Class: SPARQL::Algebra::Operator::Group

Inherits:
SPARQL::Algebra::Operator show all
Includes:
Query
Defined in:
vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb

Overview

The SPARQL group operator.

group takes either two or three operands. The first operand is an array of grouped variables. The last operand is the query to be grouped. If three operands are provided, the second is an array of temporary bindings.

[19] GroupClause ::= 'GROUP' 'BY' GroupCondition+

Examples:

SPARQL Grammar

PREFIX : <http://www.example.org>

SELECT ?P (COUNT(?O) AS ?C)
WHERE { ?S ?P ?O }
GROUP BY ?P

SSE

(prefix
 ((: <http://www.example.org>))
 (project (?P ?C)
  (extend ((?C ??.0))
   (group (?P) ((??.0 (count ?O)))
    (bgp (triple ?S ?P ?O))))))

SPARQL Grammar (HAVING aggregate)

PREFIX : <http://www.example.org/>
SELECT ?s (AVG(?o) AS ?avg)
WHERE { ?s ?p ?o }
GROUP BY ?s
HAVING (AVG(?o) <= 2.0)

SSE (HAVING aggregate)

(prefix ((: <http://www.example.org/>))
 (project (?s ?avg)
  (filter (<= ??.1 2.0)
   (extend ((?avg ??.0))
    (group (?s) ((??.0 (avg ?o)) (??.1 (avg ?o)))
     (bgp (triple ?s ?p ?o)))))))

SPARQL Grammar (non-trivial filters)

PREFIX : <http://example.com/data/#>
SELECT ?g (AVG(?p) AS ?avg) ((MIN(?p) + MAX(?p)) / 2 AS ?c)
WHERE { ?g :p ?p . }
GROUP BY ?g

SSE (non-trivial filters)

(prefix ((: <http://example.com/data/#>))
 (project (?g ?avg ?c)
  (extend ((?avg ??.0) (?c (/ (+ ??.1 ??.2) 2)))
   (group (?g)
          ((??.0 (avg ?p))
           (??.1 (min ?p))
           (??.2 (max ?p)))
    (bgp (triple ?g :p ?p)))) ))

See Also:

Constant Summary collapse

NAME = 
[:group]

Constants inherited from SPARQL::Algebra::Operator

ARITY, IsURI, URI

Constants included from Expression

Expression::PATTERN_PARENTS

Constants included from RDF::Util::Logger

RDF::Util::Logger::IOWrapper

Instance Attribute Summary

Attributes included from Query

#solutions

Attributes inherited from SPARQL::Algebra::Operator

#operands

Instance Method Summary collapse

Methods included from Query

#each_solution, #empty?, #failed?, #graph_name=, #matched?, #query_yields_boolean?, #query_yields_solutions?, #query_yields_statements?, #unshift

Methods inherited from SPARQL::Algebra::Operator

#aggregate?, arity, base_uri, #base_uri, base_uri=, #bind, #boolean, #constant?, #deep_dup, #each_descendant, #eql?, #evaluatable?, evaluate, #executable?, #first_ancestor, for, #formulae, #initialize, #inspect, #ndvars, #node?, #operand, #optimize, #optimize!, #parent, #parent=, #prefixes, prefixes, prefixes=, #rewrite, #to_binary, to_sparql, #to_sxp, #to_sxp_bin, #variable?, #vars

Methods included from Expression

cast, #constant?, #evaluate, extension, extension?, extensions, for, #invalid?, new, #node?, open, #optimize, #optimize!, parse, register_extension, #to_sxp_bin, #valid?, #variable?

Methods included from RDF::Util::Logger

#log_debug, #log_depth, #log_error, #log_fatal, #log_info, #log_recover, #log_recovering?, #log_statistics, #log_warn, #logger

Constructor Details

This class inherits a constructor from SPARQL::Algebra::Operator

Instance Method Details

#execute(queryable, **options) {|solution| ... } ⇒ RDF::Query::Solutions

Executes query with queryable and groups results based on the first operand.

Parameters:

Yields:

  • (solution) —

    each matching solution

Yield Parameters:

Yield Returns:

  • (void) —

    ignored

Returns:

See Also:



80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
 
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb', line 80

def execute(queryable, **options, &block)
  debug(options) {"Group"}
  exprlist = operands.first
  query = operands.last
  aggregates = operands.length == 3 ? operand(1) : []
  solutions = queryable.query(query, **options.merge(depth: options[:depth].to_i + 1))

  groups = solutions.group_by do |solution|
    # Evaluate each exprlist operand to get groups where each key is a new solution
    # ListEval((expr1, ..., exprn), μ) returns a list (e1, ..., en), where ei = expri(μ) or error.
    soln = RDF::Query::Solution.new
    exprlist.each do |operand|
      begin
        if operand.is_a?(Array)
          # Form is [variable, expression]
          soln[operand.first] = operand.last.evaluate(solution,
                                                      queryable: queryable,
                                                      depth: options[:depth].to_i + 1,
                                                      **options)
        else
          # Form is variable
          soln[operand] = operand.evaluate(solution, queryable: queryable,
                                                     depth: options[:depth].to_i + 1,
                                                     **options)
        end
      rescue TypeError
        # Ignore expression
      end
    end
    soln
  end

  debug(options) {"=>(groups) #{groups.inspect}"}

  # Aggregate solutions in each group using aggregates to get solutions
  @solutions = RDF::Query::Solutions(groups.map do |group_soln, solns|
    aggregates.each do |(var, aggregate)|
      begin
        group_soln[var] = aggregate.aggregate(solns, **options)
      rescue TypeError
        # Ignored in output
      end
    end
    group_soln
  end)

  # If there exprlist is empty, make sure that's at least an empty solution
  if @solutions.empty? && exprlist.empty?
    soln = RDF::Query::Solution.new
    aggregates.each do |(var, aggregate)|
      begin
        soln[var] = aggregate.aggregate([], **options)
      rescue TypeError
        # Ignored in output
      end
    end
    @solutions << soln
  end

  debug(options) {"=>(solutions) #{@solutions.inspect}"}
  @solutions.each(&block) if block_given?
  @solutions
end

#internal_variables ⇒ Hash{Symbol => RDF::Query::Variable}

The variables used within the query

Returns:



189
190
191
 
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb', line 189

def internal_variables
  operands.last.variables
end

#to_sparql(extensions: {}, filter_ops: [], **options) ⇒ String

Returns a partial SPARQL grammar for this operator.

Parameters:

Returns:



202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
 
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb', line 202

def to_sparql(extensions: {}, filter_ops: [], **options)
  having_ops = []
  if operands.length > 2
    temp_bindings = operands[1].inject({}) {|memo, (var, op)| memo.merge(var => op)}
    # Replace extensions from temporary bindings
    temp_bindings.each do |var, op|
      # Update extensions using a temporarily bound variable with its binding
      extensions = extensions.inject({}) do |memo, (ext_var, ext_op)|
        if ext_op.is_a?(Operator)
          # Try to recursivley replace variable within operator
          new_op = ext_op.deep_dup.rewrite do |operand|
            if operand.is_a?(Variable) && operand.to_sym == var.to_sym
              op.dup
            else
              operand
            end
          end
          memo.merge(ext_var.to_s => new_op)
        elsif ext_op.is_a?(Variable) && ext_op.to_sym == var.to_sym
          memo.merge(ext_var.to_s => op)
        else
          # Doesn't match this variable, so don't change
          memo.merge(ext_var.to_s => ext_op)
        end
      end

      # Filter ops using temporary bindinds are used for HAVING clauses
      filter_ops.each do |fop|
        having_ops << fop if fop.descendants.include?(var) && !having_ops.include?(fop)
      end
    end

    # If used in a HAVING clause, it's not also a filter
    filter_ops -= having_ops

    # Replace each operand in having using var with it's corresponding operation
    having_ops = having_ops.map do |op|
      op.dup.rewrite do |operand|
        # Rewrite based on temporary bindings
        temp_bindings.fetch(operand, operand)
      end
    end
  end
  operands.last.to_sparql(extensions: extensions,
                          group_ops: operands.first,
                          having_ops: having_ops,
                          **options)
end

#validate! ⇒ Object

It is an error for aggregates to project variables with a name already used in other aggregate projections, or in the WHERE clause.

It is also an error to project ungrouped variables



147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
 
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb', line 147

def validate!
  group_vars = operand(0).map {|v| Array(v).first}
  ext = first_ancestor(Extend)
  extend_vars = ext ? ext.operand(0).map(&:first).select {|v| v.is_a?(RDF::Query::Variable)} : []
  project = first_ancestor(Project)
  # If not projecting, were are effectively projecting all variables in the query
  project_vars = project ? project.operand(0) : operands.last.vars

  available_vars = (extend_vars + group_vars).compact

  # All variables must either be grouped or extended
  unless (project_vars - available_vars).empty?
    raise ArgumentError,
         "projecting ungrouped/extended variables: #{(project_vars.compact - available_vars.compact).to_sse}"
  end
  super
end

#variables ⇒ Hash{Symbol => RDF::Query::Variable}

The variables used in the extension. Includes grouped variables and temporary, but not those in the query, itself

Returns:



170
171
172
173
174
175
176
177
178
179
180
181
182
183
 
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/sparql-36baa432eb7f/lib/sparql/algebra/operator/group.rb', line 170

def variables
  group_vars = operands.first

  aggregate_vars = (operands.length == 3 ? operand(1) : [])

  # Extract first element of each and merge it's variables
  (group_vars + aggregate_vars).
    map do  |o|
      v = Array(o).first
      v if v.is_a?(RDF::Query::Variable)
    end.compact.
    map(&:variables).
    inject({}) {|memo, h| memo.merge(h)}
end