12041343242

Committed 27 Nov 2024 12:23AM UTC coverage: 93.533% (-0.5%) from 94.066%

Build # 12041343242

Build Type

push

github

Committed by

gkellogg

Commit Message

Revise logging to use standard RDF log facility.

Run Details

6 of 17 new or added lines in 3 files covered. (35.29%)

12 existing lines in 4 files now uncovered.

2126 of 2273 relevant lines covered (93.53%)

25774.02 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

96.3

/lib/ebnf/parser.rb

require_relative 'ebnf/meta'
require 'logger'

module EBNF
  class Parser
    include EBNF::PEG::Parser
    include EBNF::Terminals

    # Abstract syntax tree from parse
    #
    # @return [Array<EBNF::Rule>]
    attr_reader :ast

    # ## Terminals
    # Define rules for Terminals, placing results on the input stack, making them available to upstream non-Terminal rules.
    #
    # Terminals are defined with a symbol matching the associated rule name, and an optional (although strongly encouraged) regular expression used to match the head of the input stream.
    #
    # The result of the terminal block is the semantic value of that terminal, which if often a string, but may be any instance which reflects the semantic interpretation of that terminal.
    #
    # The `value` parameter is the value matched by the regexp, if defined, or by the sub-terminal rules otherwise.
    #
    # The `prod` parameter is the name of the parent rule for which this terminal is matched, which may have a bearing in some circumstances, although not used in this example.
    #
    # If no block is provided, then the value which would have been passed to the block is used as the result directly.

    # Match the Left hand side of a rule or terminal
    #
    #     [11] LHS        ::= ('[' SYMBOL+ ']' ' '+)? <? SYMBOL >? ' '* '::='
    terminal(:LHS, LHS) do |value, prod|
      value.to_s.scan(/(?:\[([^\]]+)\])?\s*<?(\w+)>?\s*::=/).first
    end

    # Match `SYMBOL` terminal
    #
    #     [12] SYMBOL     ::= '<' O_SYMBOL '>' | O_SYMBOL
    #     [12a] O_SYMBOL  ::= ([a-z] | [A-Z] | [0-9] | '_' | '.')+
    terminal(:SYMBOL, SYMBOL) do |value|
      value = value[1..-2] if value.start_with?('<') && value.end_with?('>')
      value.to_sym
    end

    # Match `HEX` terminal
    #
    #     [13] HEX        ::= #x' ([a-f] | [A-F] | [0-9])+
    terminal(:HEX, HEX) do |value|
      [:hex, value]
    end

    # Terminal for `RANGE` is matched as part of a `primary` rule.
    #
    #     [14] RANGE      ::= '[' ((R_CHAR '-' R_CHAR) | (HEX '-' HEX) | R_CHAR | HEX)+ '-'? ']' - LHS
    terminal(:RANGE, RANGE) do |value|
      [:range, value[1..-2]]
    end

    # Terminal for `O_RANGE` is matched as part of a `primary` rule.
    #
    #     [15] O_RANGE    ::= '[^' ((R_CHAR '-' R_CHAR) | (HEX '-' HEX) | R_CHAR | HEX)+ '-'? ']'
    terminal(:O_RANGE, O_RANGE) do |value|
      [:range, value[1..-2]]
    end

    # Match double quote string
    #
    #     [16] STRING1    ::= '"' (CHAR - '"')* '"'
    terminal(:STRING1, STRING1) do |value|
      using ::EBNF
      value[1..-2].tap {|s| s.quote_style = :dquote}
    end

    # Match single quote string
    #
    #     [17] STRING2    ::= "'" (CHAR - "'")* "'"
    terminal(:STRING2, STRING2) do |value|
      using ::EBNF
      value[1..-2].tap {|s| s.quote_style = :squote}
    end

    # The `CHAR` and `R_CHAR` productions are not used explicitly

    # Match `POSTFIX` terminal
    #
    #     [20] POSTFIX    ::= [?*+]
    terminal(:POSTFIX, POSTFIX)

    # The `PASS` productions is not used explicitly

    # ## Non-terminal productions
    # Define productions for non-Termainals. This can include `start_production` as well as `production` to hook into rule start and end. In some cases, we need to use sub-productions as generated when turning EBNF into PEG.
    #
    # Productions are defined with a symbol matching the associated rule name.
    #
    # The result of the productions is typically the abstract syntax tree matched by the rule, so far, but could be a specific semantic value, or could be ignored with the result being returned via the `callback`.
    #
    # The `value` parameter is the result returned from child productions
    #
    # The `data` parameter other data which may be returned by child productions placing information onto their input (unused in this example).
    #
    # The `callback` parameter provides access to a callback defined in the call to `parse`).

    # Production for end of `declaration` non-terminal.
    #
    # Look for `@terminals` to change parser state to parsing terminals.
    #
    # Clears the packrat parser when called.
    #
    # `@pass` is ignored here.
    #
    #     [2] declaration ::= '@terminals' | pass
    production(:declaration, clear_packrat: true) do |value, data, callback|
      # value contains a declaration.
      # Invoke callback
      callback.call(:terminals) if value == '@terminals'
      nil
    end

    # Production for end of `rule` non-terminal.
    #
    # By setting `as_hash: true` in the `start_production`, the `value` parameter will be in the form `{LHS: "v", expression: "v"}`. Otherwise, it would be expressed using an array of hashes of the form `[{LHS: "v"}, {expression: "v"}]`.
    #
    # Clears the packrat parser when called.
    #
    # Create rule from expression value and pass to callback
    #
    #     [3] rule        ::= LHS expression
    start_production(:rule, as_hash: true)
    production(:rule, clear_packrat: true) do |value, data, callback|
      # value contains an expression.
      # Invoke callback
      id, sym = value[:LHS]
      expression = value[:expression]
      callback.call(:rule, EBNF::Rule.new(sym.to_sym, id, expression))
      nil
    end

    # Production for end of `expression` non-terminal.
    # Passes through the optimized value of the alt production as follows:
    #
    # The `value` parameter, is of the form `[{alt: "v"}]`.
    #
    #     [:alt foo] => foo
    #     [:alt foo bar] => [:alt foo bar]
    #
    #     [4] expression  ::= alt
    production(:expression) do |value|
      value.first[:alt]
    end

    # Production for end of `alt` non-terminal.
    # Passes through the optimized value of the seq production as follows:
    #
    # The `value` parameter, is of the form `{seq: "v", _alt_1: "v"}`.
    #
    #     [:seq foo] => foo
    #     [:seq foo bar] => [:seq foo bar]
    #
    # Note that this also may just pass through from `_alt_1`
    #
    #     [5] alt         ::= seq ('|' seq)*
    start_production(:alt, as_hash: true)
    production(:alt) do |value|
      if value[:_alt_1].length > 0
        [:alt, value[:seq]] + value[:_alt_1]
      else
        value[:seq]
      end
    end

    # Production for end of `_alt_1` non-terminal.
    # Used to collect the `('|' seq)*` portion of the `alt` non-terminal:
    #
    # The `value` parameter, is of the form `[{seq: ["v"]}]`.
    #
    #     [5] _alt_1         ::= ('|' seq)*
    production(:_alt_1) do |value|
      value.map {|a1| a1.last[:seq]}.compact # Get rid of '|'
    end

    # Production for end of `seq` non-terminal.
    # Passes through the optimized value of the `diff` production as follows:
    #
    # The `value` parameter, is an array of values, which cannot be empty.
    #
    #     [:diff foo] => foo
    #     [:diff foo bar] => [:diff foo bar]
    #
    # Note that this also may just pass through from `_seq_1`
    #
    #     [6] seq         ::= diff+
    production(:seq) do |value|
      value.length == 1 ? value.first : ([:seq] + value)
    end

    # `Diff` production returns concatenated postfix values
    #
    # The `value` parameter, is of the form `{postfix: "v", _diff_1: "v"}`.
    #
    #     [7] diff        ::= postfix ('-' postfix)?
    start_production(:diff, as_hash: true)
    production(:diff) do |value|
      if value[:_diff_1]
        [:diff, value[:postfix], value[:_diff_1]]
      else
        value[:postfix]
      end
    end

    production(:_diff_1) do |value|
      value.last[:postfix] if value
    end

    # Production for end of `postfix` non-terminal.
    # Either returns the `primary` production value, or as modified by the `postfix`.
    #
    # The `value` parameter, is of the form `{primary: "v", _postfix_1: "v"}`.
    #
    #     [:primary] => [:primary]
    #     [:primary, '*'] => [:star, :primary]
    #     [:primary, '+'] => [:plus, :primary]
    #     [:primary, '?'] => [:opt, :primary]
    #
    #     [8] postfix     ::= primary POSTFIX?
    start_production(:postfix, as_hash: true)
    production(:postfix) do |value|
      # Push result onto input stack, as the `diff` production can have some number of `postfix` values that are applied recursively
      case value[:_postfix_1]
      when "*" then [:star, value[:primary]]
      when "+" then [:plus, value[:primary]]
      when "?" then [:opt, value[:primary]]
      else value[:primary]
      end
    end

    # Production for end of `primary` non-terminal.
    # Places `:primary` on the stack
    #
    # The `value` parameter, is either a string (for a terminal) or an array of the form `['(': '(', expression: "v", ')', ')']`.
    #
    # This may either be a terminal, or the result of an `expression`.
    #
    #     [9] primary     ::= HEX
    #                     |   SYMBOL
    #                     |   RANGE
    #                     |   O_RANGE
    #                     |   STRING1
    #                     |   STRING2
    #                     |   '(' expression ')'
    production(:primary) do |value|
      Array(value).length > 2 ? value[1][:expression] : value
    end

    # Production for end of pass non-terminal.
    #
    #     [10] pass       ::= '@pass' expression
    production(:pass) do |value, data, callback|
      # Invoke callback
      callback.call(:pass, value.last[:expression])
    end

    # ## Parser invocation.
    # On start, yield ourselves if a block is given, otherwise, return this parser instance
    #
    # @param  [#read, #to_s]          input
    # @param  [Hash{Symbol => Object}] options
    # @option options [Boolean] :level
    #   Trace level. 0(debug), 1(info), 2(warn), 3(error).
    # @return [EBNFParser]
    def initialize(input, **options, &block)
      # If the `level` option is set, instantiate a logger for collecting trace information.
      if options.key?(:level)
        options[:logger] ||= Logger.new(STDERR).
          tap {|x| x.level = options[:level]}.
          tap {|x| x.formatter = lambda {|severity, datetime, progname, msg| "#{severity} #{msg}\n"}}
      end

      # Read input, if necessary, which will be used in a Scanner.
      @input = input.respond_to?(:read) ? input.read : input.to_s

      parsing_terminals = false
      @ast = []
      parse(@input, :ebnf, EBNFMeta::RULES,
                           # Use an optimized Regexp for whitespace
                           whitespace: EBNF::Terminals::PASS,
                           **options
      ) do |context, *data|
        rule = case context
        when :terminals
          # After parsing `@terminals`
          # This changes the state of the parser to treat subsequent rules as terminals.
          parsing_terminals = true
          rule = EBNF::Rule.new(nil, nil, data.first, kind: :terminals)
        when :pass
          # After parsing `@pass`
          # This defines a specific rule for whitespace.
          rule = EBNF::Rule.new(nil, nil, data.first, kind: :pass)
        when :rule
          # A rule which has already been turned into a `Rule` object.
          rule = data.first
          rule.kind = :terminal if parsing_terminals
          rule
        end
        @ast << rule if rule
      end
    rescue EBNF::PEG::Parser::Error => e
      raise SyntaxError, e.message
    end
  end
end

1	require_relative 'ebnf/meta'	2✔
2	require 'logger'	2✔
3
4	module EBNF	2✔
5	class Parser	2✔
6	include EBNF::PEG::Parser	2✔
7	include EBNF::Terminals	2✔
8
9	# Abstract syntax tree from parse
10	#
11	# @return [Array<EBNF::Rule>]
12	attr_reader :ast	2✔
13
14	# ## Terminals
15	# Define rules for Terminals, placing results on the input stack, making them available to upstream non-Terminal rules.
16	#
17	# Terminals are defined with a symbol matching the associated rule name, and an optional (although strongly encouraged) regular expression used to match the head of the input stream.
18	#
19	# The result of the terminal block is the semantic value of that terminal, which if often a string, but may be any instance which reflects the semantic interpretation of that terminal.
20	#
21	# The `value` parameter is the value matched by the regexp, if defined, or by the sub-terminal rules otherwise.
22	#
23	# The `prod` parameter is the name of the parent rule for which this terminal is matched, which may have a bearing in some circumstances, although not used in this example.
24	#
25	# If no block is provided, then the value which would have been passed to the block is used as the result directly.
26
27	# Match the Left hand side of a rule or terminal
28	#
29	# [11] LHS ::= ('[' SYMBOL+ ']' ' '+)? <? SYMBOL >? ' '* '::='
30	terminal(:LHS, LHS) do \|value, prod\|	2✔
31	value.to_s.scan(/(?:\[([^\]]+)\])?\s<?(\w+)>?\s::=/).first	5,202✔
32	end
33
34	# Match `SYMBOL` terminal
35	#
36	# [12] SYMBOL ::= '<' O_SYMBOL '>' \| O_SYMBOL
37	# [12a] O_SYMBOL ::= ([a-z] \| [A-Z] \| [0-9] \| '_' \| '.')+
38	terminal(:SYMBOL, SYMBOL) do \|value\|	2✔
39	value = value[1..-2] if value.start_with?('<') && value.end_with?('>')	10,536✔
40	value.to_sym	10,536✔
41	end
42
43	# Match `HEX` terminal
44	#
45	# [13] HEX ::= #x' ([a-f] \| [A-F] \| [0-9])+
46	terminal(:HEX, HEX) do \|value\|	2✔
47	[:hex, value]	316✔
48	end
49
50	# Terminal for `RANGE` is matched as part of a `primary` rule.
51	#
52	# [14] RANGE ::= '[' ((R_CHAR '-' R_CHAR) \| (HEX '-' HEX) \| R_CHAR \| HEX)+ '-'? ']' - LHS
53	terminal(:RANGE, RANGE) do \|value\|	2✔
54	[:range, value[1..-2]]	2,692✔
55	end
56
57	# Terminal for `O_RANGE` is matched as part of a `primary` rule.
58	#
59	# [15] O_RANGE ::= '[^' ((R_CHAR '-' R_CHAR) \| (HEX '-' HEX) \| R_CHAR \| HEX)+ '-'? ']'
60	terminal(:O_RANGE, O_RANGE) do \|value\|	2✔
61	[:range, value[1..-2]]	488✔
62	end
63
64	# Match double quote string
65	#
66	# [16] STRING1 ::= '"' (CHAR - '"')* '"'
67	terminal(:STRING1, STRING1) do \|value\|	2✔
68	using ::EBNF	862✔
69	value[1..-2].tap {\|s\| s.quote_style = :dquote}	1,724✔
70	end
71
72	# Match single quote string
73	#
74	# [17] STRING2 ::= "'" (CHAR - "'")* "'"
75	terminal(:STRING2, STRING2) do \|value\|	2✔
76	using ::EBNF	7,778✔
77	value[1..-2].tap {\|s\| s.quote_style = :squote}	15,556✔
78	end
79
80	# The `CHAR` and `R_CHAR` productions are not used explicitly
81
82	# Match `POSTFIX` terminal
83	#
84	# [20] POSTFIX ::= [?*+]
85	terminal(:POSTFIX, POSTFIX)	2✔
86
87	# The `PASS` productions is not used explicitly
88
89	# ## Non-terminal productions
90	# Define productions for non-Termainals. This can include `start_production` as well as `production` to hook into rule start and end. In some cases, we need to use sub-productions as generated when turning EBNF into PEG.
91	#
92	# Productions are defined with a symbol matching the associated rule name.
93	#
94	# The result of the productions is typically the abstract syntax tree matched by the rule, so far, but could be a specific semantic value, or could be ignored with the result being returned via the `callback`.
95	#
96	# The `value` parameter is the result returned from child productions
97	#
98	# The `data` parameter other data which may be returned by child productions placing information onto their input (unused in this example).
99	#
100	# The `callback` parameter provides access to a callback defined in the call to `parse`).
101
102	# Production for end of `declaration` non-terminal.
103	#
104	# Look for `@terminals` to change parser state to parsing terminals.
105	#
106	# Clears the packrat parser when called.
107	#
108	# `@pass` is ignored here.
109	#
110	# [2] declaration ::= '@terminals' \| pass
111	production(:declaration, clear_packrat: true) do \|value, data, callback\|	2✔
112	# value contains a declaration.
113	# Invoke callback
114	callback.call(:terminals) if value == '@terminals'	166✔
115	nil	166✔
116	end
117
118	# Production for end of `rule` non-terminal.
119	#
120	# By setting `as_hash: true` in the `start_production`, the `value` parameter will be in the form `{LHS: "v", expression: "v"}`. Otherwise, it would be expressed using an array of hashes of the form `[{LHS: "v"}, {expression: "v"}]`.
121	#
122	# Clears the packrat parser when called.
123	#
124	# Create rule from expression value and pass to callback
125	#
126	# [3] rule ::= LHS expression
127	start_production(:rule, as_hash: true)	2✔
128	production(:rule, clear_packrat: true) do \|value, data, callback\|	2✔
129	# value contains an expression.
130	# Invoke callback
131	id, sym = value[:LHS]	5,192✔
132	expression = value[:expression]	5,192✔
133	callback.call(:rule, EBNF::Rule.new(sym.to_sym, id, expression))	5,192✔
134	nil	5,192✔
135	end
136
137	# Production for end of `expression` non-terminal.
138	# Passes through the optimized value of the alt production as follows:
139	#
140	# The `value` parameter, is of the form `[{alt: "v"}]`.
141	#
142	# [:alt foo] => foo
143	# [:alt foo bar] => [:alt foo bar]
144	#
145	# [4] expression ::= alt
146	production(:expression) do \|value\|	2✔
147	value.first[:alt]	8,638✔
148	end
149
150	# Production for end of `alt` non-terminal.
151	# Passes through the optimized value of the seq production as follows:
152	#
153	# The `value` parameter, is of the form `{seq: "v", _alt_1: "v"}`.
154	#
155	# [:seq foo] => foo
156	# [:seq foo bar] => [:seq foo bar]
157	#
158	# Note that this also may just pass through from `_alt_1`
159	#
160	# [5] alt ::= seq ('\|' seq)*
161	start_production(:alt, as_hash: true)	2✔
162	production(:alt) do \|value\|	2✔
163	if value[:_alt_1].length > 0	8,638✔
164	[:alt, value[:seq]] + value[:_alt_1]	3,120✔
165	else
166	value[:seq]	5,518✔
167	end
168	end
169
170	# Production for end of `_alt_1` non-terminal.
171	# Used to collect the `('\|' seq)*` portion of the `alt` non-terminal:
172	#
173	# The `value` parameter, is of the form `[{seq: ["v"]}]`.
174	#
175	# [5] _alt_1 ::= ('\|' seq)*
176	production(:_alt_1) do \|value\|	2✔
177	value.map {\|a1\| a1.last[:seq]}.compact # Get rid of '\|'	16,532✔
178	end
179
180	# Production for end of `seq` non-terminal.
181	# Passes through the optimized value of the `diff` production as follows:
182	#
183	# The `value` parameter, is an array of values, which cannot be empty.
184	#
185	# [:diff foo] => foo
186	# [:diff foo bar] => [:diff foo bar]
187	#
188	# Note that this also may just pass through from `_seq_1`
189	#
190	# [6] seq ::= diff+
191	production(:seq) do \|value\|	2✔
192	value.length == 1 ? value.first : ([:seq] + value)	16,532✔
193	end
194
195	# `Diff` production returns concatenated postfix values
196	#
197	# The `value` parameter, is of the form `{postfix: "v", _diff_1: "v"}`.
198	#
199	# [7] diff ::= postfix ('-' postfix)?
200	start_production(:diff, as_hash: true)	2✔
201	production(:diff) do \|value\|	2✔
202	if value[:_diff_1]	25,810✔
203	[:diff, value[:postfix], value[:_diff_1]]	260✔
204	else
205	value[:postfix]	25,550✔
206	end
207	end
208
209	production(:_diff_1) do \|value\|	2✔
210	value.last[:postfix] if value	25,810✔
211	end
212
213	# Production for end of `postfix` non-terminal.
214	# Either returns the `primary` production value, or as modified by the `postfix`.
215	#
216	# The `value` parameter, is of the form `{primary: "v", _postfix_1: "v"}`.
217	#
218	# [:primary] => [:primary]
219	# [:primary, '*'] => [:star, :primary]
220	# [:primary, '+'] => [:plus, :primary]
221	# [:primary, '?'] => [:opt, :primary]
222	#
223	# [8] postfix ::= primary POSTFIX?
224	start_production(:postfix, as_hash: true)	2✔
225	production(:postfix) do \|value\|	2✔
226	# Push result onto input stack, as the `diff` production can have some number of `postfix` values that are applied recursively
227	case value[:_postfix_1]	26,070✔
228	when "*" then [:star, value[:primary]]	1,596✔
229	when "+" then [:plus, value[:primary]]	852✔
230	when "?" then [:opt, value[:primary]]	1,472✔
231	else value[:primary]	22,150✔
232	end
233	end
234
235	# Production for end of `primary` non-terminal.
236	# Places `:primary` on the stack
237	#
238	# The `value` parameter, is either a string (for a terminal) or an array of the form `['(': '(', expression: "v", ')', ')']`.
239	#
240	# This may either be a terminal, or the result of an `expression`.
241	#
242	# [9] primary ::= HEX
243	# \| SYMBOL
244	# \| RANGE
245	# \| O_RANGE
246	# \| STRING1
247	# \| STRING2
248	# \| '(' expression ')'
249	production(:primary) do \|value\|	2✔
250	Array(value).length > 2 ? value[1][:expression] : value	26,070✔
251	end
252
253	# Production for end of pass non-terminal.
254	#
255	# [10] pass ::= '@pass' expression
256	production(:pass) do \|value, data, callback\|	2✔
257	# Invoke callback
258	callback.call(:pass, value.last[:expression])	48✔
259	end
260
261	# ## Parser invocation.
262	# On start, yield ourselves if a block is given, otherwise, return this parser instance
263	#
264	# @param [#read, #to_s] input
265	# @param [Hash{Symbol => Object}] options
266	# @option options [Boolean] :level
267	# Trace level. 0(debug), 1(info), 2(warn), 3(error).
268	# @return [EBNFParser]
269	def initialize(input, **options, &block)	2✔
270	# If the `level` option is set, instantiate a logger for collecting trace information.
271	if options.key?(:level)	354✔
272	options[:logger] \|\|= Logger.new(STDERR).	×
UNCOV 273	tap {\|x\| x.level = options[:level]}.	×
UNCOV 274	tap {\|x\| x.formatter = lambda {\|severity, datetime, progname, msg\| "#{severity} #{msg}\n"}}	×
275	end
276
277	# Read input, if necessary, which will be used in a Scanner.
278	@input = input.respond_to?(:read) ? input.read : input.to_s	354✔
279
280	parsing_terminals = false	354✔
281	@ast = []	354✔
282	parse(@input, :ebnf, EBNFMeta::RULES,	354✔
283	# Use an optimized Regexp for whitespace
284	whitespace: EBNF::Terminals::PASS,
285	**options
286	) do \|context, *data\|
287	rule = case context	5,358✔
288	when :terminals
289	# After parsing `@terminals`
290	# This changes the state of the parser to treat subsequent rules as terminals.
291	parsing_terminals = true	118✔
292	rule = EBNF::Rule.new(nil, nil, data.first, kind: :terminals)	118✔
293	when :pass
294	# After parsing `@pass`
295	# This defines a specific rule for whitespace.
296	rule = EBNF::Rule.new(nil, nil, data.first, kind: :pass)	48✔
297	when :rule
298	# A rule which has already been turned into a `Rule` object.
299	rule = data.first	5,192✔
300	rule.kind = :terminal if parsing_terminals	5,192✔
301	rule	5,192✔
302	end
303	@ast << rule if rule	5,358✔
304	end
305	rescue EBNF::PEG::Parser::Error => e
306	raise SyntaxError, e.message	16✔
307	end
308	end
309	end

dryruby / ebnf / 12041343242

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous