shex 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/AUTHORS +1 -0
- data/CREDITS +0 -0
- data/LICENSE +24 -0
- data/README.md +124 -0
- data/VERSION +1 -0
- data/etc/doap.ttl +33 -0
- data/lib/shex.rb +133 -0
- data/lib/shex/algebra.rb +39 -0
- data/lib/shex/algebra/and.rb +32 -0
- data/lib/shex/algebra/annotation.rb +6 -0
- data/lib/shex/algebra/base.rb +6 -0
- data/lib/shex/algebra/each_of.rb +53 -0
- data/lib/shex/algebra/external.rb +24 -0
- data/lib/shex/algebra/inclusion.rb +59 -0
- data/lib/shex/algebra/node_constraint.rb +148 -0
- data/lib/shex/algebra/not.rb +19 -0
- data/lib/shex/algebra/one_of.rb +50 -0
- data/lib/shex/algebra/operator.rb +274 -0
- data/lib/shex/algebra/or.rb +37 -0
- data/lib/shex/algebra/prefix.rb +6 -0
- data/lib/shex/algebra/satisfiable.rb +44 -0
- data/lib/shex/algebra/schema.rb +125 -0
- data/lib/shex/algebra/semact.rb +38 -0
- data/lib/shex/algebra/shape.rb +93 -0
- data/lib/shex/algebra/shape_ref.rb +46 -0
- data/lib/shex/algebra/start.rb +20 -0
- data/lib/shex/algebra/stem.rb +20 -0
- data/lib/shex/algebra/stem_range.rb +42 -0
- data/lib/shex/algebra/triple_constraint.rb +72 -0
- data/lib/shex/algebra/triple_expression.rb +46 -0
- data/lib/shex/algebra/unary_shape.rb +6 -0
- data/lib/shex/algebra/value.rb +28 -0
- data/lib/shex/meta.rb +7914 -0
- data/lib/shex/parser.rb +801 -0
- data/lib/shex/terminals.rb +106 -0
- metadata +224 -0
data/lib/shex/parser.rb
ADDED
@@ -0,0 +1,801 @@
|
|
1
|
+
require 'ebnf'
|
2
|
+
require 'ebnf/ll1/parser'
|
3
|
+
require 'shex/meta'
|
4
|
+
|
5
|
+
module ShEx
|
6
|
+
##
|
7
|
+
# A parser for the ShEx grammar.
|
8
|
+
#
|
9
|
+
# @see https://www.w3.org/2005/01/yacker/uploads/ShEx3?lang=perl&markup=html#productions
|
10
|
+
# @see http://en.wikipedia.org/wiki/LR_parser
|
11
|
+
class Parser
|
12
|
+
include ShEx::Meta
|
13
|
+
include ShEx::Terminals
|
14
|
+
include EBNF::LL1::Parser
|
15
|
+
|
16
|
+
##
|
17
|
+
# Any additional options for the parser.
|
18
|
+
#
|
19
|
+
# @return [Hash]
|
20
|
+
attr_reader :options
|
21
|
+
|
22
|
+
##
|
23
|
+
# The current input string being processed.
|
24
|
+
#
|
25
|
+
# @return [String]
|
26
|
+
attr_accessor :input
|
27
|
+
|
28
|
+
##
|
29
|
+
# The current input tokens being processed.
|
30
|
+
#
|
31
|
+
# @return [Array<Token>]
|
32
|
+
attr_reader :tokens
|
33
|
+
|
34
|
+
##
|
35
|
+
# The internal representation of the result using hierarchy of RDF objects and ShEx::Operator
|
36
|
+
# objects.
|
37
|
+
# @return [Array]
|
38
|
+
# @see http://sparql.rubyforge.org/algebra
|
39
|
+
attr_accessor :result
|
40
|
+
|
41
|
+
# Terminals passed to lexer. Order matters!
|
42
|
+
terminal(:CODE, CODE, unescape: true) do |prod, token, input|
|
43
|
+
# { foo %}
|
44
|
+
# Keep surrounding whitespace for now
|
45
|
+
input[:code] = token.value[1..-2].sub(/%\s*$/, '') # Drop {} and %
|
46
|
+
end
|
47
|
+
terminal(:REPEAT_RANGE, REPEAT_RANGE) do |prod, token, input|
|
48
|
+
card = token.value[1..-2].split(',').map {|v| v =~ /^\d+$/ ? v.to_i : v}
|
49
|
+
card[1] = token.value.include?(',') ? '*' : card[0] if card.length == 1
|
50
|
+
input[:cardinality] = {min: card[0], max: card[1]}
|
51
|
+
end
|
52
|
+
terminal(:BLANK_NODE_LABEL, BLANK_NODE_LABEL) do |prod, token, input|
|
53
|
+
input[:blankNode] = bnode(token.value[2..-1])
|
54
|
+
end
|
55
|
+
terminal(:IRIREF, IRIREF, unescape: true) do |prod, token, input|
|
56
|
+
begin
|
57
|
+
input[:iri] = iri(token.value[1..-2])
|
58
|
+
rescue ArgumentError => e
|
59
|
+
raise Error, e.message
|
60
|
+
end
|
61
|
+
end
|
62
|
+
terminal(:DOUBLE, DOUBLE) do |prod, token, input|
|
63
|
+
# Note that a Turtle Double may begin with a '.[eE]', so tack on a leading
|
64
|
+
# zero if necessary
|
65
|
+
value = token.value.sub(/\.([eE])/, '.0\1')
|
66
|
+
input[:literal] = literal(value, datatype: RDF::XSD.double)
|
67
|
+
end
|
68
|
+
terminal(:DECIMAL, DECIMAL) do |prod, token, input|
|
69
|
+
# Note that a Turtle Decimal may begin with a '.', so tack on a leading
|
70
|
+
# zero if necessary
|
71
|
+
value = token.value
|
72
|
+
#value = "0#{token.value}" if token.value[0,1] == "."
|
73
|
+
input[:literal] = literal(value, datatype: RDF::XSD.decimal)
|
74
|
+
end
|
75
|
+
terminal(:INTEGER, INTEGER) do |prod, token, input|
|
76
|
+
input[:literal] = literal(token.value, datatype: RDF::XSD.integer)
|
77
|
+
end
|
78
|
+
terminal(:PNAME_LN, PNAME_LN, unescape: true) do |prod, token, input|
|
79
|
+
prefix, suffix = token.value.split(":", 2)
|
80
|
+
input[:iri] = ns(prefix, suffix)
|
81
|
+
error(nil, "Compact IRI missing prefix definition: #{token.value}", production: :PNAME_LN) unless input[:iri].absolute?
|
82
|
+
end
|
83
|
+
terminal(:PNAME_NS, PNAME_NS) do |prod, token, input|
|
84
|
+
prefix = token.value[0..-2]
|
85
|
+
|
86
|
+
input[:iri] = ns(prefix, nil)
|
87
|
+
input[:prefix] = prefix && prefix.to_sym
|
88
|
+
end
|
89
|
+
terminal(:ATPNAME_LN, ATPNAME_LN, unescape: true) do |prod, token, input|
|
90
|
+
prefix, suffix = token.value.split(":", 2)
|
91
|
+
prefix.sub!(/^@#{WS}*/, '')
|
92
|
+
input[:shapeLabel] = ns(prefix, suffix)
|
93
|
+
error(nil, "Compact IRI missing prefix definition: #{token.value}", production: :ATPNAME_LN) unless input[:shapeLabel].absolute?
|
94
|
+
end
|
95
|
+
terminal(:ATPNAME_NS, ATPNAME_NS) do |prod, token, input|
|
96
|
+
prefix = token.value[0..-2]
|
97
|
+
prefix.sub!(/^@\s*/, '')
|
98
|
+
|
99
|
+
input[:shapeLabel] = ns(prefix, nil)
|
100
|
+
end
|
101
|
+
terminal(:LANGTAG, LANGTAG) do |prod, token, input|
|
102
|
+
input[:language] = token.value[1..-1]
|
103
|
+
end
|
104
|
+
terminal(:STRING_LITERAL_LONG1, STRING_LITERAL_LONG1, unescape: true) do |prod, token, input|
|
105
|
+
input[:string] = token.value[3..-4]
|
106
|
+
end
|
107
|
+
terminal(:STRING_LITERAL_LONG2, STRING_LITERAL_LONG2, unescape: true) do |prod, token, input|
|
108
|
+
input[:string] = token.value[3..-4]
|
109
|
+
end
|
110
|
+
terminal(:STRING_LITERAL1, STRING_LITERAL1, unescape: true) do |prod, token, input|
|
111
|
+
input[:string] = token.value[1..-2]
|
112
|
+
end
|
113
|
+
terminal(:STRING_LITERAL2, STRING_LITERAL2, unescape: true) do |prod, token, input|
|
114
|
+
input[:string] = token.value[1..-2]
|
115
|
+
end
|
116
|
+
terminal(:RDF_TYPE, RDF_TYPE) do |prod, token, input|
|
117
|
+
input[:iri] = (a = RDF.type.dup; a.lexical = 'a'; a)
|
118
|
+
end
|
119
|
+
|
120
|
+
# String terminals
|
121
|
+
terminal(nil, STR_EXPR, map: STR_MAP) do |prod, token, input|
|
122
|
+
case token.value
|
123
|
+
when '*' then input[:cardinality] = {min: 0, max: "*"}
|
124
|
+
when '+' then input[:cardinality] = {min: 1, max: "*"}
|
125
|
+
when '?' then input[:cardinality] = {min: 0, max: 1}
|
126
|
+
when '!' then input[:not] = token.value
|
127
|
+
when '^' then input[:inverse] = token.value
|
128
|
+
when '.' then input[:dot] = token.value
|
129
|
+
when 'true', 'false' then input[:literal] = RDF::Literal::Boolean.new(token.value)
|
130
|
+
when '~' then input[:pattern] = token.value
|
131
|
+
when 'BNODE', 'IRI',
|
132
|
+
'NONLITERAL' then input[:nonLiteralKind] = token.value.downcase.to_sym
|
133
|
+
when 'CLOSED' then input[:closed] = token.value.downcase.to_sym
|
134
|
+
when 'EXTERNAL' then input[:external] = token.value.downcase.to_sym
|
135
|
+
when 'FRACTIONDIGITS',
|
136
|
+
'TOTALDIGITS' then input[:numericLength] = token.value.downcase.to_sym
|
137
|
+
when 'LITERAL' then input[:shapeAtomLiteral] = token.value.downcase.to_sym
|
138
|
+
when 'LENGTH',
|
139
|
+
'MINLENGTH',
|
140
|
+
'MAXLENGTH' then input[:stringLength] = token.value.downcase.to_sym
|
141
|
+
when 'MININCLUSIVE',
|
142
|
+
'MINEXCLUSIVE',
|
143
|
+
'MAXINCLUSIVE',
|
144
|
+
'MAXEXCLUSIVE' then input[:numericRange] = token.value.downcase.to_sym
|
145
|
+
when 'NOT' then input[:not] = token.value.downcase.to_sym
|
146
|
+
when 'PATTERN' then input[:pattern] = token.value.downcase.to_sym
|
147
|
+
when 'START' then input[:start] = token.value.downcase.to_sym
|
148
|
+
else
|
149
|
+
#raise "Unexpected MC terminal: #{token.inspect}"
|
150
|
+
end
|
151
|
+
end
|
152
|
+
|
153
|
+
# Productions
|
154
|
+
# [1] shexDoc ::= directive* ((notStartAction | startActions) statement*)?
|
155
|
+
production(:shexDoc) do |input, data, callback|
|
156
|
+
data[:start] = data[:start] if data[:start]
|
157
|
+
|
158
|
+
expressions = []
|
159
|
+
expressions << [:base, data[:baseDecl]] if data[:baseDecl]
|
160
|
+
expressions << [:prefix, data[:prefixDecl]] if data[:prefixDecl]
|
161
|
+
expressions += Array(data[:codeDecl])
|
162
|
+
expressions << Algebra::Start.new(data[:start]) if data[:start]
|
163
|
+
expressions << [:shapes, data[:shapes]] if data[:shapes]
|
164
|
+
|
165
|
+
input[:schema] = Algebra::Schema.new(*expressions, options)
|
166
|
+
|
167
|
+
# Set schema accessor for all included expressions
|
168
|
+
input[:schema].each_descendant do |op|
|
169
|
+
op.schema = input[:schema] if op.respond_to?(:schema=)
|
170
|
+
end
|
171
|
+
self
|
172
|
+
end
|
173
|
+
|
174
|
+
# [2] directive ::= baseDecl | prefixDecl
|
175
|
+
|
176
|
+
# [3] baseDecl ::= "BASE" IRIREF
|
177
|
+
production(:baseDecl) do |input, data, callback|
|
178
|
+
input[:baseDecl] = self.base_uri = iri(data[:iri])
|
179
|
+
end
|
180
|
+
|
181
|
+
# [4] prefixDecl ::= "PREFIX" PNAME_NS IRIREF
|
182
|
+
production(:prefixDecl) do |input, data, callback|
|
183
|
+
pfx = data[:prefix]
|
184
|
+
self.prefix(pfx, data[:iri])
|
185
|
+
(input[:prefixDecl] ||= []) << [pfx.to_s, data[:iri]]
|
186
|
+
end
|
187
|
+
|
188
|
+
# [5] notStartAction ::= start | shapeExprDecl
|
189
|
+
# [6] start ::= "start" '=' shapeExpression
|
190
|
+
production(:start) do |input, data, callback|
|
191
|
+
input[:start] = data[:shapeExpression]
|
192
|
+
end
|
193
|
+
# [7] startActions ::= codeDecl+
|
194
|
+
|
195
|
+
# [8] statement ::= directive | notStartAction
|
196
|
+
|
197
|
+
# [9] shapeExprDecl ::= shapeLabel (shapeExpression|"EXTERNAL")
|
198
|
+
production(:shapeExprDecl) do |input, data, callback|
|
199
|
+
label = Array(data[:shapeLabel]).first
|
200
|
+
expression = case data[:shapeExpression]
|
201
|
+
when Algebra::NodeConstraint, Algebra::Or, Algebra::And, Algebra::Not, Algebra::ShapeRef, Algebra::Shape
|
202
|
+
data[:shapeExpression]
|
203
|
+
else
|
204
|
+
data[:external] ? Algebra::External.new() : Algebra::Shape.new()
|
205
|
+
end
|
206
|
+
|
207
|
+
(input[:shapes] ||= {})[label] = expression
|
208
|
+
end
|
209
|
+
|
210
|
+
# [10] shapeExpression ::= shapeOr
|
211
|
+
# [11] inlineShapeExpression ::= inlineShapeOr
|
212
|
+
|
213
|
+
# [12] shapeOr ::= shapeAnd ("OR" shapeAnd)*
|
214
|
+
production(:shapeOr) do |input, data, callback|
|
215
|
+
shape_or(input, data)
|
216
|
+
end
|
217
|
+
# [13] inlineShapeOr ::= inlineShapeAnd ("OR" inlineShapeAnd)*
|
218
|
+
production(:inlineShapeOr) do |input, data, callback|
|
219
|
+
shape_or(input, data)
|
220
|
+
end
|
221
|
+
def shape_or(input, data)
|
222
|
+
input.merge!(data.dup.keep_if {|k, v| [:closed, :extraPropertySet, :codeDecl].include?(k)})
|
223
|
+
expression = if Array(data[:shapeExpression]).length > 1
|
224
|
+
Algebra::Or.new(*data[:shapeExpression])
|
225
|
+
else
|
226
|
+
Array(data[:shapeExpression]).first
|
227
|
+
end
|
228
|
+
input[:shapeExpression] = expression if expression
|
229
|
+
rescue ArgumentError => e
|
230
|
+
error(nil, "Argument Error on OR: #{e.message}")
|
231
|
+
end
|
232
|
+
private :shape_or
|
233
|
+
|
234
|
+
# [14] shapeAnd ::= shapeNot ("AND" shapeNot)*
|
235
|
+
production(:shapeAnd) do |input, data, callback|
|
236
|
+
shape_and(input, data)
|
237
|
+
end
|
238
|
+
# [15] inlineShapeAnd ::= inlineShapeNot ("AND" inlineShapeNot)*
|
239
|
+
production(:inlineShapeAnd) do |input, data, callback|
|
240
|
+
shape_and(input, data)
|
241
|
+
end
|
242
|
+
def shape_and(input, data)
|
243
|
+
input.merge!(data.dup.keep_if {|k, v| [:closed, :extraPropertySet, :codeDecl].include?(k)})
|
244
|
+
expressions = Array(data[:shapeExpression]).inject([]) do |memo, expr|
|
245
|
+
memo.concat(expr.is_a?(Algebra::And) ? expr.operands : [expr])
|
246
|
+
end
|
247
|
+
expression = if expressions.length > 1
|
248
|
+
Algebra::And.new(*expressions)
|
249
|
+
else
|
250
|
+
expressions.first
|
251
|
+
end
|
252
|
+
(input[:shapeExpression] ||= []) << expression if expression
|
253
|
+
rescue ArgumentError => e
|
254
|
+
error(nil, "Argument Error on AND: #{e.message}")
|
255
|
+
end
|
256
|
+
private :shape_and
|
257
|
+
|
258
|
+
# [16] shapeNot ::= "NOT"? shapeAtom
|
259
|
+
production(:shapeNot) do |input, data, callback|
|
260
|
+
shape_not(input, data)
|
261
|
+
end
|
262
|
+
# [17] inlineShapeNot ::= "NOT"? inlineShapeAtom
|
263
|
+
production(:inlineShapeNot) do |input, data, callback|
|
264
|
+
shape_not(input, data)
|
265
|
+
end
|
266
|
+
def shape_not(input, data)
|
267
|
+
input.merge!(data.dup.keep_if {|k, v| [:closed, :extraPropertySet, :codeDecl].include?(k)})
|
268
|
+
expression = data[:shapeExpression]
|
269
|
+
expression = Algebra::Not.new(expression) if data[:not]
|
270
|
+
#error(nil, "Expected an atom for NOT") unless expression
|
271
|
+
(input[:shapeExpression] ||= []) << expression if expression
|
272
|
+
end
|
273
|
+
private :shape_not
|
274
|
+
|
275
|
+
# [18] shapeAtom ::= nodeConstraint shapeOrRef?
|
276
|
+
# | shapeOrRef
|
277
|
+
# | "(" shapeExpression ")"
|
278
|
+
# | '.' # no constraint
|
279
|
+
production(:shapeAtom) do |input, data, callback|
|
280
|
+
shape_atom(input, data)
|
281
|
+
end
|
282
|
+
# [19] inlineShapeAtom ::= nodeConstraint inlineShapeOrRef?
|
283
|
+
# | inlineShapeOrRef nodeConstraint?
|
284
|
+
# | "(" shapeExpression ")"
|
285
|
+
# | '.' # no constraint
|
286
|
+
production(:inlineShapeAtom) do |input, data, callback|
|
287
|
+
shape_atom(input, data)
|
288
|
+
end
|
289
|
+
def shape_atom(input, data)
|
290
|
+
constraint = data[:nodeConstraint]
|
291
|
+
shape = data[:shapeOrRef] || data[:shapeExpression]
|
292
|
+
input.merge!(data.dup.keep_if {|k, v| [:closed, :extraPropertySet, :codeDecl].include?(k)})
|
293
|
+
|
294
|
+
expression = [constraint, shape].compact
|
295
|
+
expression = case expression.length
|
296
|
+
when 0 then nil
|
297
|
+
when 1
|
298
|
+
expression.first
|
299
|
+
else Algebra::And.new(*expression)
|
300
|
+
end
|
301
|
+
|
302
|
+
input[:shapeExpression] = expression if expression
|
303
|
+
end
|
304
|
+
private :shape_atom
|
305
|
+
|
306
|
+
# [20] shapeOrRef ::= ATPNAME_LN | ATPNAME_NS | '@' shapeLabel | shapeDefinition
|
307
|
+
production(:shapeOrRef) do |input, data, callback|
|
308
|
+
shape_or_ref(input, data)
|
309
|
+
end
|
310
|
+
# [21] inlineShapeOrRef ::= ATPNAME_LN | ATPNAME_NS | '@' shapeLabel | inlineShapeDefinition
|
311
|
+
production(:inlineShapeOrRef) do |input, data, callback|
|
312
|
+
shape_or_ref(input, data)
|
313
|
+
end
|
314
|
+
def shape_or_ref(input, data)
|
315
|
+
input.merge!(data.dup.keep_if {|k, v| [:closed, :extraPropertySet, :codeDecl].include?(k)})
|
316
|
+
if data[:shape] || Array(data[:shapeLabel]).first
|
317
|
+
input[:shapeOrRef] = data[:shape] || Algebra::ShapeRef.new(Array(data[:shapeLabel]).first)
|
318
|
+
end
|
319
|
+
rescue ArgumentError => e
|
320
|
+
error(nil, "Argument Error on ShapeOrRef: #{e.message}")
|
321
|
+
end
|
322
|
+
private :shape_or_ref
|
323
|
+
|
324
|
+
# [22] nodeConstraint ::= "LITERAL" xsFacet*
|
325
|
+
# | nonLiteralKind stringFacet*
|
326
|
+
# | datatype xsFacet*
|
327
|
+
# | valueSet xsFacet*
|
328
|
+
# | xsFacet+
|
329
|
+
production(:nodeConstraint) do |input, data, callback|
|
330
|
+
# Semantic validate (A Syntax error)
|
331
|
+
case
|
332
|
+
when data[:datatype] && data[:numericFacet]
|
333
|
+
# Can only use a numeric facet on a numeric datatype
|
334
|
+
l = RDF::Literal.new("1", datatype: data[:datatype])
|
335
|
+
error(nil, "Numeric facet used on non-numeric datatype: #{data[:datatype]}", production: :nodeConstraint) unless l.is_a?(RDF::Literal::Numeric)
|
336
|
+
end
|
337
|
+
|
338
|
+
attrs = []
|
339
|
+
attrs += [:datatype, data[:datatype]] if data [:datatype]
|
340
|
+
attrs += [data[:shapeAtomLiteral], data[:nonLiteralKind]]
|
341
|
+
attrs += Array(data[:valueSetValue])
|
342
|
+
attrs += Array(data[:numericFacet])
|
343
|
+
attrs += Array(data[:stringFacet])
|
344
|
+
|
345
|
+
input[:nodeConstraint] = Algebra::NodeConstraint.new(*attrs.compact)
|
346
|
+
end
|
347
|
+
|
348
|
+
# [23] nonLiteralKind ::= "IRI" | "BNODE" | "NONLITERAL"
|
349
|
+
|
350
|
+
# [24] xsFacet ::= stringFacet | numericFacet
|
351
|
+
# [25] stringFacet ::= stringLength INTEGER
|
352
|
+
# | "PATTERN" string
|
353
|
+
# | '~' string # shortcut for "PATTERN"
|
354
|
+
production(:stringFacet) do |input, data, callback|
|
355
|
+
input[:stringFacet] ||= []
|
356
|
+
input[:stringFacet] << if data[:stringLength]
|
357
|
+
if input[:stringFacet].flatten.include?(data[:stringLength])
|
358
|
+
error(nil, "#{data[:stringLength]} constraint may only be used once in a Node Constraint", production: :stringFacet)
|
359
|
+
end
|
360
|
+
[data[:stringLength], data[:literal]]
|
361
|
+
elsif data[:pattern]
|
362
|
+
[:pattern, data[:string]]
|
363
|
+
end
|
364
|
+
end
|
365
|
+
|
366
|
+
# [26] stringLength ::= "LENGTH" | "MINLENGTH" | "MAXLENGTH"
|
367
|
+
|
368
|
+
# [27] numericFacet ::= numericRange (numericLiteral | string '^^' datatype )
|
369
|
+
# | numericLength INTEGER
|
370
|
+
production(:numericFacet) do |input, data, callback|
|
371
|
+
input[:numericFacet] ||= []
|
372
|
+
input[:numericFacet] << if data[:numericRange]
|
373
|
+
literal = data[:literal] || literal(data[:string], datatype: data[:datatype])
|
374
|
+
error(nil, "numericRange must use a numeric datatype: #{data[:datatype]}", production: :numericFacet) unless literal.is_a?(RDF::Literal::Numeric)
|
375
|
+
[data[:numericRange], literal]
|
376
|
+
elsif data[:numericLength]
|
377
|
+
[data[:numericLength], data[:literal]]
|
378
|
+
end
|
379
|
+
end
|
380
|
+
|
381
|
+
# [28] numericRange ::= "MININCLUSIVE" | "MINEXCLUSIVE" | "MAXINCLUSIVE" | "MAXEXCLUSIVE"
|
382
|
+
# [29] numericLength ::= "TOTALDIGITS" | "FRACTIONDIGITS"
|
383
|
+
|
384
|
+
# [30] shapeDefinition ::= (extraPropertySet | "CLOSED")* '{' tripleExpression? '}' annotation* semanticActions
|
385
|
+
production(:shapeDefinition) do |input, data, callback|
|
386
|
+
shape_definition(input, data)
|
387
|
+
end
|
388
|
+
# [31] inlineShapeDefinition ::= (extraPropertySet | "CLOSED")* '{' tripleExpression? '}'
|
389
|
+
production(:inlineShapeDefinition) do |input, data, callback|
|
390
|
+
shape_definition(input, data)
|
391
|
+
end
|
392
|
+
def shape_definition(input, data)
|
393
|
+
expression = data[:tripleExpression]
|
394
|
+
attrs = Array(data[:extraPropertySet])
|
395
|
+
attrs << :closed if data[:closed]
|
396
|
+
attrs += Array(data[:annotation])
|
397
|
+
attrs += Array(data[:codeDecl])
|
398
|
+
|
399
|
+
input[:shape] = Algebra::Shape.new(expression, *attrs) if expression
|
400
|
+
end
|
401
|
+
private :shape_definition
|
402
|
+
|
403
|
+
# [32] extraPropertySet ::= "EXTRA" predicate+
|
404
|
+
production(:extraPropertySet) do |input, data, callback|
|
405
|
+
(input[:extraPropertySet] ||= []) << data[:predicate].unshift(:extra)
|
406
|
+
end
|
407
|
+
|
408
|
+
# [33] tripleExpression ::= oneOfTripleExpr
|
409
|
+
# [34] oneOfTripleExpr ::= groupTripleExpr ('|' groupTripleExpr)*
|
410
|
+
production(:oneOfTripleExpr) do |input, data, callback|
|
411
|
+
expression = if Array(data[:tripleExpression]).length > 1
|
412
|
+
Algebra::OneOf.new(*data[:tripleExpression])
|
413
|
+
else
|
414
|
+
Array(data[:tripleExpression]).first
|
415
|
+
end
|
416
|
+
input[:tripleExpression] = expression if expression
|
417
|
+
end
|
418
|
+
|
419
|
+
# [37] groupTripleExpr ::= unaryTripleExpr (';' unaryTripleExpr?)*
|
420
|
+
production(:groupTripleExpr) do |input, data, callback|
|
421
|
+
expression = if Array(data[:tripleExpression]).length > 1
|
422
|
+
Algebra::EachOf.new(*data[:tripleExpression])
|
423
|
+
else
|
424
|
+
Array(data[:tripleExpression]).first
|
425
|
+
end
|
426
|
+
(input[:tripleExpression] ||= []) << expression if expression
|
427
|
+
end
|
428
|
+
|
429
|
+
# [40] unaryTripleExpr ::= productionLabel? (tripleConstraint | bracketedTripleExpr) | include
|
430
|
+
production(:unaryTripleExpr) do |input, data, callback|
|
431
|
+
expression = data[:tripleExpression]
|
432
|
+
expression.operands << data[:productionLabel] if expression && data[:productionLabel]
|
433
|
+
|
434
|
+
(input[:tripleExpression] ||= []) << expression if expression
|
435
|
+
end
|
436
|
+
|
437
|
+
# [41] bracketedTripleExpr ::= '(' oneOfTripleExpr ')' cardinality? annotation* semanticActions
|
438
|
+
production(:bracketedTripleExpr) do |input, data, callback|
|
439
|
+
# XXX cardinality? annotation* semanticActions
|
440
|
+
case expression = data[:tripleExpression]
|
441
|
+
when Algebra::OneOf, Algebra::EachOf
|
442
|
+
else
|
443
|
+
error(nil, "Bracketed Expression requires multiple contained expressions", production: :bracketedTripleExpr)
|
444
|
+
end
|
445
|
+
cardinality = data.fetch(:cardinality, {})
|
446
|
+
attrs = [
|
447
|
+
([:min, cardinality[:min]] if cardinality[:min]),
|
448
|
+
([:max, cardinality[:max]] if cardinality[:max])
|
449
|
+
].compact
|
450
|
+
attrs += Array(data[:codeDecl])
|
451
|
+
attrs += Array(data[:annotation])
|
452
|
+
|
453
|
+
expression.operands.concat(attrs)
|
454
|
+
input[:tripleExpression] = expression
|
455
|
+
end
|
456
|
+
|
457
|
+
# [42] productionLabel ::= '$' (iri | blankNode)
|
458
|
+
production(:productionLabel) do |input, data, callback|
|
459
|
+
input[:productionLabel] = data[:iri] || data[:blankNode]
|
460
|
+
end
|
461
|
+
|
462
|
+
# [43] tripleConstraint ::= senseFlags? predicate shapeExpression cardinality? annotation* semanticActions
|
463
|
+
production(:tripleConstraint) do |input, data, callback|
|
464
|
+
cardinality = data.fetch(:cardinality, {})
|
465
|
+
attrs = [
|
466
|
+
(:inverse if data[:inverse] || data[:not]),
|
467
|
+
Array(data[:predicate]).first, # predicate
|
468
|
+
data[:shapeExpression],
|
469
|
+
([:min, cardinality[:min]] if cardinality[:min]),
|
470
|
+
([:max, cardinality[:max]] if cardinality[:max])
|
471
|
+
].compact
|
472
|
+
attrs += Array(data[:codeDecl])
|
473
|
+
attrs += Array(data[:annotation])
|
474
|
+
|
475
|
+
input[:tripleExpression] = Algebra::TripleConstraint.new(*attrs) unless attrs.empty?
|
476
|
+
end
|
477
|
+
|
478
|
+
# [44] cardinality ::= '*' | '+' | '?' | REPEAT_RANGE
|
479
|
+
# [45] senseFlags ::= '^'
|
480
|
+
# [46] valueSet ::= '[' valueSetValue* ']'
|
481
|
+
|
482
|
+
# [47] valueSetValue ::= iriRange | literal
|
483
|
+
production(:valueSetValue) do |input, data, callback|
|
484
|
+
(input[:valueSetValue] ||= []) << Algebra::Value.new(data[:iriRange] || data[:literal])
|
485
|
+
end
|
486
|
+
|
487
|
+
# [48] iriRange ::= iri ('~' exclusion*)? | '.' exclusion+
|
488
|
+
production(:iriRange) do |input, data, callback|
|
489
|
+
exclusions = data[:exclusion].unshift(:exclusions) if data[:exclusion]
|
490
|
+
input[:iriRange] = if data[:pattern] && exclusions
|
491
|
+
Algebra::StemRange.new(data[:iri], exclusions)
|
492
|
+
elsif data[:pattern]
|
493
|
+
Algebra::Stem.new(data[:iri])
|
494
|
+
elsif data[:dot]
|
495
|
+
Algebra::StemRange.new(:wildcard, exclusions)
|
496
|
+
else
|
497
|
+
data[:iri]
|
498
|
+
end
|
499
|
+
end
|
500
|
+
|
501
|
+
# [49] exclusion ::= '-' iri '~'?
|
502
|
+
production(:exclusion) do |input, data, callback|
|
503
|
+
(input[:exclusion] ||= []) << (data[:pattern] ? Algebra::Stem.new(data[:iri]) : data[:iri])
|
504
|
+
end
|
505
|
+
|
506
|
+
# [50] include ::= '&' shapeLabel
|
507
|
+
production(:include) do |input, data, callback|
|
508
|
+
input[:tripleExpression] = Algebra::Inclusion.new(*data[:shapeLabel])
|
509
|
+
end
|
510
|
+
|
511
|
+
# [51] annotation ::= '//' predicate (iri | literal)
|
512
|
+
production(:annotation) do |input, data, callback|
|
513
|
+
annotation = Algebra::Annotation.new(data[:predicate].first, (data[:iri] || data[:literal]))
|
514
|
+
(input[:annotation] ||= []) << annotation
|
515
|
+
end
|
516
|
+
|
517
|
+
# [52] semanticActions ::= codeDecl*
|
518
|
+
|
519
|
+
# [53] codeDecl ::= '%' iri (CODE | "%")
|
520
|
+
production(:codeDecl) do |input, data, callback|
|
521
|
+
(input[:codeDecl] ||= []) << Algebra::SemAct.new(*[data[:iri], data[:code]].compact)
|
522
|
+
end
|
523
|
+
|
524
|
+
# [13t] literal ::= rdfLiteral | numericLiteral | booleanLiteral
|
525
|
+
|
526
|
+
# [54] predicate ::= iri | RDF_TYPE
|
527
|
+
production(:predicate) do |input, data, callback|
|
528
|
+
(input[:predicate] ||= []) << data[:iri]
|
529
|
+
end
|
530
|
+
|
531
|
+
# [55] datatype ::= iri
|
532
|
+
production(:datatype) do |input, data, callback|
|
533
|
+
input[:datatype] = data[:iri]
|
534
|
+
end
|
535
|
+
|
536
|
+
# [56] shapeLabel ::= iri | blankNode
|
537
|
+
production(:shapeLabel) do |input, data, callback|
|
538
|
+
(input[:shapeLabel] ||= []) << (data[:iri] || data[:blankNode])
|
539
|
+
end
|
540
|
+
|
541
|
+
# [16t] numericLiteral ::= INTEGER | DECIMAL | DOUBLE
|
542
|
+
# [129s] rdfLiteral ::= string (LANGTAG | '^^' datatype)?
|
543
|
+
production(:rdfLiteral) do |input, data, callback|
|
544
|
+
input[:literal] = literal(data[:string], data)
|
545
|
+
end
|
546
|
+
|
547
|
+
# [134s] booleanLiteral ::= 'true' | 'false'
|
548
|
+
# [135s] string ::= STRING_LITERAL1 | STRING_LITERAL_LONG1
|
549
|
+
# | STRING_LITERAL2 | STRING_LITERAL_LONG2
|
550
|
+
# [136s] iri ::= IRIREF | prefixedName
|
551
|
+
# [137s] prefixedName ::= PNAME_LN | PNAME_NS
|
552
|
+
# [138s] blankNode ::= BLANK_NODE_LABEL
|
553
|
+
|
554
|
+
##
|
555
|
+
# Initializes a new parser instance.
|
556
|
+
#
|
557
|
+
# @example parsing a ShExC schema
|
558
|
+
# schema = ShEx::Parser.new(%(
|
559
|
+
# PREFIX ex: <http://schema.example/> ex:IssueShape {ex:state IRI}
|
560
|
+
# ).parse
|
561
|
+
#
|
562
|
+
# @param [String, IO, StringIO, #to_s] input
|
563
|
+
# @param [Hash{Symbol => Object}] options
|
564
|
+
# @option options [Hash] :prefixes (Hash.new)
|
565
|
+
# the prefix mappings to use (for acessing intermediate parser productions)
|
566
|
+
# @option options [#to_s] :base_uri (nil)
|
567
|
+
# the base URI to use when resolving relative URIs (for acessing intermediate parser productions)
|
568
|
+
# @option options [#to_s] :anon_base ("b0")
|
569
|
+
# Basis for generating anonymous Nodes
|
570
|
+
# @option options [Boolean] :resolve_iris (false)
|
571
|
+
# Resolve prefix and relative IRIs, otherwise, when serializing the parsed SSE
|
572
|
+
# as S-Expressions, use the original prefixed and relative URIs along with `base` and `prefix`
|
573
|
+
# definitions.
|
574
|
+
# @option options [Boolean] :validate (false)
|
575
|
+
# whether to validate the parsed statements and values
|
576
|
+
# @option options [Boolean] :progress
|
577
|
+
# Show progress of parser productions
|
578
|
+
# @option options [Boolean] :debug
|
579
|
+
# Detailed debug output
|
580
|
+
# @yield [parser] `self`
|
581
|
+
# @yieldparam [ShEx::Parser] parser
|
582
|
+
# @yieldreturn [void] ignored
|
583
|
+
# @return [ShEx::Parser]
|
584
|
+
# @raise [ShEx::NotSatisfied] if not satisfied
|
585
|
+
# @raise [ShEx::ParseError] when a syntax error is detected
|
586
|
+
# @raise [ShEx::StructureError, ArgumentError] on structural problems with schema
|
587
|
+
def initialize(input = nil, options = {}, &block)
|
588
|
+
@input = case input
|
589
|
+
when IO, StringIO then input.read
|
590
|
+
else input.to_s.dup
|
591
|
+
end
|
592
|
+
@input.encode!(Encoding::UTF_8) if @input.respond_to?(:encode!)
|
593
|
+
@options = {anon_base: "b0", validate: false}.merge(options)
|
594
|
+
@options[:debug] ||= case
|
595
|
+
when options[:progress] then 2
|
596
|
+
when options[:validate] then 1
|
597
|
+
end
|
598
|
+
|
599
|
+
debug("base IRI") {base_uri.inspect}
|
600
|
+
debug("validate") {validate?.inspect}
|
601
|
+
|
602
|
+
if block_given?
|
603
|
+
case block.arity
|
604
|
+
when 0 then instance_eval(&block)
|
605
|
+
else block.call(self)
|
606
|
+
end
|
607
|
+
end
|
608
|
+
end
|
609
|
+
|
610
|
+
# @return [String]
|
611
|
+
def to_sxp_bin
|
612
|
+
@result
|
613
|
+
end
|
614
|
+
|
615
|
+
def to_s
|
616
|
+
@result.to_sxp
|
617
|
+
end
|
618
|
+
|
619
|
+
alias_method :ll1_parse, :parse
|
620
|
+
|
621
|
+
# Parse query
|
622
|
+
#
|
623
|
+
# The result is a SPARQL Algebra S-List. Productions return an array such as the following:
|
624
|
+
#
|
625
|
+
# (prefix ((: <http://example/>))
|
626
|
+
# (union
|
627
|
+
# (bgp (triple ?s ?p ?o))
|
628
|
+
# (graph ?g
|
629
|
+
# (bgp (triple ?s ?p ?o)))))
|
630
|
+
#
|
631
|
+
# @param [Symbol, #to_s] prod The starting production for the parser.
|
632
|
+
# It may be a URI from the grammar, or a symbol representing the local_name portion of the grammar URI.
|
633
|
+
# @return [Array]
|
634
|
+
# @see http://www.w3.org/TR/sparql11-query/#sparqlAlgebra
|
635
|
+
# @see http://axel.deri.ie/sparqltutorial/ESWC2007_SPARQL_Tutorial_unit2b.pdf
|
636
|
+
def parse(prod = START)
|
637
|
+
ll1_parse(@input, prod.to_sym, @options.merge(branch: BRANCH,
|
638
|
+
first: FIRST,
|
639
|
+
follow: FOLLOW,
|
640
|
+
whitespace: WS)
|
641
|
+
) do |context, *data|
|
642
|
+
case context
|
643
|
+
when :trace
|
644
|
+
level, lineno, depth, *args = data
|
645
|
+
message = args.to_sse
|
646
|
+
d_str = depth > 100 ? ' ' * 100 + '+' : ' ' * depth
|
647
|
+
str = "[#{lineno}](#{level})#{d_str}#{message}".chop
|
648
|
+
case @options[:debug]
|
649
|
+
when Array
|
650
|
+
@options[:debug] << str unless level > 2
|
651
|
+
when TrueClass
|
652
|
+
$stderr.puts str
|
653
|
+
when Integer
|
654
|
+
$stderr.puts(str) if level <= @options[:debug]
|
655
|
+
end
|
656
|
+
end
|
657
|
+
end
|
658
|
+
|
659
|
+
# The last thing on the @prod_data stack is the result
|
660
|
+
@result = case
|
661
|
+
when !prod_data.is_a?(Hash)
|
662
|
+
prod_data
|
663
|
+
when prod_data.empty?
|
664
|
+
nil
|
665
|
+
when prod_data[:schema]
|
666
|
+
prod_data[:schema]
|
667
|
+
else
|
668
|
+
key = prod_data.keys.first
|
669
|
+
[key] + Array(prod_data[key]) # Creates [:key, [:triple], ...]
|
670
|
+
end
|
671
|
+
|
672
|
+
# Validate resulting expression
|
673
|
+
@result.validate! if @result && validate?
|
674
|
+
@result
|
675
|
+
rescue EBNF::LL1::Parser::Error, EBNF::LL1::Lexer::Error => e
|
676
|
+
raise ShEx::ParseError.new(e.message, lineno: e.lineno, token: e.token)
|
677
|
+
end
|
678
|
+
|
679
|
+
private
|
680
|
+
##
|
681
|
+
# Returns the URI prefixes currently defined for this parser.
|
682
|
+
#
|
683
|
+
# @example
|
684
|
+
# prefixes[:dc] #=> RDF::URI('http://purl.org/dc/terms/')
|
685
|
+
#
|
686
|
+
# @return [Hash{Symbol => RDF::URI}]
|
687
|
+
# @since 0.3.0
|
688
|
+
def prefixes
|
689
|
+
@options[:prefixes] ||= {}
|
690
|
+
end
|
691
|
+
|
692
|
+
##
|
693
|
+
# Defines the given URI prefixes for this parser.
|
694
|
+
#
|
695
|
+
# @example
|
696
|
+
# prefixes = {
|
697
|
+
# dc: RDF::URI('http://purl.org/dc/terms/'),
|
698
|
+
# }
|
699
|
+
#
|
700
|
+
# @param [Hash{Symbol => RDF::URI}] prefixes
|
701
|
+
# @return [Hash{Symbol => RDF::URI}]
|
702
|
+
# @since 0.3.0
|
703
|
+
def prefixes=(prefixes)
|
704
|
+
@options[:prefixes] = prefixes
|
705
|
+
end
|
706
|
+
|
707
|
+
##
|
708
|
+
# Defines the given named URI prefix for this parser.
|
709
|
+
#
|
710
|
+
# @example Defining a URI prefix
|
711
|
+
# prefix :dc, RDF::URI('http://purl.org/dc/terms/')
|
712
|
+
#
|
713
|
+
# @example Returning a URI prefix
|
714
|
+
# prefix(:dc) #=> RDF::URI('http://purl.org/dc/terms/')
|
715
|
+
#
|
716
|
+
# @overload prefix(name, uri)
|
717
|
+
# @param [Symbol, #to_s] name
|
718
|
+
# @param [RDF::URI, #to_s] uri
|
719
|
+
#
|
720
|
+
# @overload prefix(name)
|
721
|
+
# @param [Symbol, #to_s] name
|
722
|
+
#
|
723
|
+
# @return [RDF::URI]
|
724
|
+
def prefix(name, iri = nil)
|
725
|
+
name = name.to_s.empty? ? nil : (name.respond_to?(:to_sym) ? name.to_sym : name.to_s.to_sym)
|
726
|
+
iri.nil? ? prefixes[name] : prefixes[name] = iri
|
727
|
+
end
|
728
|
+
|
729
|
+
##
|
730
|
+
# Returns the Base URI defined for the parser,
|
731
|
+
# as specified or when parsing a BASE prologue element.
|
732
|
+
#
|
733
|
+
# @example
|
734
|
+
# base #=> RDF::URI('http://example.com/')
|
735
|
+
#
|
736
|
+
# @return [HRDF::URI]
|
737
|
+
def base_uri
|
738
|
+
RDF::URI(@options[:base_uri])
|
739
|
+
end
|
740
|
+
|
741
|
+
##
|
742
|
+
# Set the Base URI to use for this parser.
|
743
|
+
#
|
744
|
+
# @param [RDF::URI, #to_s] iri
|
745
|
+
#
|
746
|
+
# @example
|
747
|
+
# base_uri = RDF::URI('http://purl.org/dc/terms/')
|
748
|
+
#
|
749
|
+
# @return [RDF::URI]
|
750
|
+
def base_uri=(iri)
|
751
|
+
@options[:base_uri] = RDF::URI(iri)
|
752
|
+
end
|
753
|
+
|
754
|
+
##
|
755
|
+
# Returns `true` when resolving IRIs, otherwise BASE and PREFIX are retained in the output algebra.
|
756
|
+
#
|
757
|
+
# @return [Boolean] `true` or `false`
|
758
|
+
# @since 1.0.3
|
759
|
+
def validate?
|
760
|
+
@options[:validate]
|
761
|
+
end
|
762
|
+
|
763
|
+
# Generate a BNode identifier
|
764
|
+
def bnode(id)
|
765
|
+
@bnode_cache ||= {}
|
766
|
+
raise Error, "Illegal attempt to reuse a BNode" if @bnode_cache[id] && @bnode_cache[id].frozen?
|
767
|
+
@bnode_cache[id] ||= RDF::Node.new(id)
|
768
|
+
end
|
769
|
+
|
770
|
+
# Create URIs
|
771
|
+
def iri(value)
|
772
|
+
# If we have a base URI, use that when constructing a new URI
|
773
|
+
value = RDF::URI(value)
|
774
|
+
if base_uri && value.relative?
|
775
|
+
base_uri.join(value)
|
776
|
+
else
|
777
|
+
value
|
778
|
+
end
|
779
|
+
end
|
780
|
+
|
781
|
+
def ns(prefix, suffix)
|
782
|
+
base = prefix(prefix).to_s
|
783
|
+
suffix = suffix.to_s.sub(/^\#/, "") if base.index("#")
|
784
|
+
debug {"ns(#{prefix.inspect}): base: '#{base}', suffix: '#{suffix}'"}
|
785
|
+
iri(base + suffix.to_s)
|
786
|
+
end
|
787
|
+
|
788
|
+
# Create a literal
|
789
|
+
def literal(value, options = {})
|
790
|
+
options = options.dup
|
791
|
+
# Internal representation is to not use xsd:string, although it could arguably go the other way.
|
792
|
+
options.delete(:datatype) if options[:datatype] == RDF::XSD.string
|
793
|
+
debug("literal") do
|
794
|
+
"value: #{value.inspect}, " +
|
795
|
+
"options: #{options.inspect}, " +
|
796
|
+
"validate: #{validate?.inspect}, "
|
797
|
+
end
|
798
|
+
RDF::Literal.new(value, options.merge(validate: validate?))
|
799
|
+
end
|
800
|
+
end # class Parser
|
801
|
+
end # module ShEx
|