bxtjson 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/bxtjson.rb +285 -0
- metadata +91 -0
data/lib/bxtjson.rb
ADDED
@@ -0,0 +1,285 @@
|
|
1
|
+
# input stream style (new-line separated) json objects
|
2
|
+
# main method map_onto_skeleton_of_schema
|
3
|
+
# JSON -> JSON
|
4
|
+
|
5
|
+
require 'multi_json'
|
6
|
+
require 'json_schema'
|
7
|
+
# == dependencies
|
8
|
+
# * gem: json_schema
|
9
|
+
# * gem: multi_json
|
10
|
+
# * JSON standard library or other (e.g. oj) json parsers
|
11
|
+
# * You implement a model in Sequel or ActiveRecord if you want.
|
12
|
+
# == Constants (examples)
|
13
|
+
# cleaner_proc = ->(str) {str.gsub(/\W+/, " ").lstrip
|
14
|
+
# .gsub(" ", "_")
|
15
|
+
# .gsub(/PPPO_|PPCO_/, "")
|
16
|
+
# .downcase
|
17
|
+
# }
|
18
|
+
# json_filename = "../../../fsms-tmp/PP_PROPOSAL_2015.JSON"
|
19
|
+
# schema_filename = "./docs/schema.json"
|
20
|
+
# authorizinng_pointer '#/departments/primary_dept'
|
21
|
+
|
22
|
+
module Bxtjson
|
23
|
+
# Initialize an empty hashmap given a json-schema[http://json-schema.org]
|
24
|
+
#
|
25
|
+
# @param [Hash] schema_data the data from json-schema
|
26
|
+
# @param [String, nil] entity the resource name to enfocus
|
27
|
+
# @return [Hash] the an empty "initialized" json-schema
|
28
|
+
def self.skeleton(schema_data:,
|
29
|
+
entity: nil)
|
30
|
+
schema = JsonSchema.parse!(schema_data)
|
31
|
+
schema.expand_references!
|
32
|
+
if entity.nil?
|
33
|
+
entity_schema = schema
|
34
|
+
else
|
35
|
+
entity_schema = schema.properties[entity]
|
36
|
+
end
|
37
|
+
return _skeleton(entity_schema, acc: {})
|
38
|
+
end
|
39
|
+
# process a file of jsonl (linefeed) and clean keys with proc
|
40
|
+
#
|
41
|
+
# @param [String] json_filename source of json objects in jsonl[https://github.com/stephenplusplus/jsonl] format
|
42
|
+
# @param [Proc] a function to clean up keys
|
43
|
+
# @return [Hash] hash of arrays `[{ }]`
|
44
|
+
def self.text_to_lazy_json(json_filename:,
|
45
|
+
clean_proc:)
|
46
|
+
File.foreach(json_filename)
|
47
|
+
.lazy
|
48
|
+
.map do |line|
|
49
|
+
_key_cleaner(data: MultiJson.load(line), clean_proc: clean_proc)
|
50
|
+
end
|
51
|
+
end
|
52
|
+
# Parse json-schema file and map contents of json file into
|
53
|
+
# initialized schema
|
54
|
+
#
|
55
|
+
# Mapping of contents will search for the first key in the source,
|
56
|
+
# that match the schema (recursively). So, the source should be
|
57
|
+
# flat for clarity, and the schema can be nested.
|
58
|
+
#
|
59
|
+
# If it cannot find a key, it will look for "top/next/final" path
|
60
|
+
# key in the source data.
|
61
|
+
#
|
62
|
+
# For example, in the skeleton
|
63
|
+
# {key: {nest: "this"} }
|
64
|
+
# Will be filled with "muscle" if the source has a ket
|
65
|
+
# {"key/nest": "data muscle"}
|
66
|
+
#
|
67
|
+
# TODO: design interface from csv to json that fits these
|
68
|
+
# principles.
|
69
|
+
#
|
70
|
+
# @param [String] json_filename filename for jsonl source
|
71
|
+
# @param [String] schema_filename filename for json-schema
|
72
|
+
# @param [Proc] clean_proc a function to clean up keys
|
73
|
+
# @param [String, #create] model the name of a model to call :create on
|
74
|
+
# @param [String] authorizing_pointer json-pointer[https://tools.ietf.org/html/rfc6901] that fills in the key "authorized_by"
|
75
|
+
def self.muscle(json_filename:,
|
76
|
+
schema_filename:,
|
77
|
+
clean_proc: ->(str){str},
|
78
|
+
model: nil,
|
79
|
+
schema_entity: nil,
|
80
|
+
authorizing_pointer:,
|
81
|
+
data_attr: :data)
|
82
|
+
skeleton = Bxtjson.skeleton(schema_data: MultiJson.load(File.read(schema_filename)),
|
83
|
+
entity: schema_entity)
|
84
|
+
if model
|
85
|
+
model = constantize(model.to_s.capitalize)
|
86
|
+
|
87
|
+
text_to_lazy_json(json_filename: json_filename, clean_proc: clean_proc )
|
88
|
+
.map {|data|
|
89
|
+
data = fillin(source_hash: _map_onto_skeleton_of_schema( data,
|
90
|
+
skeleton: skeleton ),
|
91
|
+
skeleton: skeleton)
|
92
|
+
result = model.create( data_attr => data)
|
93
|
+
}
|
94
|
+
else
|
95
|
+
out = []
|
96
|
+
text_to_lazy_json(json_filename: json_filename, clean_proc: clean_proc )
|
97
|
+
.map {|data| out << fillin(source_hash: _map_onto_skeleton_of_schema( data,
|
98
|
+
skeleton: skeleton ),
|
99
|
+
skeleton: skeleton)
|
100
|
+
}
|
101
|
+
end
|
102
|
+
end
|
103
|
+
# Recursively remove falsey values from hash
|
104
|
+
# Falsey values are those that return true from respond_to(:empty?)
|
105
|
+
# or :nil?
|
106
|
+
# @param [Hash] hash
|
107
|
+
# @return [Hash]
|
108
|
+
def self.compact_hash!(hash)
|
109
|
+
p = proc do |_, v|
|
110
|
+
v.delete_if(&p) if v.respond_to? :delete_if
|
111
|
+
v.respond_to?(:empty?) && v.empty? || v.nil?
|
112
|
+
end
|
113
|
+
hash.delete_if(&p)
|
114
|
+
end
|
115
|
+
def self.compact_values!(hash)
|
116
|
+
Hash[hash.map do |key, value|
|
117
|
+
[key,
|
118
|
+
if value.is_a?(Array)
|
119
|
+
value.map {|item| Bxtjson.compact_hash!(item) }
|
120
|
+
elsif value.respond_to?( :delete_if)
|
121
|
+
Bxtjson.compact_hash!(value)
|
122
|
+
else
|
123
|
+
value
|
124
|
+
end
|
125
|
+
]
|
126
|
+
end
|
127
|
+
]
|
128
|
+
end
|
129
|
+
private
|
130
|
+
# Creates a skeleton for object and array from a Json Schema
|
131
|
+
# Boolean, String, Number, Integer, Null are given a nil value to start.
|
132
|
+
# Hash -> Hash
|
133
|
+
def self._skeleton(json_schema, acc={})
|
134
|
+
case json_schema.type
|
135
|
+
when ["object"]
|
136
|
+
acc = Hash[json_schema.properties.map do |key, value|
|
137
|
+
[key, _skeleton(value, acc)]
|
138
|
+
end
|
139
|
+
]
|
140
|
+
when ["array"] # at this point the key is already in the Hash,
|
141
|
+
# just need to return an array with one hash
|
142
|
+
acc = [
|
143
|
+
json_schema.items.properties.map { |key, value|
|
144
|
+
[key, _skeleton(value, acc)]
|
145
|
+
}.to_h
|
146
|
+
]
|
147
|
+
else
|
148
|
+
return nil
|
149
|
+
end
|
150
|
+
return acc
|
151
|
+
end
|
152
|
+
# given a key, return value of lookup recursively
|
153
|
+
# if that lookup fails, try by path
|
154
|
+
# (String, Hash) -> Hash
|
155
|
+
def self.lookup(key, source_hash, path=[])
|
156
|
+
source_hash.fetch(key, nil) || source_hash.fetch(path.join("/"), nil)
|
157
|
+
end
|
158
|
+
|
159
|
+
# Take an array of hashes with a hash that contains values to
|
160
|
+
# insert. Expand the arrays into objects
|
161
|
+
# (e.g. key: [1,2,3] -> [{key: 1}, {key: 2}, {key: 3})
|
162
|
+
# (Array, Hash) -> {[]}
|
163
|
+
def self.expand_array_to_objects(array:, source_hash: )
|
164
|
+
matrix = array.first.map do |key, _|
|
165
|
+
# if a plain string put into array. Flatten all others.
|
166
|
+
[lookup(key, source_hash)].flatten.map {|value|
|
167
|
+
# zipmap behavior here so that if one array is shorter
|
168
|
+
# the result is nill when mapped against longer array
|
169
|
+
# ["a"].zip ["a", "b"] | reverse # => {"a":"a", "b":nil]
|
170
|
+
# h = Hash[ [[value].zip( [key]).map(&:reverse).flatten ] ]
|
171
|
+
[value].zip( [key]).map(&:reverse).flatten
|
172
|
+
|
173
|
+
}
|
174
|
+
end
|
175
|
+
# pad the array if current array length is not eq max length of
|
176
|
+
# arrays. Pad is the first element. This is how some reports treat
|
177
|
+
# repeating values (a la sql reporting)
|
178
|
+
sorted = matrix.sort_by(&:length)
|
179
|
+
max = sorted.last.length
|
180
|
+
sorted.map {|item| # padding done here. Second element in array,
|
181
|
+
# below, could be nil. Todo: paramaterize that as option
|
182
|
+
item.fill( [sorted.first.first[0], sorted.first.first[1] ], (item.length)..(max - 1) )
|
183
|
+
}
|
184
|
+
sorted
|
185
|
+
|
186
|
+
# transpose keeping a slot if empty (like a speadsheet)
|
187
|
+
head, *tail = sorted
|
188
|
+
(head.zip *tail).map(&:to_h)
|
189
|
+
|
190
|
+
end
|
191
|
+
|
192
|
+
# given a source_hash, find the first key from a skeleton hash
|
193
|
+
# and insert value. Depends on flat source hash
|
194
|
+
# remember the path during lookup with skeleton
|
195
|
+
# (Hash, Hash) -> Hash
|
196
|
+
# a bit lost here
|
197
|
+
def self.fillin(source_hash:, skeleton:, acc: {}, path: [])
|
198
|
+
case
|
199
|
+
when skeleton.kind_of?( Hash )
|
200
|
+
acc = Hash[skeleton.map do |key, value|
|
201
|
+
path.push key # save hash depth to stack-like []
|
202
|
+
# recurse on skeleton levels
|
203
|
+
[
|
204
|
+
[ path.last, (fillin(source_hash: source_hash,
|
205
|
+
skeleton: nil,
|
206
|
+
acc: lookup(key, source_hash, path),
|
207
|
+
path: path) or
|
208
|
+
fillin(source_hash: source_hash,
|
209
|
+
skeleton: value,
|
210
|
+
path: path))
|
211
|
+
],
|
212
|
+
path.pop # pop the path at end of recursion,
|
213
|
+
# and drop from returned array
|
214
|
+
][0]
|
215
|
+
end
|
216
|
+
]
|
217
|
+
when (skeleton.kind_of?( Array) and skeleton.first.empty?)
|
218
|
+
# when an array with no inner objects/hashmaps
|
219
|
+
acc = lookup(path.last, source_hash)
|
220
|
+
when skeleton.kind_of?( Array )
|
221
|
+
# when an array (eg Key: [1,2,3]) but we want obj: [{key:1}, {key: 2}]
|
222
|
+
acc = expand_array_to_objects( array: skeleton,
|
223
|
+
source_hash: source_hash)
|
224
|
+
when skeleton.nil? # the acc value should be a string, so join if possible
|
225
|
+
if acc.respond_to?(:join)
|
226
|
+
acc = acc.join
|
227
|
+
elsif acc.respond_to?(:empty?)
|
228
|
+
acc = acc.empty? ? nil : acc
|
229
|
+
else
|
230
|
+
acc = acc
|
231
|
+
end
|
232
|
+
else
|
233
|
+
acc = nil
|
234
|
+
end
|
235
|
+
return acc
|
236
|
+
end
|
237
|
+
# loop through hash, cleaning keys
|
238
|
+
# of note: if a "key/key" pointer where key == key then only the
|
239
|
+
# value of the nested key will be returned. Use a naming convention
|
240
|
+
# of "keys/key" or "unique/uniqueNest"
|
241
|
+
# Hash -> Hash
|
242
|
+
def self._map_onto_skeleton_of_schema(json_data,
|
243
|
+
acc: {},
|
244
|
+
skeleton:)
|
245
|
+
|
246
|
+
case
|
247
|
+
when json_data.kind_of?(Hash)
|
248
|
+
acc = Hash[json_data.map do |key, value|
|
249
|
+
[key,
|
250
|
+
_map_onto_skeleton_of_schema(value,
|
251
|
+
acc: acc,
|
252
|
+
skeleton: skeleton)
|
253
|
+
]
|
254
|
+
end
|
255
|
+
]
|
256
|
+
when json_data.kind_of?(Array)
|
257
|
+
acc = json_data.map do |item|
|
258
|
+
_map_onto_skeleton_of_schema(item, skeleton: skeleton)
|
259
|
+
end
|
260
|
+
else
|
261
|
+
acc = json_data
|
262
|
+
end
|
263
|
+
end
|
264
|
+
def self._key_cleaner(data:, clean_proc: ->(str){str}, acc: {})
|
265
|
+
case
|
266
|
+
when data.kind_of?(Hash)
|
267
|
+
acc = Hash[ data.map do |key, value|
|
268
|
+
[ clean_proc.call(key), _key_cleaner(data: value) ]
|
269
|
+
end
|
270
|
+
]
|
271
|
+
else
|
272
|
+
acc = data
|
273
|
+
end
|
274
|
+
acc
|
275
|
+
end
|
276
|
+
# File activesupport/lib/active_support/inflector.rb, line 278
|
277
|
+
def self.constantize(camel_cased_word)
|
278
|
+
unless /\A(?:::)?([A-Z]\w*(?:::[A-Z]\w*)*)\z/ =~ camel_cased_word
|
279
|
+
raise NameError, "#{camel_cased_word.inspect} is not a valid constant name!"
|
280
|
+
end
|
281
|
+
|
282
|
+
Object.module_eval("::#{$1}", __FILE__, __LINE__)
|
283
|
+
end
|
284
|
+
end
|
285
|
+
|
metadata
ADDED
@@ -0,0 +1,91 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: bxtjson
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Jacob Kroeze
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2015-03-26 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: multi_json
|
16
|
+
requirement: !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ~>
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: '1.10'
|
22
|
+
- - ! '>='
|
23
|
+
- !ruby/object:Gem::Version
|
24
|
+
version: 1.10.0
|
25
|
+
type: :runtime
|
26
|
+
prerelease: false
|
27
|
+
version_requirements: !ruby/object:Gem::Requirement
|
28
|
+
none: false
|
29
|
+
requirements:
|
30
|
+
- - ~>
|
31
|
+
- !ruby/object:Gem::Version
|
32
|
+
version: '1.10'
|
33
|
+
- - ! '>='
|
34
|
+
- !ruby/object:Gem::Version
|
35
|
+
version: 1.10.0
|
36
|
+
- !ruby/object:Gem::Dependency
|
37
|
+
name: json_schema
|
38
|
+
requirement: !ruby/object:Gem::Requirement
|
39
|
+
none: false
|
40
|
+
requirements:
|
41
|
+
- - ~>
|
42
|
+
- !ruby/object:Gem::Version
|
43
|
+
version: '0.5'
|
44
|
+
- - ! '>='
|
45
|
+
- !ruby/object:Gem::Version
|
46
|
+
version: 0.5.0
|
47
|
+
type: :runtime
|
48
|
+
prerelease: false
|
49
|
+
version_requirements: !ruby/object:Gem::Requirement
|
50
|
+
none: false
|
51
|
+
requirements:
|
52
|
+
- - ~>
|
53
|
+
- !ruby/object:Gem::Version
|
54
|
+
version: '0.5'
|
55
|
+
- - ! '>='
|
56
|
+
- !ruby/object:Gem::Version
|
57
|
+
version: 0.5.0
|
58
|
+
description: initialize empty hash from schema, map from hash to schema, return a
|
59
|
+
lazy enumerable
|
60
|
+
email: jlkroeze@gmail.com
|
61
|
+
executables: []
|
62
|
+
extensions: []
|
63
|
+
extra_rdoc_files: []
|
64
|
+
files:
|
65
|
+
- lib/bxtjson.rb
|
66
|
+
homepage: https://github.com/jacob-kroeze/bxtjson
|
67
|
+
licenses:
|
68
|
+
- MIT
|
69
|
+
post_install_message:
|
70
|
+
rdoc_options: []
|
71
|
+
require_paths:
|
72
|
+
- lib
|
73
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
74
|
+
none: false
|
75
|
+
requirements:
|
76
|
+
- - ! '>='
|
77
|
+
- !ruby/object:Gem::Version
|
78
|
+
version: '0'
|
79
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
80
|
+
none: false
|
81
|
+
requirements:
|
82
|
+
- - ! '>='
|
83
|
+
- !ruby/object:Gem::Version
|
84
|
+
version: '0'
|
85
|
+
requirements: []
|
86
|
+
rubyforge_project:
|
87
|
+
rubygems_version: 1.8.23
|
88
|
+
signing_key:
|
89
|
+
specification_version: 3
|
90
|
+
summary: Map between json schema
|
91
|
+
test_files: []
|