lrama 0.6.4 → 0.6.6
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/NEWS.md +40 -0
- data/lib/lrama/command.rb +5 -0
- data/lib/lrama/grammar/parameterizing_rule/resolver.rb +15 -6
- data/lib/lrama/grammar/parameterizing_rule/rule.rb +3 -2
- data/lib/lrama/grammar/rule.rb +5 -1
- data/lib/lrama/grammar/rule_builder.rb +69 -18
- data/lib/lrama/lexer/grammar_file.rb +11 -1
- data/lib/lrama/lexer/token/user_code.rb +2 -0
- data/lib/lrama/lexer.rb +1 -0
- data/lib/lrama/option_parser.rb +3 -2
- data/lib/lrama/parser.rb +539 -518
- data/lib/lrama/state.rb +3 -25
- data/lib/lrama/version.rb +1 -1
- data/parser.y +9 -1
- data/sample/calc.y +7 -4
- data/sig/lrama/grammar/parameterizing_rule/resolver.rbs +5 -3
- data/sig/lrama/grammar/parameterizing_rule/rule.rbs +2 -1
- data/sig/lrama/grammar/rule_builder.rbs +5 -1
- data/sig/lrama/lexer/grammar_file.rbs +2 -0
- data/sig/lrama/lexer/token/user_code.rbs +1 -0
- metadata +2 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 5a24c7fd3c35d08a7e602fe74d4be9987f0415a1dc621ceb02ec4a722b1e16a0
|
4
|
+
data.tar.gz: fa3e0819afd1a454c17189c05f4c02260e35a252b09fdc828238f01b7ad3c58e
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 7a5268674e4e8630a8efad69e82af94202f66da34530fdb87b7bd8d0e4cee93cf3881e1509d441fade31d1886f39e48b43c9f25178511983002a30fe19047266
|
7
|
+
data.tar.gz: 54122863a527a29a6279b9ad4c2732a378a138743d5d91e9c611ba9d66c7929f14385ee228b1dde7e20f2a4efc60fe1d9d5046fcb753c828daf64e4d9d465350
|
data/NEWS.md
CHANGED
@@ -1,5 +1,45 @@
|
|
1
1
|
# NEWS for Lrama
|
2
2
|
|
3
|
+
## Lrama 0.6.5 (2024-03-25)
|
4
|
+
|
5
|
+
### Typed Midrule Actions
|
6
|
+
|
7
|
+
User can specify the type of mid rule action by tag (`<bar>`) instead of specifying it with in an action.
|
8
|
+
|
9
|
+
```
|
10
|
+
primary: k_case expr_value terms?
|
11
|
+
{
|
12
|
+
$<val>$ = p->case_labels;
|
13
|
+
p->case_labels = Qnil;
|
14
|
+
}
|
15
|
+
case_body
|
16
|
+
k_end
|
17
|
+
{
|
18
|
+
...
|
19
|
+
}
|
20
|
+
```
|
21
|
+
|
22
|
+
can be written as
|
23
|
+
|
24
|
+
```
|
25
|
+
primary: k_case expr_value terms?
|
26
|
+
{
|
27
|
+
$$ = p->case_labels;
|
28
|
+
p->case_labels = Qnil;
|
29
|
+
}<val>
|
30
|
+
case_body
|
31
|
+
k_end
|
32
|
+
{
|
33
|
+
...
|
34
|
+
}
|
35
|
+
```
|
36
|
+
|
37
|
+
`%destructor` for midrule action is invoked only when tag is specified by Typed Midrule Actions.
|
38
|
+
|
39
|
+
Difference from Bison's Typed Midrule Actions is that tag is postposed in Lrama however it's preposed in Bison.
|
40
|
+
|
41
|
+
Bison supports this feature from 3.1.
|
42
|
+
|
3
43
|
## Lrama 0.6.4 (2024-03-22)
|
4
44
|
|
5
45
|
### Parameterizing rules (preceded, terminated, delimited)
|
data/lib/lrama/command.rb
CHANGED
@@ -47,6 +47,11 @@ module Lrama
|
|
47
47
|
puts grammar.rules
|
48
48
|
end
|
49
49
|
|
50
|
+
if options.trace_opts && options.trace_opts[:actions]
|
51
|
+
puts "Grammar rules with actions:"
|
52
|
+
grammar.rules.each { |rule| puts rule.with_actions }
|
53
|
+
end
|
54
|
+
|
50
55
|
File.open(options.outfile, "w+") do |f|
|
51
56
|
Lrama::Output.new(
|
52
57
|
out: f,
|
@@ -13,8 +13,12 @@ module Lrama
|
|
13
13
|
@rules << rule
|
14
14
|
end
|
15
15
|
|
16
|
-
def
|
17
|
-
select_rules(token).last
|
16
|
+
def find_rule(token)
|
17
|
+
select_rules(@rules, token).last
|
18
|
+
end
|
19
|
+
|
20
|
+
def find_inline(token)
|
21
|
+
@rules.select { |rule| rule.name == token.s_value && rule.is_inline }.last
|
18
22
|
end
|
19
23
|
|
20
24
|
def created_lhs(lhs_s_value)
|
@@ -23,8 +27,9 @@ module Lrama
|
|
23
27
|
|
24
28
|
private
|
25
29
|
|
26
|
-
def select_rules(token)
|
27
|
-
rules =
|
30
|
+
def select_rules(rules, token)
|
31
|
+
rules = select_not_inline_rules(rules)
|
32
|
+
rules = select_rules_by_name(rules, token.rule_name)
|
28
33
|
rules = rules.select { |rule| rule.required_parameters_count == token.args_count }
|
29
34
|
if rules.empty?
|
30
35
|
raise "Invalid number of arguments. `#{token.rule_name}`"
|
@@ -33,8 +38,12 @@ module Lrama
|
|
33
38
|
end
|
34
39
|
end
|
35
40
|
|
36
|
-
def
|
37
|
-
rules
|
41
|
+
def select_not_inline_rules(rules)
|
42
|
+
rules.select { |rule| !rule.is_inline }
|
43
|
+
end
|
44
|
+
|
45
|
+
def select_rules_by_name(rules, rule_name)
|
46
|
+
rules = rules.select { |rule| rule.name == rule_name }
|
38
47
|
if rules.empty?
|
39
48
|
raise "Parameterizing rule does not exist. `#{rule_name}`"
|
40
49
|
else
|
@@ -2,12 +2,13 @@ module Lrama
|
|
2
2
|
class Grammar
|
3
3
|
class ParameterizingRule
|
4
4
|
class Rule
|
5
|
-
attr_reader :name, :parameters, :rhs_list, :required_parameters_count
|
5
|
+
attr_reader :name, :parameters, :rhs_list, :required_parameters_count, :is_inline
|
6
6
|
|
7
|
-
def initialize(name, parameters, rhs_list)
|
7
|
+
def initialize(name, parameters, rhs_list, is_inline: false)
|
8
8
|
@name = name
|
9
9
|
@parameters = parameters
|
10
10
|
@rhs_list = rhs_list
|
11
|
+
@is_inline = is_inline
|
11
12
|
@required_parameters_count = parameters.count
|
12
13
|
end
|
13
14
|
end
|
data/lib/lrama/grammar/rule.rb
CHANGED
@@ -19,7 +19,7 @@ module Lrama
|
|
19
19
|
# TODO: Change this to display_name
|
20
20
|
def to_s
|
21
21
|
l = lhs.id.s_value
|
22
|
-
r = empty_rule? ? "ε" : rhs.map {|r| r.id.s_value }.join("
|
22
|
+
r = empty_rule? ? "ε" : rhs.map {|r| r.id.s_value }.join(" ")
|
23
23
|
|
24
24
|
"#{l} -> #{r}"
|
25
25
|
end
|
@@ -32,6 +32,10 @@ module Lrama
|
|
32
32
|
"#{l}: #{r}"
|
33
33
|
end
|
34
34
|
|
35
|
+
def with_actions
|
36
|
+
"#{to_s} {#{token_code&.s_value}}"
|
37
|
+
end
|
38
|
+
|
35
39
|
# opt_nl: ε <-- empty_rule
|
36
40
|
# | '\n' <-- not empty_rule
|
37
41
|
def empty_rule?
|
@@ -16,8 +16,13 @@ module Lrama
|
|
16
16
|
@user_code = nil
|
17
17
|
@precedence_sym = nil
|
18
18
|
@line = nil
|
19
|
+
@rules = []
|
19
20
|
@rule_builders_for_parameterizing_rules = []
|
20
21
|
@rule_builders_for_derived_rules = []
|
22
|
+
@rule_builders_for_inline_rules = []
|
23
|
+
@parameterizing_rules = []
|
24
|
+
@inline_rules = []
|
25
|
+
@midrule_action_rules = []
|
21
26
|
end
|
22
27
|
|
23
28
|
def add_rhs(rhs)
|
@@ -52,12 +57,16 @@ module Lrama
|
|
52
57
|
|
53
58
|
def setup_rules(parameterizing_rule_resolver)
|
54
59
|
preprocess_references unless @skip_preprocess_references
|
55
|
-
|
60
|
+
if rhs.any? { |token| parameterizing_rule_resolver.find_inline(token) }
|
61
|
+
resolve_inline(parameterizing_rule_resolver)
|
62
|
+
else
|
63
|
+
process_rhs(parameterizing_rule_resolver)
|
64
|
+
end
|
56
65
|
build_rules
|
57
66
|
end
|
58
67
|
|
59
68
|
def rules
|
60
|
-
@parameterizing_rules + @midrule_action_rules + @rules
|
69
|
+
@parameterizing_rules + @inline_rules + @midrule_action_rules + @rules
|
61
70
|
end
|
62
71
|
|
63
72
|
private
|
@@ -73,19 +82,25 @@ module Lrama
|
|
73
82
|
def build_rules
|
74
83
|
tokens = @replaced_rhs
|
75
84
|
|
76
|
-
|
77
|
-
|
78
|
-
|
79
|
-
|
80
|
-
|
81
|
-
|
82
|
-
rule_builder
|
83
|
-
|
84
|
-
|
85
|
-
rule_builder
|
86
|
-
|
87
|
-
|
88
|
-
|
85
|
+
if tokens
|
86
|
+
rule = Rule.new(
|
87
|
+
id: @rule_counter.increment, _lhs: lhs, _rhs: tokens, lhs_tag: lhs_tag, token_code: user_code,
|
88
|
+
position_in_original_rule_rhs: @position_in_original_rule_rhs, precedence_sym: precedence_sym, lineno: line
|
89
|
+
)
|
90
|
+
@rules = [rule]
|
91
|
+
@parameterizing_rules = @rule_builders_for_parameterizing_rules.map do |rule_builder|
|
92
|
+
rule_builder.rules
|
93
|
+
end.flatten
|
94
|
+
@midrule_action_rules = @rule_builders_for_derived_rules.map do |rule_builder|
|
95
|
+
rule_builder.rules
|
96
|
+
end.flatten
|
97
|
+
@midrule_action_rules.each do |r|
|
98
|
+
r.original_rule = rule
|
99
|
+
end
|
100
|
+
else
|
101
|
+
@inline_rules = @rule_builders_for_inline_rules.map do |rule_builder|
|
102
|
+
rule_builder.rules
|
103
|
+
end.flatten
|
89
104
|
end
|
90
105
|
end
|
91
106
|
|
@@ -103,7 +118,7 @@ module Lrama
|
|
103
118
|
when Lrama::Lexer::Token::Ident
|
104
119
|
@replaced_rhs << token
|
105
120
|
when Lrama::Lexer::Token::InstantiateRule
|
106
|
-
parameterizing_rule = parameterizing_rule_resolver.
|
121
|
+
parameterizing_rule = parameterizing_rule_resolver.find_rule(token)
|
107
122
|
raise "Unexpected token. #{token}" unless parameterizing_rule
|
108
123
|
|
109
124
|
bindings = Binding.new(parameterizing_rule, token.args)
|
@@ -115,7 +130,7 @@ module Lrama
|
|
115
130
|
@replaced_rhs << lhs_token
|
116
131
|
parameterizing_rule_resolver.created_lhs_list << lhs_token
|
117
132
|
parameterizing_rule.rhs_list.each do |r|
|
118
|
-
rule_builder = RuleBuilder.new(@rule_counter, @midrule_action_counter, lhs_tag: token.lhs_tag
|
133
|
+
rule_builder = RuleBuilder.new(@rule_counter, @midrule_action_counter, lhs_tag: token.lhs_tag)
|
119
134
|
rule_builder.lhs = lhs_token
|
120
135
|
r.symbols.each { |sym| rule_builder.add_rhs(bindings.resolve_symbol(sym)) }
|
121
136
|
rule_builder.line = line
|
@@ -128,10 +143,11 @@ module Lrama
|
|
128
143
|
end
|
129
144
|
when Lrama::Lexer::Token::UserCode
|
130
145
|
prefix = token.referred ? "@" : "$@"
|
146
|
+
tag = token.tag || lhs_tag
|
131
147
|
new_token = Lrama::Lexer::Token::Ident.new(s_value: prefix + @midrule_action_counter.increment.to_s)
|
132
148
|
@replaced_rhs << new_token
|
133
149
|
|
134
|
-
rule_builder = RuleBuilder.new(@rule_counter, @midrule_action_counter, i, lhs_tag:
|
150
|
+
rule_builder = RuleBuilder.new(@rule_counter, @midrule_action_counter, i, lhs_tag: tag, skip_preprocess_references: true)
|
135
151
|
rule_builder.lhs = new_token
|
136
152
|
rule_builder.user_code = token
|
137
153
|
rule_builder.complete_input
|
@@ -156,6 +172,41 @@ module Lrama
|
|
156
172
|
"#{token.rule_name}_#{s_values.join('_')}"
|
157
173
|
end
|
158
174
|
|
175
|
+
def resolve_inline(parameterizing_rule_resolver)
|
176
|
+
rhs.each_with_index do |token, i|
|
177
|
+
if inline_rule = parameterizing_rule_resolver.find_inline(token)
|
178
|
+
inline_rule.rhs_list.each_with_index do |inline_rhs|
|
179
|
+
rule_builder = RuleBuilder.new(@rule_counter, @midrule_action_counter, lhs_tag: lhs_tag, skip_preprocess_references: true)
|
180
|
+
resolve_inline_rhs(rule_builder, inline_rhs, i)
|
181
|
+
rule_builder.lhs = lhs
|
182
|
+
rule_builder.line = line
|
183
|
+
rule_builder.user_code = replace_inline_user_code(inline_rhs, i)
|
184
|
+
rule_builder.complete_input
|
185
|
+
rule_builder.setup_rules(parameterizing_rule_resolver)
|
186
|
+
@rule_builders_for_inline_rules << rule_builder
|
187
|
+
end
|
188
|
+
end
|
189
|
+
end
|
190
|
+
end
|
191
|
+
|
192
|
+
def resolve_inline_rhs(rule_builder, inline_rhs, index)
|
193
|
+
rhs.each_with_index do |token, i|
|
194
|
+
if index == i
|
195
|
+
inline_rhs.symbols.each { |sym| rule_builder.add_rhs(sym) }
|
196
|
+
else
|
197
|
+
rule_builder.add_rhs(token)
|
198
|
+
end
|
199
|
+
end
|
200
|
+
end
|
201
|
+
|
202
|
+
def replace_inline_user_code(inline_rhs, index)
|
203
|
+
return user_code if inline_rhs.user_code.nil?
|
204
|
+
return user_code if user_code.nil?
|
205
|
+
|
206
|
+
code = user_code.s_value.gsub(/\$#{index + 1}/, inline_rhs.user_code.s_value)
|
207
|
+
Lrama::Lexer::Token::UserCode.new(s_value: code, location: user_code.location)
|
208
|
+
end
|
209
|
+
|
159
210
|
def numberize_references
|
160
211
|
# Bison n'th component is 1-origin
|
161
212
|
(rhs + [user_code]).compact.each.with_index(1) do |token, i|
|
@@ -1,11 +1,21 @@
|
|
1
1
|
module Lrama
|
2
2
|
class Lexer
|
3
3
|
class GrammarFile
|
4
|
+
class Text < String
|
5
|
+
def inspect
|
6
|
+
length <= 50 ? super : "#{self[0..47]}...".inspect
|
7
|
+
end
|
8
|
+
end
|
9
|
+
|
4
10
|
attr_reader :path, :text
|
5
11
|
|
6
12
|
def initialize(path, text)
|
7
13
|
@path = path
|
8
|
-
@text = text.freeze
|
14
|
+
@text = Text.new(text).freeze
|
15
|
+
end
|
16
|
+
|
17
|
+
def inspect
|
18
|
+
"<#{self.class}: @path=#{path}, @text=#{text.inspect}>"
|
9
19
|
end
|
10
20
|
|
11
21
|
def ==(other)
|
data/lib/lrama/lexer.rb
CHANGED
data/lib/lrama/option_parser.rb
CHANGED
@@ -119,8 +119,9 @@ module Lrama
|
|
119
119
|
|
120
120
|
VALID_TRACES = %w[
|
121
121
|
none locations scan parse automaton bitsets
|
122
|
-
closure grammar rules resource
|
123
|
-
m4-early m4 skeleton time
|
122
|
+
closure grammar rules actions resource
|
123
|
+
sets muscles tools m4-early m4 skeleton time
|
124
|
+
ielr cex all
|
124
125
|
]
|
125
126
|
|
126
127
|
def validate_trace(trace)
|