Class: Pegex::Parser

Inherits:
Object
  • Object
show all
Defined in:
lib/pegex/parser.rb

Defined Under Namespace

Classes: PegexParseError

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize {|_self| ... } ⇒ Parser

Returns a new instance of Parser.

Yields:

  • (_self)

Yield Parameters:

  • _self (Pegex::Parser) —

    the object that the method was called on



17
18
19
20
21
22
23
24
 
# File 'lib/pegex/parser.rb', line 17

def initialize
  @position = 0
  @farthest = 0
  @optimized = false
  @throw_on_error = true
  @debug = ENV['RUBY_PEGEX_DEBUG'] || $PegexParserDebug || false
  yield self if block_given?
end

Instance Attribute Details

#debug ⇒ Object

Returns the value of attribute debug.



15
16
17
 
# File 'lib/pegex/parser.rb', line 15

def debug
  @debug
end

#grammar ⇒ Object

Returns the value of attribute grammar.



9
10
11
 
# File 'lib/pegex/parser.rb', line 9

def grammar
  @grammar
end

#input ⇒ Object

Returns the value of attribute input.



11
12
13
 
# File 'lib/pegex/parser.rb', line 11

def input
  @input
end

#parent ⇒ Object

Returns the value of attribute parent.



13
14
15
 
# File 'lib/pegex/parser.rb', line 13

def parent
  @parent
end

#receiver ⇒ Object

Returns the value of attribute receiver.



10
11
12
 
# File 'lib/pegex/parser.rb', line 10

def receiver
  @receiver
end

#rule ⇒ Object

Returns the value of attribute rule.



14
15
16
 
# File 'lib/pegex/parser.rb', line 14

def rule
  @rule
end

Instance Method Details

#format_error(msg) ⇒ Object 



258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
 
# File 'lib/pegex/parser.rb', line 258

def format_error msg
  buffer = @buffer
  position = @farthest
  real_pos = @position

  line = buffer[0, position].scan(/\n/).size + 1
  column = position - (buffer.rindex("\n", position) || -1)

  pretext = @buffer[
    position < 50 ? 0 : position - 50,
    position < 50 ? position : 50
  ]
  context = @buffer[position, 50]
  pretext.gsub! /.*\n/m, ''
  context.gsub! /\n/, "\\n"

  return <<"..."
Error parsing Pegex document:
msg:      #{msg}
line:     #{line}
column:   #{column}
context:  #{pretext}#{context}
#{' ' * (pretext.length + 10)}^
position: #{position} (#{real_pos} pre-lookahead)
...
end

#match_all(list, parent = nil) ⇒ Object 



192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
 
# File 'lib/pegex/parser.rb', line 192

def match_all list, parent=nil
  position = @position
  set = []
  len = 0
  list.each do |elem|
    if match = match_next(elem)
      if !(elem['+asr'] or elem['-skip'])
        set.concat(match)
        len += 1
      end
    else
      @farthest = position if (@position = position) > @farthest
      return false
    end
  end
  set = [set] if len > 1
  return set
end

#match_any(list, parent = nil) ⇒ Object 



211
212
213
214
215
216
217
218
 
# File 'lib/pegex/parser.rb', line 211

def match_any list, parent=nil
  list.each do |elem|
    if (match = match_next(elem))
      return match
    end
  end
  return false
end

#match_err(error, parent = nil) ⇒ Object 



220
221
222
 
# File 'lib/pegex/parser.rb', line 220

def match_err error, parent=nil
  throw_error(error)
end

#match_next(next_) ⇒ Object 



123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
 
# File 'lib/pegex/parser.rb', line 123

def match_next next_
  return match_next_with_sep(next_) if next_['.sep']

  rule, method, kind, min, max, assertion =
    next_.values_at 'rule', 'method', 'kind', '+min', '+max', '+asr'

  position, match, count = @position, [], 0

  while return_ = method.call(rule, next_)
    position = @position unless assertion
    count += 1
    match.concat return_
    break if max == 1
  end
  if max != 1
    match = [match]
    @farthest = position if (@position = position) > @farthest
  end
  result = (count >= min and (max == 0 or count <= max)) ^ (assertion == -1)
  if not(result) or assertion
    @farthest = position if (@position = position) > @farthest
  end

  return result ? next_['-skip'] ? [] : match : false
end

#match_next_with_sep(next_) ⇒ Object 



149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
 
# File 'lib/pegex/parser.rb', line 149

def match_next_with_sep next_
  rule, method, kind, min, max, sep =
    next_.values_at 'rule', 'method', 'kind', '+min', '+max', '.sep'

  position, match, count, scount, smin, smax =
    @position, [], 0, 0, *(sep.values_at('+min', '+max'))
  while return_ = method.call(rule, next_)
    position = @position
    count += 1
    match.concat(return_)
    return_ = match_next(sep) or break
    match.concat(smax == 1 ? return_ : return_[0]) if !return_.empty?
    scount += 1
  end
  match = [match] if max != 1
  result = count >= min and (max == 0 or count <= max)
  if count == scount and not sep['+eok']
    @farthest = position if (@position = position) > @farthest
  end

  return(result ? next_['-skip'] ? [] : match : false)
end

#match_ref(ref, parent) ⇒ Object 



172
173
174
175
176
177
178
179
 
# File 'lib/pegex/parser.rb', line 172

def match_ref ref, parent
  rule = @tree[ref]
  match = match_next(rule) or return false
  return Pegex::Constant::Dummy unless rule['action']
  @rule, @parent = ref, parent
  result = rule['action'].call(match.first)
  return (result.equal? Pegex::Constant::Null) ? result : [result]
end

#match_ref_trace(ref, parent) ⇒ Object 



224
225
226
227
228
229
230
231
232
233
234
235
 
# File 'lib/pegex/parser.rb', line 224

def match_ref_trace ref, parent
  rule = @tree[ref]
  trace = ! rule['+asr']
  trace("try_#{ref}") if trace
  result = nil
  if (result = match_ref(ref, parent))
    trace("got_#{ref}") if trace
  else
    trace("not_#{ref}") if trace
  end
  return result
end

#match_rgx(regexp, parent = nil) ⇒ Object 



181
182
183
184
185
186
187
188
189
190
 
# File 'lib/pegex/parser.rb', line 181

def match_rgx regexp, parent=nil
  buffer = @buffer[@position .. -1]
  (m = buffer.match regexp) or return false
  @position += m[0].length
  # TODO use m.captures
  match = m[1..-1]
  match = [ match ] if m.length > 2
  @farthest = @position if @position > @farthest
  return match
end

#optimize_grammar(start) ⇒ Object 



76
77
78
79
80
81
82
83
84
 
# File 'lib/pegex/parser.rb', line 76

def optimize_grammar start
  return if @optimized
  @tree.each_pair do |name, node|
    next if node.kind_of? String
    optimize_node(node)
  end
  optimize_node('.ref' => start)
  @optimized = true
end

#optimize_node(node) ⇒ Object 



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
 
# File 'lib/pegex/parser.rb', line 86

def optimize_node node
  ['ref', 'rgx', 'all', 'any', 'err', 'code', 'xxx'].each do |kind|
    fail if kind == 'xxx'
    if node['rule'] = node[".#{kind}"]
      node['kind'] = kind
      node['method'] = self.method "match_#{kind}"
      break
    end
  end
  min, max = node.values_at '+min', '+max'
  node['+min'] ||= max.nil? ? 1 : 0
  node['+max'] ||= min.nil? ? 1 : 0
  node['+asr'] ||= nil
  node['+min'] = node['+min'].to_i
  node['+max'] = node['+max'].to_i

  if ['any', 'all'].include? node['kind']
    node['rule'].each do |elem|
      optimize_node elem
    end
  elsif node['kind'] == 'ref'
    ref = node['rule']
    rule = @tree[ref]
    if @receiver.respond_to? "got_#{ref}"
      rule['action'] = receiver.method "got_#{ref}"
    elsif receiver.respond_to? 'gotrule'
      rule['action'] = receiver.method 'gotrule'
    end
    node['method'] = self.method 'match_ref_trace' if @debug
  elsif node['kind'] == 'rgx'
    node['rule'] = Regexp.new "\\A#{node['.rgx']}"
  end
  if sep = node['.sep']
    optimize_node(sep)
  end
end

#parse(input, start = nil) ⇒ Object 



26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
 
# File 'lib/pegex/parser.rb', line 26

def parse input, start=nil
  @position = 0

  if not input.kind_of? Pegex::Input
    input = Pegex::Input.new {|i| i.string = input}
  end
  @input = input
  @input.open unless @input.open?
  @buffer = @input.read
  @length = @buffer.length

  fail "No 'grammar'. Can't parse" unless @grammar
  @tree = @grammar.tree ||= @grammar.make_tree

  start_rule_ref = start ||
    @tree['+toprule'] ||
    (@tree['TOP'] ? 'TOP' : nil) or
      fail "No starting rule for Pegex::Parser::parse"

  optimize_grammar(start_rule_ref)

  fail "No 'receiver'. Can't parse" unless @receiver

  # XXX does ruby have problems with circulat references?
  @receiver.parser = self

  if @receiver.respond_to? 'initial'
    @rule = start_rule_ref
    @parent = {}
    @receiver.initial
  end

  match = match_ref(start_rule_ref, {})

  @input.close

  if !match or @position < @length
    throw_error "Parse document failed for some reason"
    return
  end

  if @receiver.respond_to? 'final'
    @rule = start_rule_ref
    @parent = {}
    match = [ @receiver.final(match.first) ]
  end

  return match.first
end

#throw_error(msg) ⇒ Object

Raises:



252
253
254
255
256
 
# File 'lib/pegex/parser.rb', line 252

def throw_error msg
  @error = format_error msg
  return nil unless @throw_on_error
  raise PegexParseError, @error
end

#trace(action) ⇒ Object 



237
238
239
240
241
242
243
244
245
246
247
248
 
# File 'lib/pegex/parser.rb', line 237

def trace action
  indent = !!action.match(/^try_/)
  @indent ||= 0
  @indent -= 1 unless indent
  $stderr.print ' ' * @indent
  @indent += 1 if indent
  snippet = @buffer[@position..-1]
  snippet = snippet[0..30] + '...' if snippet.length > 30;
  snippet.gsub! /\n/, "\\n"
  $stderr.printf "%-30s", action
  $stderr.print indent ? " >#{snippet}<\n" : "\n"
end