lrama 0.5.5 → 0.5.7
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.codespellignore +0 -0
- data/.github/workflows/codespell.yaml +16 -0
- data/.github/workflows/test.yaml +23 -7
- data/.gitignore +1 -0
- data/Gemfile +1 -0
- data/README.md +69 -3
- data/Rakefile +12 -0
- data/Steepfile +3 -0
- data/exe/lrama +1 -1
- data/lib/lrama/command.rb +16 -137
- data/lib/lrama/context.rb +4 -4
- data/lib/lrama/digraph.rb +1 -2
- data/lib/lrama/grammar/union.rb +2 -2
- data/lib/lrama/grammar.rb +187 -1
- data/lib/lrama/lexer/token.rb +7 -1
- data/lib/lrama/lexer.rb +131 -289
- data/lib/lrama/option_parser.rb +128 -0
- data/lib/lrama/options.rb +23 -0
- data/lib/lrama/output.rb +27 -15
- data/lib/lrama/parser.rb +1759 -255
- data/lib/lrama/version.rb +1 -1
- data/lib/lrama.rb +2 -0
- data/parser.y +416 -0
- data/rbs_collection.lock.yaml +1 -1
- data/sample/calc.y +0 -2
- data/sample/parse.y +0 -3
- data/sig/lrama/digraph.rbs +23 -0
- data/sig/lrama/lexer/token/type.rbs +17 -0
- data/template/bison/_yacc.h +71 -0
- data/template/bison/yacc.c +6 -71
- data/template/bison/yacc.h +1 -73
- metadata +11 -4
- data/lib/lrama/parser/token_scanner.rb +0 -56
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: c786e8955350ca51e6226aacd493073f7c31714ef7d5911dfb16259aa9415513
|
4
|
+
data.tar.gz: d167b1a2df9dcbc8f31912e8842388aa7d6a788a8d3eeff0278c6d3ed5ecbac6
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: c89fe932ef32b5f441b87df33c2431c8631a069c293306d0a39c858e462731e11d753117ab4e7c48b4e4e41015405328a8d8f5590fbce3544e67babdaeddd8a6
|
7
|
+
data.tar.gz: 619d1affd9f09c5c1b7748f712005b60073638be709e8bae986e39801173cf9de32555cf05837dbb7450360989989f5277e76a496eec702ace75c3b4af944471
|
data/.codespellignore
ADDED
File without changes
|
@@ -0,0 +1,16 @@
|
|
1
|
+
name: CodeSpell
|
2
|
+
on:
|
3
|
+
- pull_request
|
4
|
+
jobs:
|
5
|
+
codespell:
|
6
|
+
name: CodeSpell
|
7
|
+
runs-on: ubuntu-latest
|
8
|
+
steps:
|
9
|
+
- uses: actions/checkout@v4
|
10
|
+
- name: CodeSpell
|
11
|
+
uses: codespell-project/actions-codespell@master
|
12
|
+
with:
|
13
|
+
check_filenames: true
|
14
|
+
check_hidden: true
|
15
|
+
ignore_words_file: .codespellignore
|
16
|
+
exclude_file: lib/lrama/parser.rb
|
data/.github/workflows/test.yaml
CHANGED
@@ -15,7 +15,7 @@ jobs:
|
|
15
15
|
matrix:
|
16
16
|
ruby: ['head', '3.2', '3.1', '3.0', '2.7', '2.6', '2.5']
|
17
17
|
steps:
|
18
|
-
- uses: actions/checkout@
|
18
|
+
- uses: actions/checkout@v4
|
19
19
|
- uses: ruby/setup-ruby@v1
|
20
20
|
with:
|
21
21
|
ruby-version: ${{ matrix.ruby }}
|
@@ -29,7 +29,7 @@ jobs:
|
|
29
29
|
matrix:
|
30
30
|
ruby: ['head']
|
31
31
|
steps:
|
32
|
-
- uses: actions/checkout@
|
32
|
+
- uses: actions/checkout@v4
|
33
33
|
- uses: ruby/setup-ruby@v1
|
34
34
|
with:
|
35
35
|
ruby-version: ${{ matrix.ruby }}
|
@@ -38,13 +38,27 @@ jobs:
|
|
38
38
|
- run: bundle exec rspec
|
39
39
|
check-misc:
|
40
40
|
runs-on: ubuntu-20.04
|
41
|
+
strategy:
|
42
|
+
matrix:
|
43
|
+
ruby: ['head']
|
41
44
|
steps:
|
42
|
-
- uses: actions/checkout@
|
45
|
+
- uses: actions/checkout@v4
|
46
|
+
- uses: ruby/setup-ruby@v1
|
47
|
+
with:
|
48
|
+
ruby-version: ${{ matrix.ruby }}
|
49
|
+
bundler-cache: true
|
50
|
+
- run: bundle install
|
51
|
+
|
43
52
|
# Copy from https://github.com/ruby/ruby/blob/089227e94823542acfdafa68541d330eee42ffea/.github/workflows/check_misc.yml#L27
|
44
53
|
- name: Check for trailing spaces
|
45
54
|
run: |
|
46
|
-
git grep -I -n '[ ]$' -- '*.rb' '*.[chy]' '*.rs' && exit 1 || :
|
55
|
+
git grep -I -n '[ ]$' -- '*.rb' '*.[chy]' '*.rs' ':!spec/' && exit 1 || :
|
47
56
|
git grep -n '^[ ][ ]*$' -- '*.md' && exit 1 || :
|
57
|
+
|
58
|
+
- name: Check for parser.rb is up to date
|
59
|
+
run: |
|
60
|
+
bundle exec rake build:racc_parser
|
61
|
+
git diff --color --no-ext-diff --ignore-submodules --exit-code lib/lrama/parser.rb
|
48
62
|
steep-check:
|
49
63
|
runs-on: ubuntu-20.04
|
50
64
|
strategy:
|
@@ -52,7 +66,7 @@ jobs:
|
|
52
66
|
matrix:
|
53
67
|
ruby: ['head']
|
54
68
|
steps:
|
55
|
-
- uses: actions/checkout@
|
69
|
+
- uses: actions/checkout@v4
|
56
70
|
- uses: ruby/setup-ruby@v1
|
57
71
|
with:
|
58
72
|
ruby-version: ${{ matrix.ruby }}
|
@@ -65,13 +79,15 @@ jobs:
|
|
65
79
|
strategy:
|
66
80
|
fail-fast: false
|
67
81
|
matrix:
|
68
|
-
|
82
|
+
# '3.0' is the oldest living ruby version
|
83
|
+
# '2.5' is for BASERUBY
|
84
|
+
baseruby: ['head', '3.0', '2.5']
|
69
85
|
ruby_branch: ['master']
|
70
86
|
defaults:
|
71
87
|
run:
|
72
88
|
working-directory: ../ruby/build
|
73
89
|
steps:
|
74
|
-
- uses: actions/checkout@
|
90
|
+
- uses: actions/checkout@v4
|
75
91
|
- uses: ruby/setup-ruby@v1
|
76
92
|
with:
|
77
93
|
ruby-version: ${{ matrix.baseruby }}
|
data/.gitignore
CHANGED
data/Gemfile
CHANGED
data/README.md
CHANGED
@@ -61,12 +61,78 @@ This branch generates "parse.c" compatible with Bison 3.8.2 for ruby 3.0, 3.1, 3
|
|
61
61
|
|
62
62
|
Lrama is executed with BASERUBY when building ruby from source code. Therefore Lrama needs to support BASERUBY, currently 2.5, or later version.
|
63
63
|
|
64
|
-
This also requires Lrama to be able to run with only default gems
|
64
|
+
This also requires Lrama to be able to run with only default gems because BASERUBY runs with `--disable=gems` option.
|
65
65
|
|
66
|
-
##
|
66
|
+
## Development
|
67
|
+
|
68
|
+
### How to generate new_parser.rb
|
69
|
+
|
70
|
+
```shell
|
71
|
+
$ rake build:racc_parser
|
72
|
+
```
|
73
|
+
|
74
|
+
`new_parser.rb` is generated from `parser.y` by Racc.
|
75
|
+
Run the rake command when you update `parser.y` then commit changes of both files.
|
76
|
+
|
77
|
+
### Test
|
78
|
+
|
79
|
+
Running tests:
|
80
|
+
|
81
|
+
```shell
|
82
|
+
$ bundle install
|
83
|
+
$ bundle exec rspec
|
84
|
+
```
|
85
|
+
|
86
|
+
Running type check:
|
87
|
+
|
88
|
+
```shell
|
89
|
+
$ bundle install
|
90
|
+
$ bundle exec rbs collection install
|
91
|
+
$ bundle exec steep check
|
92
|
+
```
|
93
|
+
|
94
|
+
### Profiling Lrama
|
95
|
+
|
96
|
+
#### 1. Create parse.tmp.y in ruby/ruby
|
97
|
+
|
98
|
+
```shell
|
99
|
+
$ ruby tool/id2token.rb parse.y > parse.tmp.y
|
100
|
+
$ cp parse.tmp.y dir/lrama/tmp
|
101
|
+
```
|
102
|
+
|
103
|
+
#### 2. Enable Profiler
|
104
|
+
|
105
|
+
```diff
|
106
|
+
diff --git a/exe/lrama b/exe/lrama
|
107
|
+
index ba5fb06..2497178 100755
|
108
|
+
--- a/exe/lrama
|
109
|
+
+++ b/exe/lrama
|
110
|
+
@@ -3,4 +3,6 @@
|
111
|
+
$LOAD_PATH << File.join(__dir__, "../lib")
|
112
|
+
require "lrama"
|
113
|
+
|
114
|
+
-Lrama::Command.new.run(ARGV.dup)
|
115
|
+
+Lrama::Report::Profile.report_profile do
|
116
|
+
+ Lrama::Command.new.run(ARGV.dup)
|
117
|
+
+end
|
118
|
+
```
|
119
|
+
|
120
|
+
#### 3. Run Lrama
|
121
|
+
|
122
|
+
```shell
|
123
|
+
$ exe/lrama -o parse.tmp.c --header=parse.tmp.h tmp/parse.tmp.y
|
124
|
+
```
|
125
|
+
|
126
|
+
#### 4. Generate Flamegraph
|
127
|
+
|
128
|
+
```shell
|
129
|
+
$ stackprof --d3-flamegraph tmp/stackprof-cpu-myapp.dump > tmp/flamegraph.html
|
130
|
+
```
|
131
|
+
|
132
|
+
### Build Ruby
|
67
133
|
|
68
134
|
1. Install Lrama
|
69
|
-
2. Run `make
|
135
|
+
2. Run `make main`
|
70
136
|
|
71
137
|
## Release flow
|
72
138
|
|
data/Rakefile
CHANGED
@@ -1 +1,13 @@
|
|
1
1
|
require "bundler/gem_tasks"
|
2
|
+
|
3
|
+
namespace "build" do
|
4
|
+
desc "build parser from parser.y by using Racc"
|
5
|
+
task :racc_parser do
|
6
|
+
`bundle exec racc parser.y --embedded -o lib/lrama/parser.rb`
|
7
|
+
end
|
8
|
+
|
9
|
+
desc "build parser for debugging"
|
10
|
+
task :racc_verbose_parser do
|
11
|
+
`bundle exec racc parser.y --embedded -o lib/lrama/parser.rb -t --log-file=parser.output`
|
12
|
+
end
|
13
|
+
end
|
data/Steepfile
CHANGED
@@ -1,10 +1,13 @@
|
|
1
1
|
# D = Steep::Diagnostic
|
2
2
|
#
|
3
3
|
target :lib do
|
4
|
+
repo_path '.gem_rbs_collection/'
|
4
5
|
signature "sig"
|
5
6
|
|
6
7
|
check "lib/lrama/bitmap.rb"
|
8
|
+
check "lib/lrama/digraph.rb"
|
7
9
|
check "lib/lrama/report/duration.rb"
|
8
10
|
check "lib/lrama/report/profile.rb"
|
11
|
+
check "lib/lrama/token/type.rb"
|
9
12
|
check "lib/lrama/warning.rb"
|
10
13
|
end
|
data/exe/lrama
CHANGED
data/lib/lrama/command.rb
CHANGED
@@ -1,53 +1,35 @@
|
|
1
|
-
require 'optparse'
|
2
|
-
|
3
1
|
module Lrama
|
4
2
|
class Command
|
5
|
-
def
|
6
|
-
|
7
|
-
|
8
|
-
@skeleton = "bison/yacc.c"
|
9
|
-
@header = false
|
10
|
-
@header_file = nil
|
11
|
-
@report = []
|
12
|
-
@report_file = nil
|
13
|
-
@outfile = "y.tab.c"
|
14
|
-
@trace = []
|
15
|
-
@error_recovery = false
|
16
|
-
@grammar_file = nil
|
17
|
-
@report_file = nil
|
18
|
-
@trace_opts = nil
|
19
|
-
@report_opts = nil
|
20
|
-
end
|
3
|
+
def run(argv)
|
4
|
+
options = OptionParser.new.parse(argv)
|
21
5
|
|
22
|
-
|
23
|
-
parse_option
|
24
|
-
|
25
|
-
Report::Duration.enable if @trace_opts[:time]
|
6
|
+
Report::Duration.enable if options.trace_opts[:time]
|
26
7
|
|
27
8
|
warning = Lrama::Warning.new
|
28
|
-
|
29
|
-
|
30
|
-
|
9
|
+
text = options.y.read
|
10
|
+
options.y.close if options.y != STDIN
|
11
|
+
grammar = Lrama::Parser.new(text).parse
|
12
|
+
states = Lrama::States.new(grammar, warning, trace_state: (options.trace_opts[:automaton] || options.trace_opts[:closure]))
|
31
13
|
states.compute
|
32
14
|
context = Lrama::Context.new(states)
|
33
15
|
|
34
|
-
if
|
16
|
+
if options.report_file
|
35
17
|
reporter = Lrama::StatesReporter.new(states)
|
36
|
-
File.open(
|
37
|
-
reporter.report(f,
|
18
|
+
File.open(options.report_file, "w+") do |f|
|
19
|
+
reporter.report(f, **options.report_opts)
|
38
20
|
end
|
39
21
|
end
|
40
22
|
|
41
|
-
File.open(
|
23
|
+
File.open(options.outfile, "w+") do |f|
|
42
24
|
Lrama::Output.new(
|
43
25
|
out: f,
|
44
|
-
output_file_path:
|
45
|
-
template_name:
|
46
|
-
grammar_file_path:
|
47
|
-
header_file_path:
|
26
|
+
output_file_path: options.outfile,
|
27
|
+
template_name: options.skeleton,
|
28
|
+
grammar_file_path: options.grammar_file,
|
29
|
+
header_file_path: options.header_file,
|
48
30
|
context: context,
|
49
31
|
grammar: grammar,
|
50
|
-
error_recovery:
|
32
|
+
error_recovery: options.error_recovery,
|
51
33
|
).render
|
52
34
|
end
|
53
35
|
|
@@ -55,108 +37,5 @@ module Lrama
|
|
55
37
|
exit 1
|
56
38
|
end
|
57
39
|
end
|
58
|
-
|
59
|
-
private
|
60
|
-
|
61
|
-
def validate_report(report)
|
62
|
-
bison_list = %w[states itemsets lookaheads solved counterexamples cex all none]
|
63
|
-
others = %w[verbose]
|
64
|
-
list = bison_list + others
|
65
|
-
not_supported = %w[cex none]
|
66
|
-
h = { grammar: true }
|
67
|
-
|
68
|
-
report.each do |r|
|
69
|
-
if list.include?(r) && !not_supported.include?(r)
|
70
|
-
h[r.to_sym] = true
|
71
|
-
else
|
72
|
-
raise "Invalid report option \"#{r}\"."
|
73
|
-
end
|
74
|
-
end
|
75
|
-
|
76
|
-
if h[:all]
|
77
|
-
(bison_list - not_supported).each do |r|
|
78
|
-
h[r.to_sym] = true
|
79
|
-
end
|
80
|
-
|
81
|
-
h.delete(:all)
|
82
|
-
end
|
83
|
-
|
84
|
-
return h
|
85
|
-
end
|
86
|
-
|
87
|
-
def validate_trace(trace)
|
88
|
-
list = %w[
|
89
|
-
none locations scan parse automaton bitsets
|
90
|
-
closure grammar resource sets muscles tools
|
91
|
-
m4-early m4 skeleton time ielr cex all
|
92
|
-
]
|
93
|
-
h = {}
|
94
|
-
|
95
|
-
trace.each do |t|
|
96
|
-
if list.include?(t)
|
97
|
-
h[t.to_sym] = true
|
98
|
-
else
|
99
|
-
raise "Invalid trace option \"#{t}\"."
|
100
|
-
end
|
101
|
-
end
|
102
|
-
|
103
|
-
return h
|
104
|
-
end
|
105
|
-
|
106
|
-
def parse_option
|
107
|
-
opt = OptionParser.new
|
108
|
-
|
109
|
-
# opt.on('-h') {|v| p v }
|
110
|
-
opt.on('-V', '--version') {|v| puts "lrama #{Lrama::VERSION}"; exit 0 }
|
111
|
-
|
112
|
-
# Tuning the Parser
|
113
|
-
opt.on('-S', '--skeleton=FILE') {|v| @skeleton = v }
|
114
|
-
opt.on('-t') { } # Do nothing
|
115
|
-
|
116
|
-
# Output Files:
|
117
|
-
opt.on('-h', '--header=[FILE]') {|v| @header = true; @header_file = v }
|
118
|
-
opt.on('-d') { @header = true }
|
119
|
-
opt.on('-r', '--report=THINGS', Array) {|v| @report = v }
|
120
|
-
opt.on('--report-file=FILE') {|v| @report_file = v }
|
121
|
-
opt.on('-v') { } # Do nothing
|
122
|
-
opt.on('-o', '--output=FILE') {|v| @outfile = v }
|
123
|
-
|
124
|
-
# Hidden
|
125
|
-
opt.on('--trace=THINGS', Array) {|v| @trace = v }
|
126
|
-
|
127
|
-
# Error Recovery
|
128
|
-
opt.on('-e') {|v| @error_recovery = true }
|
129
|
-
|
130
|
-
opt.parse!(@argv)
|
131
|
-
|
132
|
-
@trace_opts = validate_trace(@trace)
|
133
|
-
@report_opts = validate_report(@report)
|
134
|
-
|
135
|
-
@grammar_file = @argv.shift
|
136
|
-
|
137
|
-
if !@grammar_file
|
138
|
-
abort "File should be specified\n"
|
139
|
-
end
|
140
|
-
|
141
|
-
if @grammar_file == '-'
|
142
|
-
@grammar_file = @argv.shift or abort "File name for STDIN should be specified\n"
|
143
|
-
@y = STDIN
|
144
|
-
else
|
145
|
-
@y = File.open(@grammar_file, 'r')
|
146
|
-
end
|
147
|
-
|
148
|
-
if !@report.empty? && @report_file.nil? && @grammar_file
|
149
|
-
@report_file = File.dirname(@grammar_file) + "/" + File.basename(@grammar_file, ".*") + ".output"
|
150
|
-
end
|
151
|
-
|
152
|
-
if !@header_file && @header
|
153
|
-
case
|
154
|
-
when @outfile
|
155
|
-
@header_file = File.dirname(@outfile) + "/" + File.basename(@outfile, ".*") + ".h"
|
156
|
-
when @grammar_file
|
157
|
-
@header_file = File.dirname(@grammar_file) + "/" + File.basename(@grammar_file, ".*") + ".h"
|
158
|
-
end
|
159
|
-
end
|
160
|
-
end
|
161
40
|
end
|
162
41
|
end
|
data/lib/lrama/context.rb
CHANGED
@@ -170,7 +170,7 @@ module Lrama
|
|
170
170
|
return a
|
171
171
|
end
|
172
172
|
|
173
|
-
# Mapping from rule number to
|
173
|
+
# Mapping from rule number to length of RHS.
|
174
174
|
# Dummy rule is appended as the first element whose value is 0
|
175
175
|
# because 0 means error in yydefact.
|
176
176
|
def yyr2
|
@@ -214,7 +214,7 @@ module Lrama
|
|
214
214
|
(rule_id + 1) * -1
|
215
215
|
end
|
216
216
|
|
217
|
-
# Symbol number is
|
217
|
+
# Symbol number is assigned to term first then nterm.
|
218
218
|
# This method calculates sequence_number for nterm.
|
219
219
|
def nterm_number_to_sequence_number(nterm_number)
|
220
220
|
nterm_number - @states.terms.count
|
@@ -259,7 +259,7 @@ module Lrama
|
|
259
259
|
actions[conflict.symbol.number] = ErrorActionNumber
|
260
260
|
end
|
261
261
|
|
262
|
-
# If default_reduction_rule,
|
262
|
+
# If default_reduction_rule, replace default_reduction_rule in
|
263
263
|
# actions with zero.
|
264
264
|
if state.default_reduction_rule
|
265
265
|
actions.map! do |e|
|
@@ -272,7 +272,7 @@ module Lrama
|
|
272
272
|
end
|
273
273
|
|
274
274
|
# If no default_reduction_rule, default behavior is an
|
275
|
-
# error then
|
275
|
+
# error then replace ErrorActionNumber with zero.
|
276
276
|
if !state.default_reduction_rule
|
277
277
|
actions.map! do |e|
|
278
278
|
if e == ErrorActionNumber
|
data/lib/lrama/digraph.rb
CHANGED
data/lib/lrama/grammar/union.rb
CHANGED
data/lib/lrama/grammar.rb
CHANGED
@@ -1,3 +1,5 @@
|
|
1
|
+
require "strscan"
|
2
|
+
|
1
3
|
require "lrama/grammar/auxiliary"
|
2
4
|
require "lrama/grammar/code"
|
3
5
|
require "lrama/grammar/error_token"
|
@@ -306,6 +308,188 @@ module Lrama
|
|
306
308
|
@nterms ||= @symbols.select(&:nterm?)
|
307
309
|
end
|
308
310
|
|
311
|
+
def extract_references
|
312
|
+
unless initial_action.nil?
|
313
|
+
scanner = StringScanner.new(initial_action.s_value)
|
314
|
+
references = []
|
315
|
+
|
316
|
+
while !scanner.eos? do
|
317
|
+
start = scanner.pos
|
318
|
+
case
|
319
|
+
# $ references
|
320
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
321
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\$/) # $$, $<long>$
|
322
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
323
|
+
references << [:dollar, "$", tag, start, scanner.pos - 1]
|
324
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?(\d+)/) # $1, $2, $<long>1
|
325
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
326
|
+
references << [:dollar, Integer(scanner[2]), tag, start, scanner.pos - 1]
|
327
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?([a-zA-Z_][a-zA-Z0-9_]*)/) # $foo, $expr, $<long>program (named reference without brackets)
|
328
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
329
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
330
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # $expr.right, $expr-right, $<long>program (named reference with brackets)
|
331
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
332
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
333
|
+
|
334
|
+
# @ references
|
335
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
336
|
+
when scanner.scan(/@\$/) # @$
|
337
|
+
references << [:at, "$", nil, start, scanner.pos - 1]
|
338
|
+
when scanner.scan(/@(\d+)/) # @1
|
339
|
+
references << [:at, Integer(scanner[1]), nil, start, scanner.pos - 1]
|
340
|
+
when scanner.scan(/@([a-zA-Z][a-zA-Z0-9_]*)/) # @foo, @expr (named reference without brackets)
|
341
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
342
|
+
when scanner.scan(/@\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # @expr.right, @expr-right (named reference with brackets)
|
343
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
344
|
+
else
|
345
|
+
scanner.getch
|
346
|
+
end
|
347
|
+
end
|
348
|
+
|
349
|
+
initial_action.token_code.references = references
|
350
|
+
build_references(initial_action.token_code)
|
351
|
+
end
|
352
|
+
|
353
|
+
@printers.each do |printer|
|
354
|
+
scanner = StringScanner.new(printer.code.s_value)
|
355
|
+
references = []
|
356
|
+
|
357
|
+
while !scanner.eos? do
|
358
|
+
start = scanner.pos
|
359
|
+
case
|
360
|
+
# $ references
|
361
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
362
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\$/) # $$, $<long>$
|
363
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
364
|
+
references << [:dollar, "$", tag, start, scanner.pos - 1]
|
365
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?(\d+)/) # $1, $2, $<long>1
|
366
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
367
|
+
references << [:dollar, Integer(scanner[2]), tag, start, scanner.pos - 1]
|
368
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?([a-zA-Z_][a-zA-Z0-9_]*)/) # $foo, $expr, $<long>program (named reference without brackets)
|
369
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
370
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
371
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # $expr.right, $expr-right, $<long>program (named reference with brackets)
|
372
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
373
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
374
|
+
|
375
|
+
# @ references
|
376
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
377
|
+
when scanner.scan(/@\$/) # @$
|
378
|
+
references << [:at, "$", nil, start, scanner.pos - 1]
|
379
|
+
when scanner.scan(/@(\d+)/) # @1
|
380
|
+
references << [:at, Integer(scanner[1]), nil, start, scanner.pos - 1]
|
381
|
+
when scanner.scan(/@([a-zA-Z][a-zA-Z0-9_]*)/) # @foo, @expr (named reference without brackets)
|
382
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
383
|
+
when scanner.scan(/@\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # @expr.right, @expr-right (named reference with brackets)
|
384
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
385
|
+
else
|
386
|
+
scanner.getch
|
387
|
+
end
|
388
|
+
end
|
389
|
+
|
390
|
+
printer.code.token_code.references = references
|
391
|
+
build_references(printer.code.token_code)
|
392
|
+
end
|
393
|
+
|
394
|
+
@error_tokens.each do |error_token|
|
395
|
+
scanner = StringScanner.new(error_token.code.s_value)
|
396
|
+
references = []
|
397
|
+
|
398
|
+
while !scanner.eos? do
|
399
|
+
start = scanner.pos
|
400
|
+
case
|
401
|
+
# $ references
|
402
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
403
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\$/) # $$, $<long>$
|
404
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
405
|
+
references << [:dollar, "$", tag, start, scanner.pos - 1]
|
406
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?(\d+)/) # $1, $2, $<long>1
|
407
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
408
|
+
references << [:dollar, Integer(scanner[2]), tag, start, scanner.pos - 1]
|
409
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?([a-zA-Z_][a-zA-Z0-9_]*)/) # $foo, $expr, $<long>program (named reference without brackets)
|
410
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
411
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
412
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # $expr.right, $expr-right, $<long>program (named reference with brackets)
|
413
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
414
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
415
|
+
|
416
|
+
# @ references
|
417
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
418
|
+
when scanner.scan(/@\$/) # @$
|
419
|
+
references << [:at, "$", nil, start, scanner.pos - 1]
|
420
|
+
when scanner.scan(/@(\d+)/) # @1
|
421
|
+
references << [:at, Integer(scanner[1]), nil, start, scanner.pos - 1]
|
422
|
+
when scanner.scan(/@([a-zA-Z][a-zA-Z0-9_]*)/) # @foo, @expr (named reference without brackets)
|
423
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
424
|
+
when scanner.scan(/@\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # @expr.right, @expr-right (named reference with brackets)
|
425
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
426
|
+
else
|
427
|
+
scanner.getch
|
428
|
+
end
|
429
|
+
end
|
430
|
+
|
431
|
+
error_token.code.token_code.references = references
|
432
|
+
build_references(error_token.code.token_code)
|
433
|
+
end
|
434
|
+
|
435
|
+
@_rules.each do |lhs, rhs, _|
|
436
|
+
rhs.each_with_index do |token, index|
|
437
|
+
next if token.class == Lrama::Grammar::Symbol || token.type != Lrama::Lexer::Token::User_code
|
438
|
+
|
439
|
+
scanner = StringScanner.new(token.s_value)
|
440
|
+
references = []
|
441
|
+
|
442
|
+
while !scanner.eos? do
|
443
|
+
start = scanner.pos
|
444
|
+
case
|
445
|
+
# $ references
|
446
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
447
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\$/) # $$, $<long>$
|
448
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
449
|
+
references << [:dollar, "$", tag, start, scanner.pos - 1]
|
450
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?(\d+)/) # $1, $2, $<long>1
|
451
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
452
|
+
references << [:dollar, Integer(scanner[2]), tag, start, scanner.pos - 1]
|
453
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?([a-zA-Z_][a-zA-Z0-9_]*)/) # $foo, $expr, $<long>program (named reference without brackets)
|
454
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
455
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
456
|
+
when scanner.scan(/\$(<[a-zA-Z0-9_]+>)?\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # $expr.right, $expr-right, $<long>program (named reference with brackets)
|
457
|
+
tag = scanner[1] ? Lrama::Lexer::Token.new(type: Lrama::Lexer::Token::Tag, s_value: scanner[1]) : nil
|
458
|
+
references << [:dollar, scanner[2], tag, start, scanner.pos - 1]
|
459
|
+
|
460
|
+
# @ references
|
461
|
+
# It need to wrap an identifier with brackets to use ".-" for identifiers
|
462
|
+
when scanner.scan(/@\$/) # @$
|
463
|
+
references << [:at, "$", nil, start, scanner.pos - 1]
|
464
|
+
when scanner.scan(/@(\d+)/) # @1
|
465
|
+
references << [:at, Integer(scanner[1]), nil, start, scanner.pos - 1]
|
466
|
+
when scanner.scan(/@([a-zA-Z][a-zA-Z0-9_]*)/) # @foo, @expr (named reference without brackets)
|
467
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
468
|
+
when scanner.scan(/@\[([a-zA-Z_.][-a-zA-Z0-9_.]*)\]/) # @expr.right, @expr-right (named reference with brackets)
|
469
|
+
references << [:at, scanner[1], nil, start, scanner.pos - 1]
|
470
|
+
|
471
|
+
when scanner.scan(/\/\*/)
|
472
|
+
scanner.scan_until(/\*\//)
|
473
|
+
else
|
474
|
+
scanner.getch
|
475
|
+
end
|
476
|
+
end
|
477
|
+
|
478
|
+
token.references = references
|
479
|
+
token.numberize_references(lhs, rhs)
|
480
|
+
build_references(token)
|
481
|
+
end
|
482
|
+
end
|
483
|
+
end
|
484
|
+
|
485
|
+
def create_token(type, s_value, line, column)
|
486
|
+
t = Token.new(type: type, s_value: s_value)
|
487
|
+
t.line = line
|
488
|
+
t.column = column
|
489
|
+
|
490
|
+
return t
|
491
|
+
end
|
492
|
+
|
309
493
|
private
|
310
494
|
|
311
495
|
def find_nterm_by_id!(id)
|
@@ -470,7 +654,9 @@ module Lrama
|
|
470
654
|
|
471
655
|
# Fill #number and #token_id
|
472
656
|
def fill_symbol_number
|
473
|
-
#
|
657
|
+
# Character literal in grammar file has
|
658
|
+
# token id corresponding to ASCII code by default,
|
659
|
+
# so start token_id from 256.
|
474
660
|
token_id = 256
|
475
661
|
|
476
662
|
# YYEMPTY = -2
|