|
53
|
+
|
54
|
+
results.each do |id, res|
|
55
|
+
puts "#{ id }, #{ res.value }"
|
33
56
|
end
|
34
57
|
|
35
58
|
```
|
36
59
|
|
37
|
-
|
60
|
+
You can also specify an id to be used:
|
61
|
+
|
62
|
+
```ruby
|
63
|
+
require 'mirador'
|
64
|
+
mc = Mirador::Client.new 'your_api_key'
|
65
|
+
|
66
|
+
# first method: use ids as keys
|
67
|
+
results = mc.classify_files(nsfw: 'nsfw.jpg', sfw: 'sfw.jpg')
|
38
68
|
|
39
|
-
|
69
|
+
assert results[:nsfw]
|
70
|
+
assert results[:sfw]
|
71
|
+
|
72
|
+
# second method: pass an array of { id:, data: } hashes
|
73
|
+
results = mc.classify_files([{ id: :nsfw, data: 'nsfw.jpg'}, { id: :sfw, data: 'sfw.jpg' }])
|
74
|
+
|
75
|
+
assert results[:nsfw]
|
76
|
+
assert results[:sfw]
|
77
|
+
```
|
78
|
+
|
79
|
+
File can be either a filename or a file object; e.g., the following is also valid:
|
80
|
+
|
81
|
+
```ruby
|
82
|
+
results = mc.classify_files(nsfw: File.open('nsfw.jpg'))
|
83
|
+
```
|
84
|
+
|
85
|
+
### <a name='classify_file'></a> Mirador::Client#classify_file
|
86
|
+
|
87
|
+
A shortcut for classifying a single file; this will return a `Mirador::Result` instead of a `Mirador::ResultList`:
|
40
88
|
|
41
89
|
```ruby
|
42
90
|
require 'mirador'
|
91
|
+
mc = Mirador::Client.new 'your_api_key'
|
43
92
|
|
44
|
-
|
45
|
-
|
46
|
-
end
|
93
|
+
# first method: use ids as keys
|
94
|
+
nsfw = mc.classify_file(nsfw: 'nsfw.jpg')
|
47
95
|
|
96
|
+
puts nsfw.value
|
48
97
|
```
|
49
98
|
|
50
|
-
###
|
99
|
+
### <a name='classify_buffers'></a> Mirador:Client#classify_buffers
|
100
|
+
|
101
|
+
Classify a buffer, e.g., an already-read file. This simplifies the classification of file uploads, e.g. POST data. The interface is identical to [classify_files](#classify_files), only differing in the actual data passed in:
|
102
|
+
|
103
|
+
```ruby
|
104
|
+
require 'mirador'
|
105
|
+
mc = Mirador::Client.new 'your_api_key'
|
106
|
+
|
107
|
+
nsfw_buf = File.read('nsfw.jpg')
|
108
|
+
sfw_buf = File.read('sfw.jpg')
|
109
|
+
|
110
|
+
# these are equivalent
|
111
|
+
results = mc.classify_buffers(nsfw: nsfw_buf, sfw: sfw_buf)
|
112
|
+
results = mc.classify_buffers([{id: :nsfw, data: nsfw_buf}, {id: :sfw, data: sfw_buf}])
|
113
|
+
|
114
|
+
# since buffers dont have a name, you just get an index as id
|
115
|
+
results = mc.classify_buffers(nsfw_buf, sfw_buf)
|
116
|
+
```
|
117
|
+
|
118
|
+
#### <a name='classify_buffer'></a> Mirador::Client#classify_buffer
|
119
|
+
|
120
|
+
As with classify_file, there is a shortcut for classifying only one buffer; see [classify_file](#classify_file) for clarifications on usage (it's identical).
|
121
|
+
|
122
|
+
### <a name='classify_encoded_strings'></a> Mirador::Client#classify_encoded_strings
|
123
|
+
|
124
|
+
The Mirador API internally represents images as base64-encoded strings (agnostic of image encoding); this method lets you pass in an alread-encoded string in the event that you're also using base64 encoding elsewhere in your system. Usage is the same as [classify_buffers](#classify_buffers):
|
125
|
+
|
126
|
+
```ruby
|
127
|
+
require 'mirador'
|
128
|
+
require 'base64'
|
129
|
+
|
130
|
+
mc = Mirador::Client.new 'your_api_key'
|
131
|
+
|
132
|
+
nsfw_buf = Base64.encode64(File.read('nsfw.jpg'))
|
133
|
+
sfw_buf = Base64.encode64(File.read('sfw.jpg'))
|
134
|
+
|
135
|
+
# these are equivalent
|
136
|
+
results = mc.classify_encoded_strings(nsfw: nsfw_buf, sfw: sfw_buf)
|
137
|
+
results = mc.classify_encoded_strings([{id: :nsfw, data: nsfw_buf}, {id: :sfw, data: sfw_buf}])
|
138
|
+
|
139
|
+
# since strings dont have a name, you just get an index as id
|
140
|
+
results = mc.classify_encoded_strings(nsfw_buf, sfw_buf)
|
141
|
+
```
|
142
|
+
|
143
|
+
#### <a name='classify_encoded_string'></a> Mirador::Client#classify_encoded_string
|
144
|
+
|
145
|
+
Another helper for only working with 1 request/result at a time. See [classify_file](#classify_file) for more info.
|
146
|
+
|
147
|
+
|
148
|
+
### <a name='classify_data_uris'></a> Mirador::Client#classify_data_uris
|
149
|
+
|
150
|
+
This simplifies data transfer between client applications and the mirador API. For example, given the following javascript:
|
151
|
+
|
152
|
+
```javascript
|
153
|
+
document.getElementById('form-field').addEventListener('change', function (e) {
|
154
|
+
|
155
|
+
var file = this.files[0];
|
156
|
+
|
157
|
+
var reader = new FileReader();
|
158
|
+
reader.onload = function (e) {
|
159
|
+
$.post('/proxy/mirador', { id: file.name, data: e.target.result });
|
160
|
+
}
|
161
|
+
|
162
|
+
reader.readAsDataURL(file);
|
163
|
+
});
|
164
|
+
```
|
165
|
+
|
166
|
+
Your could classify that data url with the following code:
|
167
|
+
|
168
|
+
```ruby
|
169
|
+
|
170
|
+
res = mc.classify_data_uris(request['id'] => request['data'])
|
171
|
+
|
172
|
+
# send the result
|
173
|
+
res[request['id']].to_json
|
174
|
+
|
175
|
+
# or, even easier
|
176
|
+
mc.classify_data_uri(request['id'] => request['data']).to_json
|
177
|
+
|
178
|
+
```
|
179
|
+
|
180
|
+
Otherwise, classify_data_uris and classify_data_uri have identical interfaces to the other methods covered so far.
|
181
|
+
|
182
|
+
|
183
|
+
## Classify URLs
|
184
|
+
|
185
|
+
You can easily classify a publically-available URL (e.g., a public s3 bucket), with [classify_urls](#classify_urls) and [classify_url](#classify_url). The interfaces for these methods are identical to the file-handling methods covered above.
|
186
|
+
|
187
|
+
|
188
|
+
### <a name='classify_urls'></a> Mirador::Client#classify_urls
|
189
|
+
|
190
|
+
The only things to keep in mind with URLs:
|
191
|
+
|
192
|
+
* must be publically-accessibly
|
193
|
+
* must be < Mirador::Client::MAX_ID_LEN if you are using the url as the item's id (see below)
|
194
|
+
* download/response time on url will affect response time of result, must be less than 60 seconds.
|
195
|
+
|
196
|
+
|
197
|
+
#### Examples:
|
198
|
+
|
199
|
+
|
200
|
+
Assigning specific ids to urls:
|
201
|
+
|
202
|
+
```ruby
|
203
|
+
require 'mirador'
|
204
|
+
|
205
|
+
mc = Mirador::Client.new 'your_api_key'
|
206
|
+
|
207
|
+
res = mc.classify_urls(nsfw: 'http://static.mirador.im/test/nsfw.jpg', sfw: 'http://static.mirador.im/test/sfw.jpg')
|
208
|
+
|
209
|
+
assert res[:nsfw]
|
210
|
+
assert res[:sfw].safe
|
211
|
+
|
212
|
+
```
|
213
|
+
|
214
|
+
Implicitly using url as its own id:
|
215
|
+
|
216
|
+
```ruby
|
217
|
+
require 'mirador'
|
218
|
+
|
219
|
+
nsfw_url = 'http://static.mirador.im/test/nsfw.jpg'
|
220
|
+
sfw_url = 'http://static.mirador.im/test/sfw.jpg'
|
221
|
+
|
222
|
+
mc = Mirador::Client.new 'your_api_key'
|
223
|
+
res = mc.classify_urls(nsfw_url, sfw_url)
|
224
|
+
|
225
|
+
puts res[nsfw_url].value
|
226
|
+
puts res[sfw_url].value
|
227
|
+
```
|
228
|
+
|
229
|
+
Classify a single URL using Mirador::Client#classify_url
|
230
|
+
|
231
|
+
```ruby
|
232
|
+
require 'mirador'
|
233
|
+
|
234
|
+
mc = Mirador::Client.new 'your_api_key'
|
235
|
+
nsfw = mc.classify_url(nsfw_url)
|
236
|
+
|
237
|
+
assert (not nsfw.safe)
|
238
|
+
puts nsfw.value
|
239
|
+
```
|
240
|
+
|
241
|
+
## <a name='result'></a> Mirador::Result
|
242
|
+
|
243
|
+
The `Mirador::Result` class wraps the output of the API for a specific image/url. It has the following attributes:
|
244
|
+
|
245
|
+
* `@id` [Mixed]: the id, as specified in the request, or implied (see above)
|
246
|
+
* `@safe` [Boolean]: whether the image should be considered flagged/containing adult content
|
247
|
+
* `@value` [Float 0.0-1.0]: A float indicating the likelyhood of the image containing adult content (useful for creating custom thresholds)
|
248
|
+
* `@error` [String]: will only be non-nil if this is an error
|
249
|
+
|
250
|
+
The `Mirador::Result` object also has a couple of convenience methods:
|
251
|
+
|
252
|
+
* `#to_h` - convert to a hash
|
253
|
+
* `#to_json` - if json is require'd, serialize to json
|
254
|
+
* `#failed?` - returns a boolean indicating whether image is a failure/error
|
255
|
+
* `#to_s` - returns a string representation of the result`
|
256
|
+
* `#name` **(deprecated)** - this simply maps to `@id`
|
257
|
+
|
258
|
+
## <a name='resultlist'></a> Mirador::ResultList [Enumerable]
|
259
|
+
|
260
|
+
Methods that return multiple results do so by returning a single `Mirador::ResultList`. This object is used in lieu of a Hash or Array as to provide mixed-access. You can treat it as an array, iterating via `each do |x|`, indexing with integers, or by simply calling `#to_a`, or as a hash, indexing with `@id`'s from image-requests.
|
261
|
+
|
262
|
+
The ResultList has the following methods:
|
51
263
|
|
52
|
-
|
264
|
+
* `#[](key)` operator override to index the ResultList. You can index by integers in range of 0 - ResultList#length, or by an `@id` for one of the Result objects within.
|
265
|
+
* `#to_a` convert to an array of `Mirador::Result` objects
|
266
|
+
* `#length` the number of items in the `ResultList`
|
267
|
+
* `#update` equivalent to Hash#update
|
268
|
+
* `#to_h` conver to a hash
|
269
|
+
* `#to_json` serialize the resultlist as json
|
270
|
+
* `#each` `ResultList` includes `Enumerable`, and this implementation of `#each` checks the arity of blocks passed in to allow iteration either as an array or as a Hash.
|
53
271
|
|
54
|
-
* `Result.name` - `string`, the filename or url for this request
|
55
|
-
* `Result.safe` - `bool`, a boolean indicating whether image contains adult content.
|
56
|
-
* `Result.value` - `float`, a number 0.0 - 1.0 indicating confidence of judgement
|
57
272
|
|
58
273
|
## Contributing
|
59
274
|
|
data/lib/mirador.rb
CHANGED
@@ -3,137 +3,343 @@ require 'base64'
|
|
3
3
|
|
4
4
|
module Mirador
|
5
5
|
|
6
|
-
|
6
|
+
class ApiError < StandardError
|
7
|
+
end
|
7
8
|
|
8
|
-
class
|
9
|
-
|
9
|
+
class ResultList
|
10
|
+
include Enumerable
|
11
|
+
|
12
|
+
def initialize(items=[])
|
13
|
+
@items = {}
|
10
14
|
|
11
|
-
|
12
|
-
|
13
|
-
|
14
|
-
@value = data['value']
|
15
|
+
items.each do |x|
|
16
|
+
@items[x.id] = x
|
17
|
+
end
|
15
18
|
end
|
16
19
|
|
17
|
-
def
|
18
|
-
|
20
|
+
def <<(item)
|
21
|
+
@items[item.id] = item
|
22
|
+
end
|
23
|
+
|
24
|
+
def [](key)
|
25
|
+
if key.is_a? Integer and not @items.has_key? key
|
26
|
+
@items.values[key]
|
27
|
+
else
|
28
|
+
@items[key.to_s]
|
29
|
+
end
|
30
|
+
end
|
31
|
+
|
32
|
+
def to_a
|
33
|
+
@items.values
|
34
|
+
end
|
35
|
+
|
36
|
+
def length
|
37
|
+
@items.values.length
|
19
38
|
end
|
20
39
|
|
21
|
-
def
|
40
|
+
def update other
|
41
|
+
@items.update(other)
|
42
|
+
end
|
43
|
+
|
44
|
+
def to_h
|
45
|
+
@items
|
46
|
+
end
|
22
47
|
|
23
|
-
|
24
|
-
|
48
|
+
def to_json
|
49
|
+
@items.to_json
|
50
|
+
end
|
51
|
+
|
52
|
+
def each &block
|
53
|
+
if block.arity == 1
|
54
|
+
@items.values.each do |x|
|
55
|
+
block.call(x)
|
56
|
+
end
|
57
|
+
else
|
58
|
+
@items.each do |k, v|
|
59
|
+
block.call(k, v)
|
60
|
+
end
|
25
61
|
end
|
62
|
+
end
|
63
|
+
|
64
|
+
def self.parse_results res
|
26
65
|
|
27
|
-
|
28
|
-
|
66
|
+
output = {}
|
67
|
+
res.each do |x|
|
68
|
+
r = Result.new(x)
|
69
|
+
output[r.id] = r
|
29
70
|
end
|
71
|
+
|
72
|
+
output
|
30
73
|
end
|
31
74
|
|
32
75
|
end
|
33
76
|
|
34
|
-
class
|
77
|
+
class Result
|
78
|
+
attr_accessor :id, :safe, :value, :error
|
79
|
+
|
80
|
+
def initialize data
|
81
|
+
|
82
|
+
if data.has_key? 'errors'
|
83
|
+
@error = data['errors']
|
84
|
+
return
|
85
|
+
end
|
86
|
+
|
87
|
+
@id = data['id']
|
88
|
+
@safe = data['result']['safe']
|
89
|
+
@value = data['result']['value']
|
90
|
+
|
91
|
+
end
|
92
|
+
|
93
|
+
def to_h
|
94
|
+
{
|
95
|
+
id: @id,
|
96
|
+
safe: @safe,
|
97
|
+
value: @value,
|
98
|
+
}
|
99
|
+
end
|
100
|
+
|
101
|
+
def to_json
|
102
|
+
as_h = self.to_h
|
103
|
+
|
104
|
+
if as_h.respond_to? :to_json
|
105
|
+
as_h.to_json
|
106
|
+
else
|
107
|
+
nil
|
108
|
+
end
|
109
|
+
end
|
110
|
+
|
111
|
+
def failed?
|
112
|
+
@error != nil
|
113
|
+
end
|
114
|
+
|
115
|
+
def to_s
|
116
|
+
"<Mirador::Result; id: #{ @id }; safe: #{ @safe }; value: #{ @value }/>"
|
117
|
+
end
|
118
|
+
|
119
|
+
def name
|
120
|
+
@id
|
121
|
+
end
|
122
|
+
|
35
123
|
end
|
36
124
|
|
37
125
|
class Client
|
38
126
|
include HTTParty
|
39
127
|
base_uri 'api.mirador.im'
|
128
|
+
|
40
129
|
default_timeout 10
|
41
130
|
|
42
|
-
MAX_LEN =
|
131
|
+
MAX_LEN = 5
|
132
|
+
MAX_ID_LEN = 256
|
133
|
+
DATA_URI_PRE = ';base64,'
|
134
|
+
DATA_URI_PRELEN = 8
|
43
135
|
|
44
136
|
def initialize(api_key)
|
45
137
|
@options = { api_key: api_key }
|
46
138
|
end
|
47
139
|
|
48
|
-
|
49
|
-
|
50
|
-
|
51
|
-
|
52
|
-
|
53
|
-
out << self.classify_urls(s)
|
140
|
+
# metaprogramming extreme
|
141
|
+
[:url, :file, :buffer, :encoded_string, :data_uri].each do |datatype|
|
142
|
+
define_method("classify_#{datatype.to_s}s") do |args, params={}|
|
143
|
+
flexible_request args, params do |item|
|
144
|
+
fmt_items(datatype, item)
|
54
145
|
end
|
146
|
+
end
|
55
147
|
|
56
|
-
|
148
|
+
define_method("classify_#{datatype.to_s}") do |args, params={}|
|
149
|
+
res = self.send("classify_#{datatype.to_s}s", args, params)
|
150
|
+
res[0]
|
57
151
|
end
|
58
152
|
|
59
|
-
|
60
|
-
"/v1/classify",
|
61
|
-
{
|
62
|
-
body: @options.merge({url: urls}),
|
63
|
-
headers: {"User-Agent" => "Mirador Client v1.0/Ruby"}
|
64
|
-
}
|
65
|
-
)
|
153
|
+
end
|
66
154
|
|
67
|
-
|
68
|
-
|
69
|
-
|
70
|
-
|
71
|
-
|
155
|
+
protected
|
156
|
+
|
157
|
+
def flexible_request(args, params={}, &cb)
|
158
|
+
req = {}
|
159
|
+
|
160
|
+
req = (if args.is_a? Hash
|
161
|
+
|
162
|
+
Hash[args.map do |k, v|
|
163
|
+
process_param(k, v)
|
164
|
+
end]
|
165
|
+
|
166
|
+
elsif args.is_a? String
|
167
|
+
Hash[[process_argument(args)]]
|
168
|
+
|
169
|
+
elsif args and args.length
|
170
|
+
Hash[args.each_with_index.map do |a, idx|
|
171
|
+
process_argument(a, idx)
|
172
|
+
end]
|
72
173
|
|
73
|
-
|
174
|
+
elsif params
|
175
|
+
Hash[params.map do |k, v|
|
176
|
+
process_param(k, v)
|
177
|
+
end]
|
178
|
+
end)
|
179
|
+
|
180
|
+
chunked_request(req) do |item|
|
181
|
+
formatted = cb.call(item)
|
182
|
+
make_request(formatted)
|
183
|
+
end
|
74
184
|
end
|
75
185
|
|
76
|
-
def
|
77
|
-
|
78
|
-
|
79
|
-
|
80
|
-
|
186
|
+
def process_argument arg, idx=0
|
187
|
+
|
188
|
+
if arg.is_a?(String)
|
189
|
+
if arg.length < MAX_ID_LEN
|
190
|
+
[arg, arg]
|
191
|
+
else
|
192
|
+
[idx, arg]
|
81
193
|
end
|
82
194
|
|
83
|
-
|
84
|
-
end
|
195
|
+
elsif arg.respond_to?(:name) and arg.respond_to?(:read)
|
85
196
|
|
86
|
-
|
87
|
-
|
88
|
-
|
197
|
+
[arg.name, arg]
|
198
|
+
|
199
|
+
elsif arg.respond_to?(:id) and arg.respond_to?(:data)
|
89
200
|
|
201
|
+
[arg.id, arg.data]
|
90
202
|
|
91
|
-
|
203
|
+
elsif arg.is_a?(Hash)
|
92
204
|
|
93
|
-
|
94
|
-
|
95
|
-
|
96
|
-
|
205
|
+
if arg.has_key? :id and arg.has_key? :data
|
206
|
+
[arg[:id], arg[:data]]
|
207
|
+
elsif arg.has_key? 'id' and arg.has_key? 'data'
|
208
|
+
[arg['id'], arg['data']]
|
97
209
|
end
|
98
210
|
|
99
|
-
|
211
|
+
else
|
212
|
+
raise ApiError, "Invalid argument: #{ arg }"
|
100
213
|
end
|
101
214
|
|
102
|
-
|
103
|
-
|
104
|
-
|
105
|
-
|
106
|
-
|
107
|
-
|
215
|
+
end
|
216
|
+
|
217
|
+
# given a parameter passed in,
|
218
|
+
# assuming that its a id => data mapping, return
|
219
|
+
# the correct formatting/check for any fuck ups
|
220
|
+
# @arguments:
|
221
|
+
# k - key
|
222
|
+
# v - value
|
223
|
+
# @returns:
|
224
|
+
# { k => v } pair
|
225
|
+
def process_param k, v
|
226
|
+
|
227
|
+
if v.is_a?(File)
|
228
|
+
[ k, v.read ]
|
229
|
+
elsif k.respond_to?(:to_s) and v.is_a?(String)
|
230
|
+
[ k.to_s, v ]
|
231
|
+
else
|
232
|
+
raise ApiError, "Invalid Argument: #{ k } => #{ v }"
|
233
|
+
end
|
234
|
+
|
235
|
+
end
|
236
|
+
|
237
|
+
# given a request and a block,
|
238
|
+
# call the block X number of times
|
239
|
+
# where X is request.length / MAX_LEN
|
240
|
+
def chunked_request req, &mthd
|
241
|
+
output = ResultList.new
|
242
|
+
req.each_slice(MAX_LEN).each do |slice|
|
243
|
+
output.update(mthd.call(slice))
|
108
244
|
end
|
109
245
|
|
110
|
-
|
111
|
-
return self.classify_encoded names, processed
|
246
|
+
return output
|
112
247
|
end
|
113
248
|
|
114
|
-
|
115
|
-
|
116
|
-
|
249
|
+
# basically, transform hash h into a hash
|
250
|
+
# where the key-value pairs are all formatted
|
251
|
+
# by 'fmt-item' (should double the number of key-value
|
252
|
+
# pairs in the hash)
|
253
|
+
def fmt_items name, h
|
254
|
+
out = {}
|
255
|
+
h.each_with_index do |kv, idx|
|
256
|
+
out.update fmt_item(name, idx, kv[0], kv[1])
|
257
|
+
end
|
258
|
+
return out
|
117
259
|
end
|
118
260
|
|
119
|
-
|
261
|
+
@@name_map = {
|
262
|
+
file: 'image',
|
263
|
+
buffer: 'image',
|
264
|
+
raw: 'image',
|
265
|
+
url: 'url',
|
266
|
+
encoded_string: 'image',
|
267
|
+
data_uri: 'image',
|
268
|
+
}
|
269
|
+
|
270
|
+
@@formatters = {
|
271
|
+
url: Proc.new { |url| url },
|
272
|
+
|
273
|
+
file: Proc.new { |file|
|
274
|
+
|
275
|
+
Base64.encode64(if file.respond_to? :read
|
276
|
+
file.read
|
277
|
+
else
|
278
|
+
File.read(file)
|
279
|
+
end).gsub(/\n/, '')
|
280
|
+
|
281
|
+
},
|
282
|
+
|
283
|
+
buffer: Proc.new { |file|
|
284
|
+
|
285
|
+
Base64.encode64(file).gsub(/\n/, '')
|
286
|
+
|
287
|
+
},
|
288
|
+
|
289
|
+
raw: Proc.new { |file|
|
290
|
+
|
291
|
+
Base64.encode64(file).gsub(/\n/, '')
|
292
|
+
|
293
|
+
},
|
294
|
+
|
295
|
+
encoded_string: Proc.new { |b64str|
|
296
|
+
b64str.gsub(/\n/, '')
|
297
|
+
},
|
298
|
+
|
299
|
+
data_uri: Proc.new { |datauri|
|
300
|
+
datauri.sub(/^.+;base64,/, '').gsub(/\n/,'')
|
301
|
+
},
|
302
|
+
|
303
|
+
}
|
304
|
+
|
305
|
+
# produce a k-v mapping internal to the API,
|
306
|
+
# so that 'name' is the datatype:
|
307
|
+
# e.g., name[idx][id], name[idx][data]
|
308
|
+
def fmt_item name, idx, id, data
|
309
|
+
formatted = @@formatters[name].call(data)
|
310
|
+
datatype = @@name_map[name]
|
311
|
+
{
|
312
|
+
"#{datatype}[#{idx}][id]" => id,
|
313
|
+
"#{datatype}[#{idx}][data]" => formatted,
|
314
|
+
}
|
315
|
+
end
|
316
|
+
|
317
|
+
# base method to actually make the request
|
318
|
+
def make_request params
|
319
|
+
|
120
320
|
res = self.class.post(
|
121
321
|
"/v1/classify",
|
122
322
|
{
|
123
|
-
body: @options.merge(
|
124
|
-
headers: {
|
323
|
+
body: @options.merge(params),
|
324
|
+
headers: {"User-Agent" => "Mirador Client v1.0/Ruby"}
|
125
325
|
}
|
126
326
|
)
|
127
327
|
|
328
|
+
k = 'results'
|
329
|
+
|
128
330
|
if res['errors']
|
129
|
-
raise ApiError, res['errors']
|
130
|
-
end
|
131
331
|
|
132
|
-
|
133
|
-
|
332
|
+
if not res['result']
|
333
|
+
raise ApiError, res
|
334
|
+
else
|
335
|
+
k = 'result'
|
336
|
+
end
|
337
|
+
|
338
|
+
elsif not res
|
339
|
+
raise ApiError, "no response: #{ res.code }"
|
134
340
|
end
|
135
341
|
|
136
|
-
return
|
342
|
+
return ResultList.parse_results res[k]
|
137
343
|
end
|
138
344
|
|
139
345
|
end
|
data/lib/mirador/version.rb
CHANGED
data/test/test_mirador.rb
CHANGED
@@ -1,5 +1,6 @@
|
|
1
1
|
require 'test/unit'
|
2
|
-
require 'mirador'
|
2
|
+
require './lib/mirador'
|
3
|
+
require 'base64'
|
3
4
|
|
4
5
|
class MiradorTest < Test::Unit::TestCase
|
5
6
|
|
@@ -11,15 +12,14 @@ class MiradorTest < Test::Unit::TestCase
|
|
11
12
|
SFW_URL = "http://demo.mirador.im/test/sfw.jpg"
|
12
13
|
NSFW_URL = "http://demo.mirador.im/test/nsfw.jpg"
|
13
14
|
|
14
|
-
MM = Mirador::Client.new('')
|
15
|
+
MM = Mirador::Client.new(ENV['MIRADOR_API_KEY'])
|
15
16
|
|
16
17
|
def test_classify_files
|
17
18
|
|
18
19
|
res = MM.classify_files([NSFW_IM, SFW_IM])
|
19
|
-
|
20
20
|
assert_equal res.length, 2
|
21
21
|
|
22
|
-
nsfw, sfw = res
|
22
|
+
nsfw, sfw = res[NSFW_IM], res[SFW_IM]
|
23
23
|
|
24
24
|
assert_operator nsfw.value, :>=, 0.50
|
25
25
|
assert_operator sfw.value, :<, 0.50
|
@@ -27,92 +27,147 @@ class MiradorTest < Test::Unit::TestCase
|
|
27
27
|
assert nsfw.name.eql?(NSFW_IM), "nsfw name does not match"
|
28
28
|
assert sfw.name.eql?(SFW_IM), "sfw name does not match"
|
29
29
|
|
30
|
+
assert nsfw.id.eql?(NSFW_IM)
|
31
|
+
assert sfw.id.eql?(SFW_IM)
|
32
|
+
|
30
33
|
assert sfw.safe
|
31
34
|
assert (not nsfw.safe)
|
32
35
|
|
33
36
|
end
|
34
37
|
|
35
|
-
def
|
36
|
-
nsfw_files = [NSFW_IM]*10
|
37
|
-
sfw_files = [SFW_IM]*10
|
38
|
+
def test_classify_urls
|
38
39
|
|
39
|
-
|
40
|
-
assert_equal
|
40
|
+
res = MM.classify_urls([NSFW_URL, SFW_URL])
|
41
|
+
assert_equal 2, res.length
|
41
42
|
|
42
|
-
|
43
|
-
assert_operator r.value, :>=, 0.50
|
44
|
-
assert r.name.eql?(NSFW_IM)
|
45
|
-
assert (not r.safe)
|
46
|
-
end
|
43
|
+
nsfw, sfw = res[NSFW_URL], res[SFW_URL]
|
47
44
|
|
48
|
-
|
49
|
-
|
45
|
+
assert_operator nsfw.value, :>=, 0.50
|
46
|
+
assert_operator sfw.value, :<, 0.50
|
47
|
+
|
48
|
+
assert nsfw.name.eql?(NSFW_URL), "nsfw name does not match"
|
49
|
+
assert sfw.name.eql?(SFW_URL), "sfw name does not match"
|
50
|
+
|
51
|
+
assert nsfw.id.eql?(NSFW_URL)
|
52
|
+
assert sfw.id.eql?(SFW_URL)
|
53
|
+
|
54
|
+
assert sfw.safe
|
55
|
+
assert (not nsfw.safe)
|
50
56
|
|
51
|
-
sres.each do |r|
|
52
|
-
assert_operator r.value, :<, 0.50
|
53
|
-
assert r.name.eql?(SFW_IM)
|
54
|
-
assert r.safe
|
55
|
-
end
|
56
57
|
end
|
57
58
|
|
58
|
-
def
|
59
|
-
nsfw_urls = [NSFW_URL]*10
|
60
|
-
sfw_urls = [SFW_URL]*10
|
59
|
+
def test_classify_chunked_urls
|
61
60
|
|
62
|
-
|
63
|
-
|
61
|
+
r = Hash[([NSFW_URL]*10).each_with_index.map do |url, idx|
|
62
|
+
[ "#{ idx }-im", url ]
|
63
|
+
end]
|
64
64
|
|
65
|
-
|
66
|
-
assert_not_nil r
|
67
|
-
assert_operator r.value, :>=, 0.50
|
68
|
-
assert r.name.eql?(NSFW_URL)
|
69
|
-
assert (not r.safe)
|
70
|
-
end
|
65
|
+
res = MM.classify_urls(r)
|
71
66
|
|
72
|
-
|
73
|
-
assert_equal sres.length, 10
|
67
|
+
assert_equal 10, res.length
|
74
68
|
|
75
|
-
|
76
|
-
|
77
|
-
assert_not_nil r.value
|
78
|
-
assert_operator r.value, :<, 0.50
|
79
|
-
assert r.name.eql?(SFW_URL)
|
80
|
-
assert r.safe
|
69
|
+
res.each do |id, r|
|
70
|
+
assert_operator r.value, :>=, 0.50
|
81
71
|
end
|
82
72
|
|
83
73
|
end
|
84
74
|
|
85
|
-
def
|
86
|
-
res = MM.classify_urls([NSFW_URL, SFW_URL])
|
75
|
+
def test_hash_call
|
87
76
|
|
88
|
-
|
89
|
-
|
77
|
+
res = MM.classify_urls(nsfw: NSFW_URL, sfw: SFW_URL)
|
78
|
+
|
79
|
+
assert res[:nsfw]
|
80
|
+
assert res[:sfw]
|
90
81
|
|
91
|
-
|
92
|
-
|
82
|
+
nsfw = res[:nsfw]
|
83
|
+
sfw = res[:sfw]
|
93
84
|
|
94
85
|
assert_operator nsfw.value, :>=, 0.50
|
95
86
|
assert_operator sfw.value, :<, 0.50
|
96
87
|
|
88
|
+
assert nsfw.name.eql?('nsfw'), "nsfw name does not match"
|
89
|
+
assert sfw.name.eql?('sfw'), "sfw name does not match"
|
90
|
+
|
91
|
+
assert nsfw.id.eql?('nsfw')
|
92
|
+
assert sfw.id.eql?('sfw')
|
93
|
+
|
97
94
|
assert sfw.safe
|
98
95
|
assert (not nsfw.safe)
|
96
|
+
|
99
97
|
end
|
100
98
|
|
101
|
-
def
|
102
|
-
|
103
|
-
res = MM.
|
99
|
+
def test_single_url
|
100
|
+
|
101
|
+
res = MM.classify_url(nsfw: NSFW_URL)
|
102
|
+
res1 = MM.classify_url(NSFW_URL)
|
103
|
+
|
104
|
+
assert_equal res.value, res1.value
|
105
|
+
|
106
|
+
end
|
107
|
+
|
108
|
+
def test_items_call
|
109
|
+
|
110
|
+
res = MM.classify_urls([{ id: :nsfw, data: NSFW_URL }, { id: :sfw, data: SFW_URL }])
|
104
111
|
|
105
112
|
assert_equal res.length, 2
|
106
|
-
nsfw, sfw = res
|
107
113
|
|
108
|
-
assert
|
109
|
-
assert sfw
|
114
|
+
assert res[:nsfw]
|
115
|
+
assert res[:sfw]
|
110
116
|
|
111
|
-
|
112
|
-
|
117
|
+
end
|
118
|
+
|
119
|
+
def test_classify_buffers
|
120
|
+
|
121
|
+
bufs = Hash[([File.read(NSFW_IM)]*3).each_with_index.map do |b, idx|
|
122
|
+
["#{idx}-buf", b]
|
123
|
+
end]
|
124
|
+
|
125
|
+
|
126
|
+
res = MM.classify_buffers(bufs)
|
127
|
+
|
128
|
+
res1 = MM.classify_buffer(File.read(SFW_IM))
|
129
|
+
|
130
|
+
assert_equal res.length, 3
|
131
|
+
assert_operator res1.value, :<=, 0.50
|
132
|
+
|
133
|
+
res.each do |r|
|
134
|
+
assert_operator r.value, :>=, 0.5
|
135
|
+
end
|
136
|
+
end
|
137
|
+
|
138
|
+
def test_data_uris
|
139
|
+
duri = 'data:image/jpg;base64,' + Base64.encode64(File.read(NSFW_IM)).gsub(/\n/, '')
|
140
|
+
|
141
|
+
res = MM.classify_data_uris(nsfw: duri)
|
142
|
+
|
143
|
+
assert res[:nsfw]
|
144
|
+
assert_operator res[:nsfw].value, :>=, 0.50
|
145
|
+
end
|
146
|
+
|
147
|
+
def test_encoded_string
|
148
|
+
|
149
|
+
tdata = Hash[[SFW_IM, NSFW_IM].map do |fname|
|
150
|
+
[fname, Base64.encode64(File.read(fname))]
|
151
|
+
end]
|
152
|
+
|
153
|
+
res = MM.classify_encoded_strings(tdata)
|
154
|
+
|
155
|
+
assert res[SFW_IM]
|
156
|
+
assert res[NSFW_IM]
|
157
|
+
|
158
|
+
assert_operator res[NSFW_IM].value, :>=, 0.50
|
159
|
+
|
160
|
+
end
|
161
|
+
|
162
|
+
def test_item_error
|
163
|
+
|
164
|
+
res = MM.classify_urls([{ id: :nsfw, data: 'invalid-url'}, { id: :sfw, data: SFW_URL }])
|
165
|
+
|
166
|
+
assert_equal res.length, 2
|
167
|
+
assert res[:sfw]
|
168
|
+
|
169
|
+
assert res.any? do |r| r.failed? end
|
113
170
|
|
114
|
-
assert sfw.safe
|
115
|
-
assert (not nsfw.safe)
|
116
171
|
end
|
117
172
|
|
118
173
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: mirador
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0
|
4
|
+
version: 0.1.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Nick Jacob
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2014-
|
11
|
+
date: 2014-08-05 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: httparty
|