ruby_parser 3.2.2 → 3.3.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- checksums.yaml.gz.sig +0 -0
- data.tar.gz.sig +0 -0
- data/.autotest +0 -2
- data/History.txt +38 -0
- data/Rakefile +2 -6
- data/lib/ruby18_parser.rb +2003 -2010
- data/lib/ruby18_parser.y +6 -8
- data/lib/ruby19_parser.rb +2016 -2004
- data/lib/ruby19_parser.y +6 -8
- data/lib/ruby20_parser.rb +2189 -2154
- data/lib/ruby20_parser.y +21 -13
- data/lib/ruby_lexer.rb +645 -812
- data/lib/ruby_parser_extras.rb +17 -46
- data/test/test_ruby_lexer.rb +1280 -1108
- data/test/test_ruby_parser.rb +101 -1
- data/test/test_ruby_parser_extras.rb +1 -63
- metadata +112 -133
- metadata.gz.sig +0 -0
data/lib/ruby20_parser.y
CHANGED
@@ -263,7 +263,7 @@ rule
|
|
263
263
|
}
|
264
264
|
opt_block_param
|
265
265
|
{
|
266
|
-
result = self.env.dynamic.keys
|
266
|
+
result = nil # self.env.dynamic.keys
|
267
267
|
}
|
268
268
|
compstmt tRCURLY
|
269
269
|
{
|
@@ -785,7 +785,6 @@ rule
|
|
785
785
|
}
|
786
786
|
| arg tEH arg opt_nl tCOLON arg
|
787
787
|
{
|
788
|
-
lexer.tern.pop
|
789
788
|
result = s(:if, val[0], val[2], val[5])
|
790
789
|
}
|
791
790
|
| primary
|
@@ -1136,7 +1135,7 @@ rule
|
|
1136
1135
|
}
|
1137
1136
|
| kDEF fname
|
1138
1137
|
{
|
1139
|
-
result =
|
1138
|
+
result = self.in_def
|
1140
1139
|
|
1141
1140
|
self.comments.push self.lexer.comments
|
1142
1141
|
self.in_def = true
|
@@ -1144,10 +1143,9 @@ rule
|
|
1144
1143
|
}
|
1145
1144
|
f_arglist bodystmt kEND
|
1146
1145
|
{
|
1147
|
-
|
1146
|
+
in_def = val[2]
|
1148
1147
|
|
1149
1148
|
result = new_defn val
|
1150
|
-
result[2].line line
|
1151
1149
|
|
1152
1150
|
self.env.unextend
|
1153
1151
|
self.in_def = in_def
|
@@ -1429,7 +1427,9 @@ opt_block_args_tail: tCOMMA block_args_tail
|
|
1429
1427
|
| f_bad_arg
|
1430
1428
|
|
1431
1429
|
lambda: {
|
1432
|
-
|
1430
|
+
self.env.extend :dynamic
|
1431
|
+
result = self.lexer.lineno
|
1432
|
+
|
1433
1433
|
result = lexer.lpar_beg
|
1434
1434
|
lexer.paren_nest += 1
|
1435
1435
|
lexer.lpar_beg = lexer.paren_nest
|
@@ -1443,6 +1443,7 @@ opt_block_args_tail: tCOMMA block_args_tail
|
|
1443
1443
|
|
1444
1444
|
call = new_call nil, :lambda
|
1445
1445
|
result = new_iter call, args, body
|
1446
|
+
self.env.unextend
|
1446
1447
|
}
|
1447
1448
|
|
1448
1449
|
f_larglist: tLPAREN2 f_args opt_bv_decl rparen
|
@@ -1470,7 +1471,7 @@ opt_block_args_tail: tCOMMA block_args_tail
|
|
1470
1471
|
}
|
1471
1472
|
opt_block_param
|
1472
1473
|
{
|
1473
|
-
result = self.env.dynamic.keys
|
1474
|
+
result = nil # self.env.dynamic.keys
|
1474
1475
|
}
|
1475
1476
|
compstmt kEND
|
1476
1477
|
{
|
@@ -1556,7 +1557,7 @@ opt_block_args_tail: tCOMMA block_args_tail
|
|
1556
1557
|
}
|
1557
1558
|
opt_block_param
|
1558
1559
|
{
|
1559
|
-
result = self.env.dynamic.keys
|
1560
|
+
result = nil # self.env.dynamic.keys
|
1560
1561
|
}
|
1561
1562
|
compstmt tRCURLY
|
1562
1563
|
{
|
@@ -1574,7 +1575,7 @@ opt_block_args_tail: tCOMMA block_args_tail
|
|
1574
1575
|
}
|
1575
1576
|
opt_block_param
|
1576
1577
|
{
|
1577
|
-
result = self.env.dynamic.keys
|
1578
|
+
result = nil # self.env.dynamic.keys
|
1578
1579
|
}
|
1579
1580
|
compstmt kEND
|
1580
1581
|
{
|
@@ -1972,13 +1973,14 @@ keyword_variable: kNIL { result = s(:nil) }
|
|
1972
1973
|
f_arglist: tLPAREN2 f_args rparen
|
1973
1974
|
{
|
1974
1975
|
result = val[1]
|
1975
|
-
lexer.lex_state = :expr_beg
|
1976
|
+
self.lexer.lex_state = :expr_beg
|
1976
1977
|
self.lexer.command_start = true
|
1977
1978
|
}
|
1978
1979
|
| f_args term
|
1979
1980
|
{
|
1980
|
-
self.lexer.lex_state = :expr_beg
|
1981
1981
|
result = val[0]
|
1982
|
+
self.lexer.lex_state = :expr_beg
|
1983
|
+
self.lexer.command_start = true
|
1982
1984
|
}
|
1983
1985
|
|
1984
1986
|
args_tail: f_kwarg tCOMMA f_kwrest opt_f_block_arg
|
@@ -2127,14 +2129,20 @@ keyword_variable: kNIL { result = s(:nil) }
|
|
2127
2129
|
{
|
2128
2130
|
# TODO: call_args
|
2129
2131
|
label, _ = val[0] # TODO: fix lineno?
|
2130
|
-
|
2132
|
+
identifier = label.to_sym
|
2133
|
+
self.env[identifier] = :lvar
|
2134
|
+
|
2135
|
+
result = s(:array, s(:kwarg, identifier, val[1]))
|
2131
2136
|
}
|
2132
2137
|
|
2133
2138
|
f_block_kw: tLABEL primary_value
|
2134
2139
|
{
|
2135
2140
|
# TODO: call_args
|
2136
2141
|
label, _ = val[0] # TODO: fix lineno?
|
2137
|
-
|
2142
|
+
identifier = label.to_sym
|
2143
|
+
self.env[identifier] = :lvar
|
2144
|
+
|
2145
|
+
result = s(:array, s(:kwarg, identifier, val[1]))
|
2138
2146
|
}
|
2139
2147
|
|
2140
2148
|
f_block_kwarg: f_block_kw
|
data/lib/ruby_lexer.rb
CHANGED
@@ -5,61 +5,21 @@ class RubyLexer
|
|
5
5
|
# :stopdoc:
|
6
6
|
RUBY19 = "".respond_to? :encoding
|
7
7
|
|
8
|
-
|
9
|
-
|
10
|
-
|
11
|
-
|
12
|
-
|
13
|
-
|
14
|
-
IDENT_RE = /^#{IDENT_CHAR_RE}+/o
|
15
|
-
|
16
|
-
attr_accessor :command_start
|
17
|
-
attr_accessor :cmdarg
|
18
|
-
attr_accessor :cond
|
19
|
-
attr_accessor :tern # TODO: rename ternary damnit... wtf
|
20
|
-
attr_accessor :string_nest
|
21
|
-
|
22
|
-
ESC_RE = /\\((?>[0-7]{1,3}|x[0-9a-fA-F]{1,2}|M-[^\\]|(C-|c)[^\\]|[^0-7xMCc]))/u
|
23
|
-
# :startdoc:
|
24
|
-
|
25
|
-
##
|
26
|
-
# What version of ruby to parse. 18 and 19 are the only valid values
|
27
|
-
# currently supported.
|
28
|
-
|
29
|
-
attr_accessor :version
|
30
|
-
|
31
|
-
# Additional context surrounding tokens that both the lexer and
|
32
|
-
# grammar use.
|
33
|
-
attr_reader :lex_state
|
34
|
-
|
35
|
-
attr_accessor :lex_strterm
|
36
|
-
|
37
|
-
attr_accessor :parser # HACK for very end of lexer... *sigh*
|
38
|
-
|
39
|
-
# Stream of data that yylex examines.
|
40
|
-
attr_reader :src
|
41
|
-
|
42
|
-
# Last token read via yylex.
|
43
|
-
attr_accessor :token
|
44
|
-
|
45
|
-
attr_accessor :string_buffer
|
46
|
-
|
47
|
-
# Value of last token which had a value associated with it.
|
48
|
-
attr_accessor :yacc_value
|
49
|
-
|
50
|
-
# What handles warnings
|
51
|
-
attr_accessor :warnings
|
8
|
+
IDENT_CHAR = if RUBY19 then
|
9
|
+
/[\w\u0080-\u{10ffff}]/u
|
10
|
+
else
|
11
|
+
/[\w\x80-\xFF]/n
|
12
|
+
end
|
52
13
|
|
53
|
-
|
54
|
-
|
55
|
-
|
56
|
-
|
14
|
+
IDENT = /^#{IDENT_CHAR}+/o
|
15
|
+
ESC = /\\((?>[0-7]{1,3}|x[0-9a-fA-F]{1,2}|M-[^\\]|(C-|c)[^\\]|u[0-9a-fA-F]+|u\{[0-9a-fA-F]+\}|[^0-7xMCc]))/u
|
16
|
+
SIMPLE_STRING = /(#{ESC}|#(#{ESC}|[^\{\#\@\$\"\\])|[^\"\\\#])*/o
|
17
|
+
SIMPLE_SSTRING = /(\\.|[^\'])*/
|
57
18
|
|
58
19
|
EOF = :eof_haha!
|
59
20
|
|
60
21
|
# ruby constants for strings (should this be moved somewhere else?)
|
61
22
|
|
62
|
-
# :stopdoc:
|
63
23
|
STR_FUNC_BORING = 0x00
|
64
24
|
STR_FUNC_ESCAPE = 0x01 # TODO: remove and replace with REGEXP
|
65
25
|
STR_FUNC_EXPAND = 0x02
|
@@ -75,6 +35,22 @@ class RubyLexer
|
|
75
35
|
STR_SSYM = STR_FUNC_SYMBOL
|
76
36
|
STR_DSYM = STR_FUNC_SYMBOL | STR_FUNC_EXPAND
|
77
37
|
|
38
|
+
ESCAPES = {
|
39
|
+
"a" => "\007",
|
40
|
+
"b" => "\010",
|
41
|
+
"e" => "\033",
|
42
|
+
"f" => "\f",
|
43
|
+
"n" => "\n",
|
44
|
+
"r" => "\r",
|
45
|
+
"s" => " ",
|
46
|
+
"t" => "\t",
|
47
|
+
"v" => "\13",
|
48
|
+
"\\" => '\\',
|
49
|
+
"\n" => "",
|
50
|
+
"C-\?" => 127.chr,
|
51
|
+
"c\?" => 127.chr,
|
52
|
+
}
|
53
|
+
|
78
54
|
TOKENS = {
|
79
55
|
"!" => :tBANG,
|
80
56
|
"!=" => :tNEQ,
|
@@ -90,8 +66,58 @@ class RubyLexer
|
|
90
66
|
"=~" => :tMATCH,
|
91
67
|
"->" => :tLAMBDA,
|
92
68
|
}
|
69
|
+
|
70
|
+
@@regexp_cache = Hash.new { |h,k| h[k] = Regexp.new(Regexp.escape(k)) }
|
71
|
+
@@regexp_cache[nil] = nil
|
72
|
+
|
93
73
|
# :startdoc:
|
94
74
|
|
75
|
+
attr_accessor :brace_nest
|
76
|
+
attr_accessor :cmdarg
|
77
|
+
attr_accessor :command_start
|
78
|
+
attr_accessor :cond
|
79
|
+
|
80
|
+
##
|
81
|
+
# Additional context surrounding tokens that both the lexer and
|
82
|
+
# grammar use.
|
83
|
+
|
84
|
+
attr_accessor :lex_state
|
85
|
+
|
86
|
+
attr_accessor :lex_strterm
|
87
|
+
attr_accessor :lpar_beg
|
88
|
+
attr_accessor :paren_nest
|
89
|
+
attr_accessor :parser # HACK for very end of lexer... *sigh*
|
90
|
+
attr_accessor :space_seen
|
91
|
+
attr_accessor :string_buffer
|
92
|
+
attr_accessor :string_nest
|
93
|
+
|
94
|
+
# Stream of data that yylex examines.
|
95
|
+
attr_reader :src
|
96
|
+
alias :ss :src
|
97
|
+
|
98
|
+
# Last token read via yylex.
|
99
|
+
attr_accessor :token
|
100
|
+
|
101
|
+
##
|
102
|
+
# What version of ruby to parse. 18 and 19 are the only valid values
|
103
|
+
# currently supported.
|
104
|
+
|
105
|
+
attr_accessor :version
|
106
|
+
|
107
|
+
# Value of last token which had a value associated with it.
|
108
|
+
attr_accessor :yacc_value
|
109
|
+
|
110
|
+
attr_writer :lineno # reader is lazy initalizer
|
111
|
+
|
112
|
+
attr_writer :comments
|
113
|
+
|
114
|
+
def initialize v = 18
|
115
|
+
self.version = v
|
116
|
+
|
117
|
+
reset
|
118
|
+
end
|
119
|
+
|
120
|
+
##
|
95
121
|
# How the parser advances to the next token.
|
96
122
|
#
|
97
123
|
# @return true if not at end of file (EOF).
|
@@ -100,7 +126,7 @@ class RubyLexer
|
|
100
126
|
r = yylex
|
101
127
|
self.token = r
|
102
128
|
|
103
|
-
raise "yylex returned nil" unless r
|
129
|
+
raise "yylex returned nil, near #{ss.rest[0,10].inspect}" unless r
|
104
130
|
|
105
131
|
return RubyLexer::EOF != r
|
106
132
|
end
|
@@ -109,28 +135,35 @@ class RubyLexer
|
|
109
135
|
self.warning("Ambiguous first argument. make sure.")
|
110
136
|
end
|
111
137
|
|
112
|
-
def
|
138
|
+
def arg_state
|
139
|
+
in_arg_state? ? :expr_arg : :expr_beg
|
140
|
+
end
|
141
|
+
|
142
|
+
def beginning_of_line?
|
143
|
+
ss.bol?
|
144
|
+
end
|
145
|
+
|
146
|
+
def check re
|
147
|
+
ss.check re
|
148
|
+
end
|
149
|
+
|
150
|
+
def comments # TODO: remove this... maybe comment_string + attr_accessor
|
113
151
|
c = @comments.join
|
114
152
|
@comments.clear
|
115
153
|
c
|
116
154
|
end
|
117
155
|
|
118
|
-
def
|
119
|
-
|
120
|
-
cmdarg.push false
|
121
|
-
self.lex_state = :expr_beg
|
122
|
-
self.yacc_value = val
|
156
|
+
def end_of_stream?
|
157
|
+
ss.eos?
|
123
158
|
end
|
124
159
|
|
125
|
-
def
|
126
|
-
|
127
|
-
|
128
|
-
|
129
|
-
:expr_beg
|
130
|
-
end
|
160
|
+
def expr_result token, text
|
161
|
+
cond.push false
|
162
|
+
cmdarg.push false
|
163
|
+
result :expr_beg, token, text
|
131
164
|
end
|
132
165
|
|
133
|
-
def heredoc here #
|
166
|
+
def heredoc here # TODO: rewrite / remove
|
134
167
|
_, eos, func, last_line = here
|
135
168
|
|
136
169
|
indent = (func & STR_FUNC_INDENT) != 0 ? "[ \t]*" : nil
|
@@ -138,11 +171,10 @@ class RubyLexer
|
|
138
171
|
eos_re = /#{indent}#{Regexp.escape eos}(\r*\n|\z)/
|
139
172
|
err_msg = "can't match #{eos_re.inspect} anywhere in "
|
140
173
|
|
141
|
-
rb_compile_error err_msg if
|
142
|
-
src.eos?
|
174
|
+
rb_compile_error err_msg if end_of_stream?
|
143
175
|
|
144
|
-
if
|
145
|
-
|
176
|
+
if beginning_of_line? && scan(eos_re) then
|
177
|
+
ss.unread_many last_line # TODO: figure out how to remove this
|
146
178
|
self.yacc_value = eos
|
147
179
|
return :tSTRING_END
|
148
180
|
end
|
@@ -151,14 +183,14 @@ class RubyLexer
|
|
151
183
|
|
152
184
|
if expand then
|
153
185
|
case
|
154
|
-
when
|
155
|
-
|
156
|
-
self.yacc_value =
|
186
|
+
when scan(/#[$@]/) then
|
187
|
+
ss.pos -= 1 # FIX omg stupid
|
188
|
+
self.yacc_value = matched
|
157
189
|
return :tSTRING_DVAR
|
158
|
-
when
|
159
|
-
self.yacc_value =
|
190
|
+
when scan(/#[{]/) then
|
191
|
+
self.yacc_value = matched
|
160
192
|
return :tSTRING_DBEG
|
161
|
-
when
|
193
|
+
when scan(/#/) then
|
162
194
|
string_buffer << '#'
|
163
195
|
end
|
164
196
|
|
@@ -172,34 +204,32 @@ class RubyLexer
|
|
172
204
|
self.yacc_value = string_buffer.join.delete("\r")
|
173
205
|
return :tSTRING_CONTENT
|
174
206
|
else
|
175
|
-
string_buffer <<
|
207
|
+
string_buffer << scan(/\n/)
|
176
208
|
end
|
177
209
|
|
178
|
-
rb_compile_error err_msg if
|
179
|
-
|
180
|
-
end until src.check(eos_re)
|
210
|
+
rb_compile_error err_msg if end_of_stream?
|
211
|
+
end until check(eos_re)
|
181
212
|
else
|
182
|
-
until
|
183
|
-
string_buffer <<
|
184
|
-
rb_compile_error err_msg if
|
185
|
-
src.eos?
|
213
|
+
until check(eos_re) do
|
214
|
+
string_buffer << scan(/.*(\n|\z)/)
|
215
|
+
rb_compile_error err_msg if end_of_stream?
|
186
216
|
end
|
187
217
|
end
|
188
218
|
|
189
219
|
self.lex_strterm = [:heredoc, eos, func, last_line]
|
190
|
-
self.yacc_value = string_buffer.join.delete("\r")
|
191
220
|
|
221
|
+
self.yacc_value = string_buffer.join.delete("\r")
|
192
222
|
return :tSTRING_CONTENT
|
193
223
|
end
|
194
224
|
|
195
|
-
def heredoc_identifier #
|
225
|
+
def heredoc_identifier # TODO: remove / rewrite
|
196
226
|
term, func = nil, STR_FUNC_BORING
|
197
227
|
self.string_buffer = []
|
198
228
|
|
199
229
|
case
|
200
|
-
when
|
201
|
-
term =
|
202
|
-
func |= STR_FUNC_INDENT unless
|
230
|
+
when scan(/(-?)([\'\"\`])(.*?)\2/) then
|
231
|
+
term = ss[2]
|
232
|
+
func |= STR_FUNC_INDENT unless ss[1].empty?
|
203
233
|
func |= case term
|
204
234
|
when "\'" then
|
205
235
|
STR_SQUOTE
|
@@ -208,24 +238,24 @@ class RubyLexer
|
|
208
238
|
else
|
209
239
|
STR_XQUOTE
|
210
240
|
end
|
211
|
-
string_buffer <<
|
212
|
-
when
|
241
|
+
string_buffer << ss[3]
|
242
|
+
when scan(/-?([\'\"\`])(?!\1*\Z)/) then
|
213
243
|
rb_compile_error "unterminated here document identifier"
|
214
|
-
when
|
244
|
+
when scan(/(-?)(#{IDENT_CHAR}+)/) then
|
215
245
|
term = '"'
|
216
246
|
func |= STR_DQUOTE
|
217
|
-
unless
|
247
|
+
unless ss[1].empty? then
|
218
248
|
func |= STR_FUNC_INDENT
|
219
249
|
end
|
220
|
-
string_buffer <<
|
250
|
+
string_buffer << ss[2]
|
221
251
|
else
|
222
252
|
return nil
|
223
253
|
end
|
224
254
|
|
225
|
-
if
|
255
|
+
if scan(/.*\n/) then
|
226
256
|
# TODO: think about storing off the char range instead
|
227
|
-
line =
|
228
|
-
|
257
|
+
line = matched
|
258
|
+
ss.extra_lines_added += 1 # FIX: ugh
|
229
259
|
else
|
230
260
|
line = nil
|
231
261
|
end
|
@@ -241,41 +271,47 @@ class RubyLexer
|
|
241
271
|
end
|
242
272
|
end
|
243
273
|
|
274
|
+
def in_arg_state? # TODO: rename is_after_operator?
|
275
|
+
in_lex_state? :expr_fname, :expr_dot
|
276
|
+
end
|
277
|
+
|
244
278
|
def in_lex_state?(*states)
|
245
279
|
states.include? lex_state
|
246
280
|
end
|
247
281
|
|
248
|
-
def
|
249
|
-
|
250
|
-
self.cond = RubyParserStuff::StackState.new(:cond)
|
251
|
-
self.cmdarg = RubyParserStuff::StackState.new(:cmdarg)
|
252
|
-
self.tern = RubyParserStuff::StackState.new(:tern)
|
253
|
-
self.string_nest = 0
|
254
|
-
self.paren_nest = 0
|
255
|
-
self.brace_nest = 0
|
256
|
-
self.lpar_beg = nil
|
282
|
+
def int_with_base base
|
283
|
+
rb_compile_error "Invalid numeric format" if matched =~ /__/
|
257
284
|
|
258
|
-
|
285
|
+
self.yacc_value = matched.to_i(base)
|
286
|
+
return :tINTEGER
|
287
|
+
end
|
259
288
|
|
260
|
-
|
289
|
+
def is_arg?
|
290
|
+
in_lex_state? :expr_arg, :expr_cmdarg
|
261
291
|
end
|
262
292
|
|
263
|
-
def
|
264
|
-
|
293
|
+
def is_beg?
|
294
|
+
in_lex_state? :expr_beg, :expr_value, :expr_mid, :expr_class
|
295
|
+
end
|
265
296
|
|
266
|
-
|
267
|
-
|
297
|
+
def is_end?
|
298
|
+
in_lex_state? :expr_end, :expr_endarg, :expr_endfn
|
268
299
|
end
|
269
300
|
|
270
|
-
def
|
271
|
-
|
272
|
-
|
273
|
-
|
301
|
+
def is_label_possible? command_state
|
302
|
+
(in_lex_state?(:expr_beg, :expr_endfn) && !command_state) || is_arg?
|
303
|
+
end
|
304
|
+
|
305
|
+
def is_space_arg? c = "x"
|
306
|
+
is_arg? and space_seen and c !~ /\s/
|
274
307
|
end
|
275
308
|
|
276
|
-
attr_writer :lineno
|
277
309
|
def lineno
|
278
|
-
@lineno ||=
|
310
|
+
@lineno ||= ss.lineno
|
311
|
+
end
|
312
|
+
|
313
|
+
def matched
|
314
|
+
ss.matched
|
279
315
|
end
|
280
316
|
|
281
317
|
##
|
@@ -288,45 +324,45 @@ class RubyLexer
|
|
288
324
|
self.lex_state = :expr_end
|
289
325
|
|
290
326
|
case
|
291
|
-
when
|
327
|
+
when scan(/[+-]?0[xXbBdD]\b/) then
|
292
328
|
rb_compile_error "Invalid numeric format"
|
293
|
-
when
|
329
|
+
when scan(/[+-]?(?:(?:[1-9][\d_]*|0)(?!\.\d)\b|0[Dd][0-9_]+)/) then
|
294
330
|
int_with_base(10)
|
295
|
-
when
|
331
|
+
when scan(/[+-]?0x[a-f0-9_]+/i) then
|
296
332
|
int_with_base(16)
|
297
|
-
when
|
333
|
+
when scan(/[+-]?0[Bb][01_]+/) then
|
298
334
|
int_with_base(2)
|
299
|
-
when
|
335
|
+
when scan(/[+-]?0[Oo]?[0-7_]*[89]/) then
|
300
336
|
rb_compile_error "Illegal octal digit."
|
301
|
-
when
|
337
|
+
when scan(/[+-]?0[Oo]?[0-7_]+|0[Oo]/) then
|
302
338
|
int_with_base(8)
|
303
|
-
when
|
339
|
+
when scan(/[+-]?[\d_]+_(e|\.)/) then
|
304
340
|
rb_compile_error "Trailing '_' in number."
|
305
|
-
when
|
306
|
-
number =
|
341
|
+
when scan(/[+-]?[\d_]+\.[\d_]+(e[+-]?[\d_]+)?\b|[+-]?[\d_]+e[+-]?[\d_]+\b/i) then
|
342
|
+
number = matched
|
307
343
|
if number =~ /__/ then
|
308
344
|
rb_compile_error "Invalid numeric format"
|
309
345
|
end
|
310
346
|
self.yacc_value = number.to_f
|
311
347
|
:tFLOAT
|
312
|
-
when
|
348
|
+
when scan(/[+-]?[0-9_]+(?![e])/) then
|
313
349
|
int_with_base(10)
|
314
350
|
else
|
315
351
|
rb_compile_error "Bad number format"
|
316
352
|
end
|
317
353
|
end
|
318
354
|
|
319
|
-
def parse_quote #
|
355
|
+
def parse_quote # TODO: remove / rewrite
|
320
356
|
beg, nnd, short_hand, c = nil, nil, false, nil
|
321
357
|
|
322
|
-
if
|
323
|
-
rb_compile_error "unknown type of %string" if
|
324
|
-
c, beg, short_hand =
|
358
|
+
if scan(/[a-z0-9]{1,2}/i) then # Long-hand (e.g. %Q{}).
|
359
|
+
rb_compile_error "unknown type of %string" if ss.matched_size == 2
|
360
|
+
c, beg, short_hand = matched, ss.getch, false
|
325
361
|
else # Short-hand (e.g. %{, %., %!, etc)
|
326
|
-
c, beg, short_hand = 'Q',
|
362
|
+
c, beg, short_hand = 'Q', ss.getch, true
|
327
363
|
end
|
328
364
|
|
329
|
-
if
|
365
|
+
if end_of_stream? or c == RubyLexer::EOF or beg == RubyLexer::EOF then
|
330
366
|
rb_compile_error "unterminated quoted string meets end of file"
|
331
367
|
end
|
332
368
|
|
@@ -334,19 +370,19 @@ class RubyLexer
|
|
334
370
|
nnd = { "(" => ")", "[" => "]", "{" => "}", "<" => ">" }[beg]
|
335
371
|
nnd, beg = beg, "\0" if nnd.nil?
|
336
372
|
|
337
|
-
token_type,
|
373
|
+
token_type, text = nil, "%#{c}#{beg}"
|
338
374
|
token_type, string_type = case c
|
339
375
|
when 'Q' then
|
340
376
|
ch = short_hand ? nnd : c + beg
|
341
|
-
|
377
|
+
text = "%#{ch}"
|
342
378
|
[:tSTRING_BEG, STR_DQUOTE]
|
343
379
|
when 'q' then
|
344
380
|
[:tSTRING_BEG, STR_SQUOTE]
|
345
381
|
when 'W' then
|
346
|
-
|
382
|
+
scan(/\s*/)
|
347
383
|
[:tWORDS_BEG, STR_DQUOTE | STR_FUNC_QWORDS]
|
348
384
|
when 'w' then
|
349
|
-
|
385
|
+
scan(/\s*/)
|
350
386
|
[:tQWORDS_BEG, STR_SQUOTE | STR_FUNC_QWORDS]
|
351
387
|
when 'x' then
|
352
388
|
[:tXSTRING_BEG, STR_XQUOTE]
|
@@ -356,20 +392,25 @@ class RubyLexer
|
|
356
392
|
self.lex_state = :expr_fname
|
357
393
|
[:tSYMBEG, STR_SSYM]
|
358
394
|
when 'I' then
|
395
|
+
src.scan(/\s*/)
|
359
396
|
[:tSYMBOLS_BEG, STR_DQUOTE | STR_FUNC_QWORDS]
|
360
397
|
when 'i' then
|
398
|
+
src.scan(/\s*/)
|
361
399
|
[:tQSYMBOLS_BEG, STR_SQUOTE | STR_FUNC_QWORDS]
|
362
400
|
end
|
363
401
|
|
364
|
-
rb_compile_error "Bad %string type. Expected [
|
402
|
+
rb_compile_error "Bad %string type. Expected [QqWwIixrs], found '#{c}'." if
|
365
403
|
token_type.nil?
|
366
404
|
|
367
|
-
|
405
|
+
raise "huh" unless string_type
|
406
|
+
|
407
|
+
string string_type, nnd, beg
|
368
408
|
|
409
|
+
self.yacc_value = text
|
369
410
|
return token_type
|
370
411
|
end
|
371
412
|
|
372
|
-
def parse_string
|
413
|
+
def parse_string quote # TODO: rewrite / remove
|
373
414
|
_, string_type, term, open = quote
|
374
415
|
|
375
416
|
space = false # FIX: remove these
|
@@ -381,41 +422,39 @@ class RubyLexer
|
|
381
422
|
regexp = (func & STR_FUNC_REGEXP) != 0
|
382
423
|
expand = (func & STR_FUNC_EXPAND) != 0
|
383
424
|
|
384
|
-
unless func then #
|
425
|
+
unless func then # nil'ed from qwords below. *sigh*
|
385
426
|
self.lineno = nil
|
386
427
|
return :tSTRING_END
|
387
428
|
end
|
388
429
|
|
389
|
-
space = true if qwords and
|
430
|
+
space = true if qwords and scan(/\s+/)
|
390
431
|
|
391
|
-
if self.string_nest == 0 &&
|
432
|
+
if self.string_nest == 0 && scan(/#{term_re}/) then
|
392
433
|
if qwords then
|
393
|
-
quote[1] = nil
|
434
|
+
quote[1] = nil
|
394
435
|
return :tSPACE
|
395
436
|
elsif regexp then
|
396
|
-
self.yacc_value = self.regx_options
|
397
437
|
self.lineno = nil
|
438
|
+
self.yacc_value = self.regx_options
|
398
439
|
return :tREGEXP_END
|
399
440
|
else
|
400
|
-
self.yacc_value = term
|
401
441
|
self.lineno = nil
|
442
|
+
self.yacc_value = term
|
402
443
|
return :tSTRING_END
|
403
444
|
end
|
404
445
|
end
|
405
446
|
|
406
|
-
if space
|
407
|
-
return :tSPACE
|
408
|
-
end
|
447
|
+
return :tSPACE if space
|
409
448
|
|
410
449
|
self.string_buffer = []
|
411
450
|
|
412
451
|
if expand
|
413
452
|
case
|
414
|
-
when
|
453
|
+
when scan(/#(?=[$@])/) then
|
415
454
|
return :tSTRING_DVAR
|
416
|
-
when
|
455
|
+
when scan(/#[{]/) then
|
417
456
|
return :tSTRING_DBEG
|
418
|
-
when
|
457
|
+
when scan(/#/) then
|
419
458
|
string_buffer << '#'
|
420
459
|
end
|
421
460
|
end
|
@@ -429,71 +468,159 @@ class RubyLexer
|
|
429
468
|
return :tSTRING_CONTENT
|
430
469
|
end
|
431
470
|
|
471
|
+
def process_token command_state, last_state
|
472
|
+
token = self.token
|
473
|
+
token << matched if scan(/[\!\?](?!=)/)
|
474
|
+
|
475
|
+
tok_id =
|
476
|
+
case
|
477
|
+
when token =~ /[!?]$/ then
|
478
|
+
:tFID
|
479
|
+
when in_lex_state?(:expr_fname) && scan(/=(?:(?![~>=])|(?==>))/) then
|
480
|
+
# ident=, not =~ => == or followed by =>
|
481
|
+
# TODO test lexing of a=>b vs a==>b
|
482
|
+
token << matched
|
483
|
+
:tIDENTIFIER
|
484
|
+
when token =~ /^[A-Z]/ then
|
485
|
+
:tCONSTANT
|
486
|
+
else
|
487
|
+
:tIDENTIFIER
|
488
|
+
end
|
489
|
+
|
490
|
+
if !ruby18 and is_label_possible?(command_state) and scan(/:(?!:)/) then
|
491
|
+
return result(:expr_beg, :tLABEL, [token, ss.lineno]) # HACK: array? TODO: self.lineno
|
492
|
+
end
|
493
|
+
|
494
|
+
unless in_lex_state? :expr_dot then
|
495
|
+
# See if it is a reserved word.
|
496
|
+
keyword = if ruby18 then # REFACTOR need 18/19 lexer subclasses
|
497
|
+
RubyParserStuff::Keyword.keyword18 token
|
498
|
+
else
|
499
|
+
RubyParserStuff::Keyword.keyword19 token
|
500
|
+
end
|
501
|
+
|
502
|
+
return process_token_keyword keyword if keyword
|
503
|
+
end # unless in_lex_state? :expr_dot
|
504
|
+
|
505
|
+
# TODO:
|
506
|
+
# if (mb == ENC_CODERANGE_7BIT && lex_state != EXPR_DOT) {
|
507
|
+
|
508
|
+
state = if is_beg? or is_arg? or in_lex_state? :expr_dot then
|
509
|
+
command_state ? :expr_cmdarg : :expr_arg
|
510
|
+
elsif not ruby18 and in_lex_state? :expr_fname then
|
511
|
+
:expr_endfn
|
512
|
+
else
|
513
|
+
:expr_end
|
514
|
+
end
|
515
|
+
|
516
|
+
if not [:expr_dot, :expr_fname].include? last_state and
|
517
|
+
self.parser.env[token.to_sym] == :lvar then
|
518
|
+
state = :expr_end
|
519
|
+
end
|
520
|
+
|
521
|
+
return result(state, tok_id, token)
|
522
|
+
end
|
523
|
+
|
524
|
+
def process_token_keyword keyword
|
525
|
+
state = keyword.state
|
526
|
+
value = [token, ss.lineno] # TODO: use self.lineno ?
|
527
|
+
|
528
|
+
self.command_start = true if state == :expr_beg and lex_state != :expr_fname
|
529
|
+
|
530
|
+
case
|
531
|
+
when lex_state == :expr_fname then
|
532
|
+
result(state, keyword.id0, keyword.name)
|
533
|
+
when keyword.id0 == :kDO then
|
534
|
+
case
|
535
|
+
when lpar_beg && lpar_beg == paren_nest then
|
536
|
+
self.lpar_beg = nil
|
537
|
+
self.paren_nest -= 1
|
538
|
+
result(state, :kDO_LAMBDA, value)
|
539
|
+
when cond.is_in_state then
|
540
|
+
result(state, :kDO_COND, value)
|
541
|
+
when cmdarg.is_in_state && lex_state != :expr_cmdarg then
|
542
|
+
result(state, :kDO_BLOCK, value)
|
543
|
+
when in_lex_state?(:expr_beg, :expr_endarg) then
|
544
|
+
result(state, :kDO_BLOCK, value)
|
545
|
+
else
|
546
|
+
result(state, :kDO, value)
|
547
|
+
end
|
548
|
+
when in_lex_state?(:expr_beg, :expr_value) then
|
549
|
+
result(state, keyword.id0, value)
|
550
|
+
when keyword.id0 != keyword.id1 then
|
551
|
+
result(:expr_beg, keyword.id1, value)
|
552
|
+
else
|
553
|
+
result(state, keyword.id1, value)
|
554
|
+
end
|
555
|
+
end
|
556
|
+
|
432
557
|
def rb_compile_error msg
|
433
|
-
msg += ". near line #{self.lineno}: #{
|
558
|
+
msg += ". near line #{self.lineno}: #{ss.rest[/^.*/].inspect}"
|
434
559
|
raise RubyParser::SyntaxError, msg
|
435
560
|
end
|
436
561
|
|
437
|
-
def read_escape #
|
562
|
+
def read_escape # TODO: remove / rewrite
|
438
563
|
case
|
439
|
-
when
|
564
|
+
when scan(/\\/) then # Backslash
|
440
565
|
'\\'
|
441
|
-
when
|
566
|
+
when scan(/n/) then # newline
|
442
567
|
"\n"
|
443
|
-
when
|
568
|
+
when scan(/t/) then # horizontal tab
|
444
569
|
"\t"
|
445
|
-
when
|
570
|
+
when scan(/r/) then # carriage-return
|
446
571
|
"\r"
|
447
|
-
when
|
572
|
+
when scan(/f/) then # form-feed
|
448
573
|
"\f"
|
449
|
-
when
|
574
|
+
when scan(/v/) then # vertical tab
|
450
575
|
"\13"
|
451
|
-
when
|
576
|
+
when scan(/a/) then # alarm(bell)
|
452
577
|
"\007"
|
453
|
-
when
|
578
|
+
when scan(/e/) then # escape
|
454
579
|
"\033"
|
455
|
-
when
|
580
|
+
when scan(/b/) then # backspace
|
456
581
|
"\010"
|
457
|
-
when
|
582
|
+
when scan(/s/) then # space
|
458
583
|
" "
|
459
|
-
when
|
460
|
-
(
|
461
|
-
when
|
462
|
-
|
463
|
-
when
|
464
|
-
|
584
|
+
when scan(/[0-7]{1,3}/) then # octal constant
|
585
|
+
(matched.to_i(8) & 0xFF).chr
|
586
|
+
when scan(/x([0-9a-fA-F]{1,2})/) then # hex constant
|
587
|
+
ss[1].to_i(16).chr
|
588
|
+
when check(/M-\\[\\MCc]/) then
|
589
|
+
scan(/M-\\/) # eat it
|
465
590
|
c = self.read_escape
|
466
591
|
c[0] = (c[0].ord | 0x80).chr
|
467
592
|
c
|
468
|
-
when
|
469
|
-
c =
|
593
|
+
when scan(/M-(.)/) then
|
594
|
+
c = ss[1]
|
470
595
|
c[0] = (c[0].ord | 0x80).chr
|
471
596
|
c
|
472
|
-
when
|
473
|
-
|
597
|
+
when check(/(C-|c)\\[\\MCc]/) then
|
598
|
+
scan(/(C-|c)\\/) # eat it
|
474
599
|
c = self.read_escape
|
475
600
|
c[0] = (c[0].ord & 0x9f).chr
|
476
601
|
c
|
477
|
-
when
|
602
|
+
when scan(/C-\?|c\?/) then
|
478
603
|
127.chr
|
479
|
-
when
|
480
|
-
c =
|
604
|
+
when scan(/(C-|c)(.)/) then
|
605
|
+
c = ss[2]
|
481
606
|
c[0] = (c[0].ord & 0x9f).chr
|
482
607
|
c
|
483
|
-
when
|
484
|
-
|
485
|
-
when
|
608
|
+
when scan(/^[89]/i) then # bad octal or hex... MRI ignores them :(
|
609
|
+
matched
|
610
|
+
when scan(/u([0-9a-fA-F]+|\{[0-9a-fA-F]+\})/) then
|
611
|
+
[ss[1].delete("{}").to_i(16)].pack("U")
|
612
|
+
when scan(/[McCx0-9]/) || end_of_stream? then
|
486
613
|
rb_compile_error("Invalid escape character syntax")
|
487
614
|
else
|
488
|
-
|
615
|
+
ss.getch
|
489
616
|
end
|
490
617
|
end
|
491
618
|
|
492
|
-
def regx_options #
|
619
|
+
def regx_options # TODO: rewrite / remove
|
493
620
|
good, bad = [], []
|
494
621
|
|
495
|
-
if
|
496
|
-
good, bad =
|
622
|
+
if scan(/[a-z]+/) then
|
623
|
+
good, bad = matched.split(//).partition { |s| s =~ /^[ixmonesu]$/ }
|
497
624
|
end
|
498
625
|
|
499
626
|
unless bad.empty? then
|
@@ -505,13 +632,30 @@ class RubyLexer
|
|
505
632
|
end
|
506
633
|
|
507
634
|
def reset
|
635
|
+
self.brace_nest = 0
|
508
636
|
self.command_start = true
|
637
|
+
self.comments = []
|
638
|
+
self.lex_state = nil
|
509
639
|
self.lex_strterm = nil
|
640
|
+
self.lineno = 1
|
641
|
+
self.lpar_beg = nil
|
642
|
+
self.paren_nest = 0
|
643
|
+
self.space_seen = false
|
644
|
+
self.string_nest = 0
|
510
645
|
self.token = nil
|
511
646
|
self.yacc_value = nil
|
512
647
|
|
513
|
-
|
514
|
-
|
648
|
+
self.cmdarg = RubyParserStuff::StackState.new(:cmdarg)
|
649
|
+
self.cond = RubyParserStuff::StackState.new(:cond)
|
650
|
+
|
651
|
+
@src = nil
|
652
|
+
end
|
653
|
+
|
654
|
+
def result lex_state, token, text # :nodoc:
|
655
|
+
lex_state = self.arg_state if lex_state == :arg_state
|
656
|
+
self.lex_state = lex_state if lex_state
|
657
|
+
self.yacc_value = text
|
658
|
+
token
|
515
659
|
end
|
516
660
|
|
517
661
|
def ruby18
|
@@ -522,35 +666,52 @@ class RubyLexer
|
|
522
666
|
Ruby19Parser === parser
|
523
667
|
end
|
524
668
|
|
669
|
+
def scan re
|
670
|
+
ss.scan re
|
671
|
+
end
|
672
|
+
|
673
|
+
def space_vs_beginning space_type, beg_type, fallback
|
674
|
+
if is_space_arg? check(/./m) then
|
675
|
+
warning "`**' interpreted as argument prefix"
|
676
|
+
space_type
|
677
|
+
elsif is_beg? then
|
678
|
+
beg_type
|
679
|
+
else
|
680
|
+
# TODO: warn_balanced("**", "argument prefix");
|
681
|
+
fallback
|
682
|
+
end
|
683
|
+
end
|
684
|
+
|
685
|
+
def string type, beg = matched, nnd = "\0"
|
686
|
+
self.lex_strterm = [:strterm, type, beg, nnd]
|
687
|
+
end
|
688
|
+
|
525
689
|
def src= src
|
526
690
|
raise "bad src: #{src.inspect}" unless String === src
|
527
691
|
@src = RPStringScanner.new(src)
|
528
692
|
end
|
529
693
|
|
530
|
-
def tokadd_escape term #
|
694
|
+
def tokadd_escape term # TODO: rewrite / remove
|
531
695
|
case
|
532
|
-
when
|
696
|
+
when scan(/\\\n/) then
|
533
697
|
# just ignore
|
534
|
-
when
|
535
|
-
self.string_buffer <<
|
536
|
-
when
|
537
|
-
self.string_buffer <<
|
698
|
+
when scan(/\\([0-7]{1,3}|x[0-9a-fA-F]{1,2})/) then
|
699
|
+
self.string_buffer << matched
|
700
|
+
when scan(/\\([MC]-|c)(?=\\)/) then
|
701
|
+
self.string_buffer << matched
|
538
702
|
self.tokadd_escape term
|
539
|
-
when
|
540
|
-
self.string_buffer <<
|
541
|
-
when
|
703
|
+
when scan(/\\([MC]-|c)(.)/) then
|
704
|
+
self.string_buffer << matched
|
705
|
+
when scan(/\\[McCx]/) then
|
542
706
|
rb_compile_error "Invalid escape character syntax"
|
543
|
-
when
|
544
|
-
self.string_buffer <<
|
707
|
+
when scan(/\\(.)/m) then
|
708
|
+
self.string_buffer << matched
|
545
709
|
else
|
546
710
|
rb_compile_error "Invalid escape character syntax"
|
547
711
|
end
|
548
712
|
end
|
549
713
|
|
550
|
-
|
551
|
-
@@regexp_cache[nil] = nil
|
552
|
-
|
553
|
-
def tokadd_string(func, term, paren) # 105 lines
|
714
|
+
def tokadd_string(func, term, paren) # TODO: rewrite / remove
|
554
715
|
qwords = (func & STR_FUNC_QWORDS) != 0
|
555
716
|
escape = (func & STR_FUNC_ESCAPE) != 0
|
556
717
|
expand = (func & STR_FUNC_EXPAND) != 0
|
@@ -560,49 +721,49 @@ class RubyLexer
|
|
560
721
|
paren_re = @@regexp_cache[paren]
|
561
722
|
term_re = @@regexp_cache[term]
|
562
723
|
|
563
|
-
until
|
724
|
+
until end_of_stream? do
|
564
725
|
c = nil
|
565
726
|
handled = true
|
566
727
|
|
567
728
|
case
|
568
|
-
when paren_re &&
|
729
|
+
when paren_re && scan(paren_re) then
|
569
730
|
self.string_nest += 1
|
570
|
-
when
|
731
|
+
when scan(term_re) then
|
571
732
|
if self.string_nest == 0 then
|
572
|
-
|
733
|
+
ss.pos -= 1
|
573
734
|
break
|
574
735
|
else
|
575
736
|
self.string_nest -= 1
|
576
737
|
end
|
577
|
-
when expand &&
|
578
|
-
|
738
|
+
when expand && scan(/#(?=[\$\@\{])/) then
|
739
|
+
ss.pos -= 1
|
579
740
|
break
|
580
|
-
when qwords &&
|
581
|
-
|
741
|
+
when qwords && scan(/\s/) then
|
742
|
+
ss.pos -= 1
|
582
743
|
break
|
583
|
-
when expand &&
|
744
|
+
when expand && scan(/#(?!\n)/) then
|
584
745
|
# do nothing
|
585
|
-
when
|
746
|
+
when check(/\\/) then
|
586
747
|
case
|
587
|
-
when qwords &&
|
748
|
+
when qwords && scan(/\\\n/) then
|
588
749
|
string_buffer << "\n"
|
589
750
|
next
|
590
|
-
when qwords &&
|
751
|
+
when qwords && scan(/\\\s/) then
|
591
752
|
c = ' '
|
592
|
-
when expand &&
|
753
|
+
when expand && scan(/\\\n/) then
|
593
754
|
next
|
594
|
-
when regexp &&
|
755
|
+
when regexp && check(/\\/) then
|
595
756
|
self.tokadd_escape term
|
596
757
|
next
|
597
|
-
when expand &&
|
758
|
+
when expand && scan(/\\/) then
|
598
759
|
c = self.read_escape
|
599
|
-
when
|
760
|
+
when scan(/\\\n/) then
|
600
761
|
# do nothing
|
601
|
-
when
|
762
|
+
when scan(/\\\\/) then
|
602
763
|
string_buffer << '\\' if escape
|
603
764
|
c = '\\'
|
604
|
-
when
|
605
|
-
unless
|
765
|
+
when scan(/\\/) then
|
766
|
+
unless scan(term_re) || paren.nil? || scan(paren_re) then
|
606
767
|
string_buffer << "\\"
|
607
768
|
end
|
608
769
|
else
|
@@ -625,38 +786,22 @@ class RubyLexer
|
|
625
786
|
/[^#{t}#{x}\#\0\\]+|./
|
626
787
|
end
|
627
788
|
|
628
|
-
|
629
|
-
c =
|
789
|
+
scan re
|
790
|
+
c = matched
|
630
791
|
|
631
792
|
rb_compile_error "symbol cannot contain '\\0'" if symbol && c =~ /\0/
|
632
793
|
end # unless handled
|
633
794
|
|
634
|
-
c ||=
|
795
|
+
c ||= matched
|
635
796
|
string_buffer << c
|
636
797
|
end # until
|
637
798
|
|
638
|
-
c ||=
|
639
|
-
c = RubyLexer::EOF if
|
799
|
+
c ||= matched
|
800
|
+
c = RubyLexer::EOF if end_of_stream?
|
640
801
|
|
641
802
|
return c
|
642
803
|
end
|
643
804
|
|
644
|
-
ESCAPES = {
|
645
|
-
"a" => "\007",
|
646
|
-
"b" => "\010",
|
647
|
-
"e" => "\033",
|
648
|
-
"f" => "\f",
|
649
|
-
"n" => "\n",
|
650
|
-
"r" => "\r",
|
651
|
-
"s" => " ",
|
652
|
-
"t" => "\t",
|
653
|
-
"v" => "\13",
|
654
|
-
"\\" => '\\',
|
655
|
-
"\n" => "",
|
656
|
-
"C-\?" => 127.chr,
|
657
|
-
"c\?" => 127.chr,
|
658
|
-
}
|
659
|
-
|
660
805
|
def unescape s
|
661
806
|
r = ESCAPES[s]
|
662
807
|
|
@@ -675,6 +820,8 @@ class RubyLexer
|
|
675
820
|
s
|
676
821
|
when /^[McCx0-9]/ then
|
677
822
|
rb_compile_error("Invalid escape character syntax")
|
823
|
+
when /u([0-9a-fA-F]+|\{[0-9a-fA-F]+\})/ then
|
824
|
+
[$1.delete("{}").to_i(16)].pack("U")
|
678
825
|
else
|
679
826
|
s
|
680
827
|
end
|
@@ -691,11 +838,11 @@ class RubyLexer
|
|
691
838
|
#
|
692
839
|
# @return Description of the Returned Value
|
693
840
|
|
694
|
-
def yylex #
|
841
|
+
def yylex # 461 lines
|
695
842
|
c = ''
|
696
843
|
self.space_seen = false
|
697
844
|
command_state = false
|
698
|
-
|
845
|
+
ss = self.src
|
699
846
|
|
700
847
|
self.token = nil
|
701
848
|
self.yacc_value = nil
|
@@ -708,41 +855,42 @@ class RubyLexer
|
|
708
855
|
last_state = lex_state
|
709
856
|
|
710
857
|
loop do # START OF CASE
|
711
|
-
if
|
858
|
+
if scan(/[\ \t\r\f\v]/) then # \s - \n + \v
|
712
859
|
self.space_seen = true
|
713
860
|
next
|
714
|
-
elsif
|
715
|
-
if
|
861
|
+
elsif check(/[^a-zA-Z]/) then
|
862
|
+
if scan(/\n|\#/) then
|
716
863
|
self.lineno = nil
|
717
|
-
c =
|
864
|
+
c = matched
|
718
865
|
if c == '#' then
|
719
|
-
|
866
|
+
ss.pos -= 1
|
720
867
|
|
721
|
-
while
|
722
|
-
|
868
|
+
while scan(/\s*#.*(\n+|\z)/) do
|
869
|
+
# TODO: self.lineno += matched.lines.to_a.size
|
870
|
+
@comments << matched.gsub(/^ +#/, '#').gsub(/^ +$/, '')
|
723
871
|
end
|
724
872
|
|
725
|
-
return RubyLexer::EOF if
|
873
|
+
return RubyLexer::EOF if end_of_stream?
|
726
874
|
end
|
727
875
|
|
728
876
|
# Replace a string of newlines with a single one
|
729
|
-
|
877
|
+
scan(/\n+/)
|
730
878
|
|
731
879
|
next if in_lex_state?(:expr_beg, :expr_value, :expr_class,
|
732
880
|
:expr_fname, :expr_dot)
|
733
881
|
|
734
|
-
if
|
735
|
-
self.space_seen = true unless
|
882
|
+
if scan(/([\ \t\r\f\v]*)\./) then
|
883
|
+
self.space_seen = true unless ss[1].empty?
|
736
884
|
|
737
|
-
|
738
|
-
next unless
|
885
|
+
ss.pos -= 1
|
886
|
+
next unless check(/\.\./)
|
739
887
|
end
|
740
888
|
|
741
889
|
self.command_start = true
|
742
|
-
|
743
|
-
return :tNL
|
744
|
-
elsif
|
745
|
-
if
|
890
|
+
|
891
|
+
return result(:expr_beg, :tNL, nil)
|
892
|
+
elsif scan(/[\]\)\}]/) then
|
893
|
+
if matched == "}" then
|
746
894
|
self.brace_nest -= 1
|
747
895
|
else
|
748
896
|
self.paren_nest -= 1
|
@@ -750,54 +898,34 @@ class RubyLexer
|
|
750
898
|
|
751
899
|
cond.lexpop
|
752
900
|
cmdarg.lexpop
|
753
|
-
tern.lexpop
|
754
|
-
|
755
|
-
self.lex_state = if src.matched == ")" then
|
756
|
-
:expr_endfn
|
757
|
-
else
|
758
|
-
:expr_endarg
|
759
|
-
end
|
760
901
|
|
761
|
-
|
762
|
-
|
902
|
+
text = matched
|
903
|
+
state = text == ")" ? :expr_endfn : :expr_endarg
|
904
|
+
token = {
|
763
905
|
")" => :tRPAREN,
|
764
906
|
"]" => :tRBRACK,
|
765
907
|
"}" => :tRCURLY
|
766
|
-
}[
|
767
|
-
return result
|
768
|
-
elsif src.scan(/\!/) then
|
769
|
-
if in_lex_state?(:expr_fname, :expr_dot) then
|
770
|
-
self.lex_state = :expr_arg
|
771
|
-
|
772
|
-
if src.scan(/@/) then
|
773
|
-
self.yacc_value = "!@"
|
774
|
-
return :tUBANG
|
775
|
-
end
|
776
|
-
else
|
777
|
-
self.lex_state = :expr_beg
|
778
|
-
end
|
908
|
+
}[text]
|
779
909
|
|
780
|
-
|
781
|
-
|
782
|
-
|
783
|
-
|
910
|
+
return result(state, token, text)
|
911
|
+
elsif scan(/\!/) then
|
912
|
+
if in_arg_state? then
|
913
|
+
return result(:expr_arg, :tUBANG, "!@") if scan(/@/)
|
784
914
|
end
|
785
915
|
|
786
|
-
|
787
|
-
|
788
|
-
|
789
|
-
|
790
|
-
return TOKENS[
|
791
|
-
elsif
|
792
|
-
if
|
916
|
+
text = scan(/[=~]/) ? "!#{matched}" : "!"
|
917
|
+
|
918
|
+
return result(arg_state, TOKENS[text], text)
|
919
|
+
elsif scan(/\.\.\.?|,|![=~]?/) then
|
920
|
+
return result(:expr_beg, TOKENS[matched], matched)
|
921
|
+
elsif check(/\./) then
|
922
|
+
if scan(/\.\d/) then
|
793
923
|
rb_compile_error "no .<digit> floating literal anymore put 0 before dot"
|
794
|
-
elsif
|
795
|
-
|
796
|
-
self.yacc_value = "."
|
797
|
-
return :tDOT
|
924
|
+
elsif scan(/\./) then
|
925
|
+
return result(:expr_dot, :tDOT, ".")
|
798
926
|
end
|
799
|
-
elsif
|
800
|
-
|
927
|
+
elsif scan(/\(/) then
|
928
|
+
token = if ruby18 then
|
801
929
|
yylex_paren18
|
802
930
|
else
|
803
931
|
yylex_paren19
|
@@ -805,259 +933,188 @@ class RubyLexer
|
|
805
933
|
|
806
934
|
self.paren_nest += 1
|
807
935
|
|
808
|
-
|
936
|
+
return expr_result(token, "(")
|
937
|
+
elsif check(/\=/) then
|
938
|
+
if scan(/\=\=\=|\=\=|\=~|\=>|\=(?!begin\b)/) then
|
939
|
+
tok = matched
|
940
|
+
return result(:arg_state, TOKENS[tok], tok)
|
941
|
+
elsif beginning_of_line? and scan(/\=begin(?=\s)/) then
|
942
|
+
@comments << matched
|
809
943
|
|
810
|
-
|
811
|
-
elsif src.check(/\=/) then
|
812
|
-
if src.scan(/\=\=\=|\=\=|\=~|\=>|\=(?!begin\b)/) then
|
813
|
-
self.fix_arg_lex_state
|
814
|
-
tok = self.yacc_value = src.matched
|
815
|
-
return TOKENS[tok]
|
816
|
-
elsif src.scan(/\=begin(?=\s)/) then
|
817
|
-
@comments << src.matched
|
818
|
-
|
819
|
-
unless src.scan(/.*?\n=end( |\t|\f)*[^\n]*(\n|\z)/m) then
|
944
|
+
unless scan(/.*?\n=end( |\t|\f)*[^\n]*(\n|\z)/m) then
|
820
945
|
@comments.clear
|
821
946
|
rb_compile_error("embedded document meets end of file")
|
822
947
|
end
|
823
948
|
|
824
|
-
@comments <<
|
949
|
+
@comments << matched
|
825
950
|
|
826
951
|
next
|
952
|
+
elsif scan(/\=(?=begin\b)/) then # h[k]=begin ... end
|
953
|
+
tok = matched
|
954
|
+
return result(:arg_state, TOKENS[tok], tok)
|
827
955
|
else
|
828
956
|
raise "you shouldn't be able to get here"
|
829
957
|
end
|
830
|
-
elsif
|
831
|
-
|
832
|
-
|
833
|
-
|
834
|
-
|
835
|
-
|
836
|
-
|
837
|
-
|
838
|
-
|
839
|
-
self.token
|
840
|
-
|
841
|
-
|
842
|
-
|
843
|
-
|
844
|
-
|
845
|
-
elsif src.scan(/\:\:/) then
|
958
|
+
elsif scan(/\"(#{SIMPLE_STRING})\"/o) then
|
959
|
+
string = matched[1..-2].gsub(ESC) { unescape $1 }
|
960
|
+
return result(:expr_end, :tSTRING, string)
|
961
|
+
elsif scan(/\"/) then # FALLBACK
|
962
|
+
string STR_DQUOTE, '"' # TODO: question this
|
963
|
+
return result(nil, :tSTRING_BEG, '"')
|
964
|
+
elsif scan(/\@\@?#{IDENT_CHAR}+/o) then
|
965
|
+
self.token = matched
|
966
|
+
|
967
|
+
rb_compile_error "`#{self.token}` is not allowed as a variable name" if
|
968
|
+
self.token =~ /\@\d/
|
969
|
+
|
970
|
+
tok_id = matched =~ /^@@/ ? :tCVAR : :tIVAR
|
971
|
+
return result(:expr_end, tok_id, self.token)
|
972
|
+
elsif scan(/\:\:/) then
|
846
973
|
if is_beg? || in_lex_state?(:expr_class) || is_space_arg? then
|
847
|
-
|
848
|
-
self.yacc_value = "::"
|
849
|
-
return :tCOLON3
|
974
|
+
return result(:expr_beg, :tCOLON3, "::")
|
850
975
|
end
|
851
976
|
|
852
|
-
|
853
|
-
|
854
|
-
return :tCOLON2
|
855
|
-
elsif ! is_end? && src.scan(/:([a-zA-Z_]#{IDENT_CHAR_RE}*(?:[?!]|=(?==>)|=(?![=>]))?)/) then
|
977
|
+
return result(:expr_dot, :tCOLON2, "::")
|
978
|
+
elsif ! is_end? && scan(/:([a-zA-Z_]#{IDENT_CHAR}*(?:[?!]|=(?==>)|=(?![=>]))?)/) then
|
856
979
|
# scanning shortcut to symbols
|
857
|
-
|
858
|
-
|
859
|
-
|
860
|
-
|
980
|
+
return result(:expr_end, :tSYMBOL, ss[1])
|
981
|
+
elsif ! is_end? && (scan(/\:\"(#{SIMPLE_STRING})\"/) ||
|
982
|
+
scan(/\:\'(#{SIMPLE_SSTRING})\'/)) then
|
983
|
+
symbol = ss[1].gsub(ESC) { unescape $1 }
|
984
|
+
|
985
|
+
rb_compile_error "symbol cannot contain '\\0'" if
|
986
|
+
ruby18 && symbol =~ /\0/
|
987
|
+
|
988
|
+
return result(:expr_end, :tSYMBOL, symbol)
|
989
|
+
elsif scan(/\:/) then
|
861
990
|
# ?: / then / when
|
862
|
-
if is_end? ||
|
863
|
-
self.lex_state = :expr_beg
|
991
|
+
if is_end? || check(/\s/) then
|
864
992
|
# TODO warn_balanced(":", "symbol literal");
|
865
|
-
|
866
|
-
return :tCOLON
|
993
|
+
return result(:expr_beg, :tCOLON, ":")
|
867
994
|
end
|
868
995
|
|
869
996
|
case
|
870
|
-
when
|
871
|
-
|
872
|
-
when
|
873
|
-
|
997
|
+
when scan(/\'/) then
|
998
|
+
string STR_SSYM, matched
|
999
|
+
when scan(/\"/) then
|
1000
|
+
string STR_DSYM, matched
|
874
1001
|
end
|
875
1002
|
|
876
|
-
|
877
|
-
|
878
|
-
return :tSYMBEG
|
879
|
-
elsif src.check(/[0-9]/) then
|
1003
|
+
return result(:expr_fname, :tSYMBEG, ":")
|
1004
|
+
elsif check(/[0-9]/) then
|
880
1005
|
return parse_number
|
881
|
-
elsif
|
1006
|
+
elsif scan(/\[/) then
|
882
1007
|
self.paren_nest += 1
|
883
1008
|
|
884
|
-
|
1009
|
+
token = nil
|
885
1010
|
|
886
1011
|
if in_lex_state? :expr_fname, :expr_dot then
|
887
|
-
self.lex_state = :expr_arg
|
888
1012
|
case
|
889
|
-
when
|
1013
|
+
when scan(/\]\=/) then
|
890
1014
|
self.paren_nest -= 1 # HACK? I dunno, or bug in MRI
|
891
|
-
|
892
|
-
|
893
|
-
when src.scan(/\]/) then
|
1015
|
+
return result(:expr_arg, :tASET, "[]=")
|
1016
|
+
when scan(/\]/) then
|
894
1017
|
self.paren_nest -= 1 # HACK? I dunno, or bug in MRI
|
895
|
-
|
896
|
-
return :tAREF
|
1018
|
+
return result(:expr_arg, :tAREF, "[]")
|
897
1019
|
else
|
898
1020
|
rb_compile_error "unexpected '['"
|
899
1021
|
end
|
900
1022
|
elsif is_beg? then
|
901
|
-
|
902
|
-
result = :tLBRACK
|
1023
|
+
token = :tLBRACK
|
903
1024
|
elsif is_arg? && space_seen then
|
904
|
-
|
905
|
-
result = :tLBRACK
|
1025
|
+
token = :tLBRACK
|
906
1026
|
else
|
907
|
-
|
1027
|
+
token = :tLBRACK2
|
908
1028
|
end
|
909
1029
|
|
910
|
-
|
911
|
-
|
912
|
-
|
913
|
-
|
914
|
-
|
915
|
-
|
916
|
-
|
917
|
-
|
918
|
-
|
919
|
-
|
920
|
-
|
921
|
-
|
922
|
-
|
923
|
-
self.lex_state = :expr_beg
|
924
|
-
self.yacc_value = "||"
|
925
|
-
return :tOROP
|
926
|
-
elsif src.scan(/\|\=/) then
|
927
|
-
self.lex_state = :expr_beg
|
928
|
-
self.yacc_value = "|"
|
929
|
-
return :tOP_ASGN
|
930
|
-
elsif src.scan(/\|/) then
|
931
|
-
self.fix_arg_lex_state
|
932
|
-
self.yacc_value = "|"
|
933
|
-
return :tPIPE
|
1030
|
+
return expr_result(token, "[")
|
1031
|
+
elsif scan(/\'#{SIMPLE_SSTRING}\'/) then
|
1032
|
+
text = matched[1..-2].gsub(/\\\\/, "\\").gsub(/\\'/, "'") # "
|
1033
|
+
return result(:expr_end, :tSTRING, text)
|
1034
|
+
elsif check(/\|/) then
|
1035
|
+
if scan(/\|\|\=/) then
|
1036
|
+
return result(:expr_beg, :tOP_ASGN, "||")
|
1037
|
+
elsif scan(/\|\|/) then
|
1038
|
+
return result(:expr_beg, :tOROP, "||")
|
1039
|
+
elsif scan(/\|\=/) then
|
1040
|
+
return result(:expr_beg, :tOP_ASGN, "|")
|
1041
|
+
elsif scan(/\|/) then
|
1042
|
+
return result(:arg_state, :tPIPE, "|")
|
934
1043
|
end
|
935
|
-
elsif
|
1044
|
+
elsif scan(/\{/) then
|
936
1045
|
self.brace_nest += 1
|
937
1046
|
if lpar_beg && lpar_beg == paren_nest then
|
938
1047
|
self.lpar_beg = nil
|
939
1048
|
self.paren_nest -= 1
|
940
1049
|
|
941
|
-
|
942
|
-
|
943
|
-
return :tLAMBEG
|
1050
|
+
return expr_result(:tLAMBEG, "{")
|
944
1051
|
end
|
945
1052
|
|
946
|
-
|
1053
|
+
token = if is_arg? || in_lex_state?(:expr_end, :expr_endfn) then
|
947
1054
|
:tLCURLY # block (primary)
|
948
1055
|
elsif in_lex_state?(:expr_endarg) then
|
949
1056
|
:tLBRACE_ARG # block (expr)
|
950
1057
|
else
|
951
|
-
self.tern.push false
|
952
1058
|
:tLBRACE # hash
|
953
1059
|
end
|
954
1060
|
|
955
|
-
self.
|
956
|
-
self.command_start = true unless result == :tLBRACE
|
1061
|
+
self.command_start = true unless token == :tLBRACE
|
957
1062
|
|
958
|
-
return
|
959
|
-
elsif
|
960
|
-
|
961
|
-
|
962
|
-
|
963
|
-
sign = src.matched
|
1063
|
+
return expr_result(token, "{")
|
1064
|
+
elsif scan(/->/) then
|
1065
|
+
return result(:expr_endfn, :tLAMBDA, nil)
|
1066
|
+
elsif scan(/[+-]/) then
|
1067
|
+
sign = matched
|
964
1068
|
utype, type = if sign == "+" then
|
965
1069
|
[:tUPLUS, :tPLUS]
|
966
1070
|
else
|
967
1071
|
[:tUMINUS, :tMINUS]
|
968
1072
|
end
|
969
1073
|
|
970
|
-
if
|
971
|
-
|
972
|
-
|
973
|
-
self.yacc_value = "#{sign}@"
|
974
|
-
return utype
|
1074
|
+
if in_arg_state? then
|
1075
|
+
if scan(/@/) then
|
1076
|
+
return result(:expr_arg, utype, "#{sign}@")
|
975
1077
|
else
|
976
|
-
|
977
|
-
return type
|
1078
|
+
return result(:expr_arg, type, sign)
|
978
1079
|
end
|
979
1080
|
end
|
980
1081
|
|
981
|
-
if
|
982
|
-
self.lex_state = :expr_beg
|
983
|
-
self.yacc_value = sign
|
984
|
-
return :tOP_ASGN
|
985
|
-
end
|
1082
|
+
return result(:expr_beg, :tOP_ASGN, sign) if scan(/\=/)
|
986
1083
|
|
987
|
-
if (is_beg? || (is_arg? && space_seen && !
|
988
|
-
if is_arg?
|
989
|
-
arg_ambiguous
|
990
|
-
end
|
1084
|
+
if (is_beg? || (is_arg? && space_seen && !check(/\s/))) then
|
1085
|
+
arg_ambiguous if is_arg?
|
991
1086
|
|
992
|
-
|
993
|
-
|
994
|
-
|
995
|
-
if src.check(/\d/) then
|
996
|
-
if utype == :tUPLUS then
|
997
|
-
return self.parse_number
|
998
|
-
else
|
999
|
-
return :tUMINUS_NUM
|
1000
|
-
end
|
1087
|
+
if check(/\d/) then
|
1088
|
+
return self.parse_number if utype == :tUPLUS
|
1089
|
+
return result(:expr_beg, :tUMINUS_NUM, sign)
|
1001
1090
|
end
|
1002
1091
|
|
1003
|
-
return utype
|
1092
|
+
return result(:expr_beg, utype, sign)
|
1004
1093
|
end
|
1005
1094
|
|
1006
|
-
|
1007
|
-
|
1008
|
-
|
1009
|
-
|
1010
|
-
|
1011
|
-
|
1012
|
-
|
1013
|
-
return :
|
1014
|
-
elsif
|
1015
|
-
result
|
1016
|
-
|
1017
|
-
|
1018
|
-
elsif is_beg? then
|
1019
|
-
:tDSTAR
|
1020
|
-
else
|
1021
|
-
# TODO: warn_balanced("**", "argument prefix");
|
1022
|
-
:tPOW
|
1023
|
-
end
|
1024
|
-
self.yacc_value = "**"
|
1025
|
-
self.fix_arg_lex_state
|
1026
|
-
return result
|
1027
|
-
elsif src.scan(/\*\=/) then
|
1028
|
-
self.lex_state = :expr_beg
|
1029
|
-
self.yacc_value = "*"
|
1030
|
-
return :tOP_ASGN
|
1031
|
-
elsif src.scan(/\*/) then
|
1032
|
-
result = if is_space_arg? src.check(/./m) then
|
1033
|
-
warning("`*' interpreted as argument prefix")
|
1034
|
-
:tSTAR
|
1035
|
-
elsif is_beg? then
|
1036
|
-
:tSTAR
|
1037
|
-
else
|
1038
|
-
# TODO: warn_balanced("*", "argument prefix");
|
1039
|
-
:tSTAR2 # TODO: rename
|
1040
|
-
end
|
1095
|
+
return result(:expr_beg, type, sign)
|
1096
|
+
elsif check(/\*/) then
|
1097
|
+
if scan(/\*\*=/) then
|
1098
|
+
return result(:expr_beg, :tOP_ASGN, "**")
|
1099
|
+
elsif scan(/\*\*/) then
|
1100
|
+
token = space_vs_beginning :tDSTAR, :tDSTAR, :tPOW
|
1101
|
+
|
1102
|
+
return result(:arg_state, token, "**")
|
1103
|
+
elsif scan(/\*\=/) then
|
1104
|
+
return result(:expr_beg, :tOP_ASGN, "*")
|
1105
|
+
elsif scan(/\*/) then
|
1106
|
+
token = space_vs_beginning :tSTAR, :tSTAR, :tSTAR2
|
1041
1107
|
|
1042
|
-
|
1043
|
-
self.fix_arg_lex_state
|
1044
|
-
return result
|
1108
|
+
return result(:arg_state, token, "*")
|
1045
1109
|
end
|
1046
|
-
elsif
|
1047
|
-
if
|
1048
|
-
|
1049
|
-
|
1050
|
-
return :
|
1051
|
-
elsif
|
1052
|
-
|
1053
|
-
|
1054
|
-
return :tLEQ
|
1055
|
-
elsif src.scan(/\<\<\=/) then
|
1056
|
-
self.fix_arg_lex_state
|
1057
|
-
self.lex_state = :expr_beg
|
1058
|
-
self.yacc_value = "\<\<"
|
1059
|
-
return :tOP_ASGN
|
1060
|
-
elsif src.scan(/\<\</) then
|
1110
|
+
elsif check(/\</) then
|
1111
|
+
if scan(/\<\=\>/) then
|
1112
|
+
return result(:arg_state, :tCMP, "<=>")
|
1113
|
+
elsif scan(/\<\=/) then
|
1114
|
+
return result(:arg_state, :tLEQ, "<=")
|
1115
|
+
elsif scan(/\<\<\=/) then
|
1116
|
+
return result(:arg_state, :tOP_ASGN, "<<")
|
1117
|
+
elsif scan(/\<\</) then
|
1061
1118
|
if (!in_lex_state?(:expr_dot, :expr_class) &&
|
1062
1119
|
!is_end? &&
|
1063
1120
|
(!is_arg? || space_seen)) then
|
@@ -1065,70 +1122,49 @@ class RubyLexer
|
|
1065
1122
|
return tok if tok
|
1066
1123
|
end
|
1067
1124
|
|
1068
|
-
|
1069
|
-
|
1070
|
-
return :
|
1071
|
-
elsif src.scan(/\</) then
|
1072
|
-
self.fix_arg_lex_state
|
1073
|
-
self.yacc_value = "<"
|
1074
|
-
return :tLT
|
1125
|
+
return result(:arg_state, :tLSHFT, "\<\<")
|
1126
|
+
elsif scan(/\</) then
|
1127
|
+
return result(:arg_state, :tLT, "<")
|
1075
1128
|
end
|
1076
|
-
elsif
|
1077
|
-
if
|
1078
|
-
|
1079
|
-
|
1080
|
-
return :
|
1081
|
-
elsif
|
1082
|
-
|
1083
|
-
|
1084
|
-
|
1085
|
-
return :tOP_ASGN
|
1086
|
-
elsif src.scan(/\>\>/) then
|
1087
|
-
self.fix_arg_lex_state
|
1088
|
-
self.yacc_value = ">>"
|
1089
|
-
return :tRSHFT
|
1090
|
-
elsif src.scan(/\>/) then
|
1091
|
-
self.fix_arg_lex_state
|
1092
|
-
self.yacc_value = ">"
|
1093
|
-
return :tGT
|
1129
|
+
elsif check(/\>/) then
|
1130
|
+
if scan(/\>\=/) then
|
1131
|
+
return result(:arg_state, :tGEQ, ">=")
|
1132
|
+
elsif scan(/\>\>=/) then
|
1133
|
+
return result(:arg_state, :tOP_ASGN, ">>")
|
1134
|
+
elsif scan(/\>\>/) then
|
1135
|
+
return result(:arg_state, :tRSHFT, ">>")
|
1136
|
+
elsif scan(/\>/) then
|
1137
|
+
return result(:arg_state, :tGT, ">")
|
1094
1138
|
end
|
1095
|
-
elsif
|
1096
|
-
self.yacc_value = "`"
|
1139
|
+
elsif scan(/\`/) then
|
1097
1140
|
case lex_state
|
1098
1141
|
when :expr_fname then
|
1099
|
-
|
1100
|
-
return :tBACK_REF2
|
1142
|
+
return result(:expr_end, :tBACK_REF2, "`")
|
1101
1143
|
when :expr_dot then
|
1102
|
-
|
1103
|
-
|
1104
|
-
|
1105
|
-
|
1106
|
-
|
1107
|
-
return :tBACK_REF2
|
1144
|
+
state = command_state ? :expr_cmdarg : :expr_arg
|
1145
|
+
return result(state, :tBACK_REF2, "`")
|
1146
|
+
else
|
1147
|
+
string STR_XQUOTE, '`'
|
1148
|
+
return result(nil, :tXSTRING_BEG, "`")
|
1108
1149
|
end
|
1109
|
-
|
1110
|
-
return :tXSTRING_BEG
|
1111
|
-
elsif src.scan(/\?/) then
|
1112
|
-
|
1150
|
+
elsif scan(/\?/) then
|
1113
1151
|
if is_end? then
|
1114
|
-
|
1115
|
-
|
1116
|
-
self.yacc_value = "?"
|
1117
|
-
return :tEH
|
1152
|
+
state = ruby18 ? :expr_beg : :expr_value # HACK?
|
1153
|
+
return result(state, :tEH, "?")
|
1118
1154
|
end
|
1119
1155
|
|
1120
|
-
if
|
1156
|
+
if end_of_stream? then
|
1121
1157
|
rb_compile_error "incomplete character syntax"
|
1122
1158
|
end
|
1123
1159
|
|
1124
|
-
if
|
1160
|
+
if check(/\s|\v/) then
|
1125
1161
|
unless is_arg? then
|
1126
1162
|
c2 = { " " => 's',
|
1127
1163
|
"\n" => 'n',
|
1128
1164
|
"\t" => 't',
|
1129
1165
|
"\v" => 'v',
|
1130
1166
|
"\r" => 'r',
|
1131
|
-
"\f" => 'f' }[
|
1167
|
+
"\f" => 'f' }[matched]
|
1132
1168
|
|
1133
1169
|
if c2 then
|
1134
1170
|
warning("invalid character syntax; use ?\\" + c2)
|
@@ -1136,47 +1172,32 @@ class RubyLexer
|
|
1136
1172
|
end
|
1137
1173
|
|
1138
1174
|
# ternary
|
1139
|
-
|
1140
|
-
|
1141
|
-
|
1142
|
-
return :tEH
|
1143
|
-
elsif src.check(/\w(?=\w)/) then # ternary, also
|
1144
|
-
self.lex_state = :expr_beg
|
1145
|
-
self.tern.push true
|
1146
|
-
self.yacc_value = "?"
|
1147
|
-
return :tEH
|
1175
|
+
state = ruby18 ? :expr_beg : :expr_value # HACK?
|
1176
|
+
return result(state, :tEH, "?")
|
1177
|
+
elsif check(/\w(?=\w)/) then # ternary, also
|
1178
|
+
return result(:expr_beg, :tEH, "?")
|
1148
1179
|
end
|
1149
1180
|
|
1150
|
-
c = if
|
1181
|
+
c = if scan(/\\/) then
|
1151
1182
|
self.read_escape
|
1152
1183
|
else
|
1153
|
-
|
1184
|
+
ss.getch
|
1154
1185
|
end
|
1155
|
-
self.lex_state = :expr_end
|
1156
1186
|
|
1157
1187
|
if version == 18 then
|
1158
|
-
|
1159
|
-
return :tINTEGER
|
1188
|
+
return result(:expr_end, :tINTEGER, c[0].ord & 0xff)
|
1160
1189
|
else
|
1161
|
-
|
1162
|
-
return :tSTRING
|
1190
|
+
return result(:expr_end, :tSTRING, c)
|
1163
1191
|
end
|
1164
|
-
elsif
|
1165
|
-
if
|
1166
|
-
|
1167
|
-
|
1168
|
-
return :
|
1169
|
-
elsif
|
1170
|
-
|
1171
|
-
|
1172
|
-
|
1173
|
-
elsif src.scan(/\&\=/) then
|
1174
|
-
self.yacc_value = "&"
|
1175
|
-
self.lex_state = :expr_beg
|
1176
|
-
return :tOP_ASGN
|
1177
|
-
elsif src.scan(/&/) then
|
1178
|
-
result = if is_arg? && space_seen &&
|
1179
|
-
!src.check(/\s/) then
|
1192
|
+
elsif check(/\&/) then
|
1193
|
+
if scan(/\&\&\=/) then
|
1194
|
+
return result(:expr_beg, :tOP_ASGN, "&&")
|
1195
|
+
elsif scan(/\&\&/) then
|
1196
|
+
return result(:expr_beg, :tANDOP, "&&")
|
1197
|
+
elsif scan(/\&\=/) then
|
1198
|
+
return result(:expr_beg, :tOP_ASGN, "&")
|
1199
|
+
elsif scan(/&/) then
|
1200
|
+
token = if is_arg? && space_seen && !check(/\s/) then
|
1180
1201
|
warning("`&' interpreted as argument prefix")
|
1181
1202
|
:tAMPER
|
1182
1203
|
elsif in_lex_state? :expr_beg, :expr_mid then
|
@@ -1185,170 +1206,121 @@ class RubyLexer
|
|
1185
1206
|
:tAMPER2
|
1186
1207
|
end
|
1187
1208
|
|
1188
|
-
|
1189
|
-
self.yacc_value = "&"
|
1190
|
-
return result
|
1209
|
+
return result(:arg_state, token, "&")
|
1191
1210
|
end
|
1192
|
-
elsif
|
1211
|
+
elsif scan(/\//) then
|
1193
1212
|
if is_beg? then
|
1194
|
-
|
1195
|
-
|
1196
|
-
return :tREGEXP_BEG
|
1213
|
+
string STR_REGEXP, '/'
|
1214
|
+
return result(nil, :tREGEXP_BEG, "/")
|
1197
1215
|
end
|
1198
1216
|
|
1199
|
-
if
|
1200
|
-
|
1201
|
-
self.lex_state = :expr_beg
|
1202
|
-
return :tOP_ASGN
|
1217
|
+
if scan(/\=/) then
|
1218
|
+
return result(:expr_beg, :tOP_ASGN, "/")
|
1203
1219
|
end
|
1204
1220
|
|
1205
1221
|
if is_arg? && space_seen then
|
1206
|
-
unless
|
1222
|
+
unless scan(/\s/) then
|
1207
1223
|
arg_ambiguous
|
1208
|
-
|
1209
|
-
|
1210
|
-
return :tREGEXP_BEG
|
1224
|
+
string STR_REGEXP, '/'
|
1225
|
+
return result(nil, :tREGEXP_BEG, "/")
|
1211
1226
|
end
|
1212
1227
|
end
|
1213
1228
|
|
1214
|
-
|
1215
|
-
|
1216
|
-
|
1217
|
-
|
1218
|
-
|
1219
|
-
|
1220
|
-
self.yacc_value = "^"
|
1221
|
-
return :tOP_ASGN
|
1222
|
-
elsif src.scan(/\^/) then
|
1223
|
-
self.fix_arg_lex_state
|
1224
|
-
self.yacc_value = "^"
|
1225
|
-
return :tCARET
|
1226
|
-
elsif src.scan(/\;/) then
|
1229
|
+
return result(:arg_state, :tDIVIDE, "/")
|
1230
|
+
elsif scan(/\^=/) then
|
1231
|
+
return result(:expr_beg, :tOP_ASGN, "^")
|
1232
|
+
elsif scan(/\^/) then
|
1233
|
+
return result(:arg_state, :tCARET, "^")
|
1234
|
+
elsif scan(/\;/) then
|
1227
1235
|
self.command_start = true
|
1228
|
-
|
1229
|
-
|
1230
|
-
|
1231
|
-
|
1232
|
-
|
1233
|
-
|
1234
|
-
end
|
1235
|
-
|
1236
|
-
self.fix_arg_lex_state
|
1237
|
-
self.yacc_value = "~"
|
1238
|
-
|
1239
|
-
return :tTILDE
|
1240
|
-
elsif src.scan(/\\/) then
|
1241
|
-
if src.scan(/\r?\n/) then
|
1236
|
+
return result(:expr_beg, :tSEMI, ";")
|
1237
|
+
elsif scan(/\~/) then
|
1238
|
+
scan(/@/) if in_lex_state? :expr_fname, :expr_dot
|
1239
|
+
return result(:arg_state, :tTILDE, "~")
|
1240
|
+
elsif scan(/\\/) then
|
1241
|
+
if scan(/\r?\n/) then
|
1242
1242
|
self.lineno = nil
|
1243
1243
|
self.space_seen = true
|
1244
1244
|
next
|
1245
1245
|
end
|
1246
1246
|
rb_compile_error "bare backslash only allowed before newline"
|
1247
|
-
elsif
|
1248
|
-
if is_beg?
|
1249
|
-
|
1250
|
-
|
1251
|
-
|
1252
|
-
if
|
1253
|
-
|
1254
|
-
|
1255
|
-
|
1256
|
-
|
1257
|
-
|
1258
|
-
|
1259
|
-
|
1260
|
-
|
1261
|
-
|
1262
|
-
|
1263
|
-
|
1264
|
-
elsif src.check(/\$/) then
|
1265
|
-
if src.scan(/(\$_)(\w+)/) then
|
1266
|
-
self.lex_state = :expr_end
|
1267
|
-
self.token = src.matched
|
1268
|
-
return process_token(command_state)
|
1269
|
-
elsif src.scan(/\$_/) then
|
1270
|
-
self.lex_state = :expr_end
|
1271
|
-
self.token = src.matched
|
1272
|
-
self.yacc_value = src.matched
|
1273
|
-
return :tGVAR
|
1274
|
-
elsif src.scan(/\$[~*$?!@\/\\;,.=:<>\"]|\$-\w?/) then
|
1275
|
-
self.lex_state = :expr_end
|
1276
|
-
self.yacc_value = src.matched
|
1277
|
-
return :tGVAR
|
1278
|
-
elsif src.scan(/\$([\&\`\'\+])/) then
|
1279
|
-
self.lex_state = :expr_end
|
1247
|
+
elsif scan(/\%/) then
|
1248
|
+
return parse_quote if is_beg?
|
1249
|
+
|
1250
|
+
return result(:expr_beg, :tOP_ASGN, "%") if scan(/\=/)
|
1251
|
+
|
1252
|
+
return parse_quote if is_arg? && space_seen && ! check(/\s/)
|
1253
|
+
|
1254
|
+
return result(:arg_state, :tPERCENT, "%")
|
1255
|
+
elsif check(/\$/) then
|
1256
|
+
if scan(/(\$_)(\w+)/) then
|
1257
|
+
self.token = matched
|
1258
|
+
return result(:expr_end, :tGVAR, matched)
|
1259
|
+
elsif scan(/\$_/) then
|
1260
|
+
return result(:expr_end, :tGVAR, matched)
|
1261
|
+
elsif scan(/\$[~*$?!@\/\\;,.=:<>\"]|\$-\w?/) then
|
1262
|
+
return result(:expr_end, :tGVAR, matched)
|
1263
|
+
elsif scan(/\$([\&\`\'\+])/) then
|
1280
1264
|
# Explicit reference to these vars as symbols...
|
1281
|
-
if
|
1282
|
-
|
1283
|
-
return :tGVAR
|
1265
|
+
if lex_state == :expr_fname then
|
1266
|
+
return result(:expr_end, :tGVAR, matched)
|
1284
1267
|
else
|
1285
|
-
|
1286
|
-
return :tBACK_REF
|
1268
|
+
return result(:expr_end, :tBACK_REF, ss[1].to_sym)
|
1287
1269
|
end
|
1288
|
-
elsif
|
1289
|
-
|
1290
|
-
|
1291
|
-
self.yacc_value = src.matched
|
1292
|
-
return :tGVAR
|
1270
|
+
elsif scan(/\$([1-9]\d*)/) then
|
1271
|
+
if lex_state == :expr_fname then
|
1272
|
+
return result(:expr_end, :tGVAR, matched)
|
1293
1273
|
else
|
1294
|
-
|
1295
|
-
return :tNTH_REF
|
1274
|
+
return result(:expr_end, :tNTH_REF, ss[1].to_i)
|
1296
1275
|
end
|
1297
|
-
elsif
|
1298
|
-
|
1299
|
-
|
1300
|
-
return
|
1301
|
-
elsif
|
1302
|
-
|
1303
|
-
self.yacc_value = "$"
|
1304
|
-
return "$"
|
1305
|
-
elsif src.scan(/\$\w+/)
|
1306
|
-
self.lex_state = :expr_end
|
1307
|
-
self.token = src.matched
|
1308
|
-
return process_token(command_state)
|
1276
|
+
elsif scan(/\$0/) then
|
1277
|
+
return result(:expr_end, :tGVAR, matched)
|
1278
|
+
elsif scan(/\$\W|\$\z/) then # TODO: remove?
|
1279
|
+
return result(:expr_end, "$", "$") # FIX: "$"??
|
1280
|
+
elsif scan(/\$\w+/)
|
1281
|
+
return result(:expr_end, :tGVAR, matched)
|
1309
1282
|
end
|
1310
|
-
elsif
|
1311
|
-
if
|
1283
|
+
elsif check(/\_/) then
|
1284
|
+
if beginning_of_line? && scan(/\__END__(\r?\n|\Z)/) then
|
1312
1285
|
self.lineno = nil
|
1313
1286
|
return RubyLexer::EOF
|
1314
|
-
elsif
|
1315
|
-
self.token =
|
1316
|
-
return process_token
|
1287
|
+
elsif scan(/\_\w*/) then
|
1288
|
+
self.token = matched
|
1289
|
+
return process_token command_state, last_state
|
1317
1290
|
end
|
1318
1291
|
end
|
1319
1292
|
end # END OF CASE
|
1320
1293
|
|
1321
|
-
if
|
1294
|
+
if scan(/\004|\032|\000/) || end_of_stream? then # ^D, ^Z, EOF
|
1322
1295
|
return RubyLexer::EOF
|
1323
1296
|
else # alpha check
|
1324
|
-
rb_compile_error "Invalid char #{
|
1325
|
-
|
1297
|
+
rb_compile_error "Invalid char #{ss.rest[0].chr} in expression" unless
|
1298
|
+
check IDENT
|
1326
1299
|
end
|
1327
1300
|
|
1328
|
-
self.token =
|
1301
|
+
self.token = matched if self.scan IDENT
|
1329
1302
|
|
1330
|
-
return process_token
|
1303
|
+
return process_token command_state, last_state
|
1331
1304
|
end
|
1332
1305
|
end
|
1333
1306
|
|
1334
1307
|
def yylex_paren18
|
1335
1308
|
self.command_start = true
|
1336
|
-
|
1309
|
+
token = :tLPAREN2
|
1337
1310
|
|
1338
1311
|
if in_lex_state? :expr_beg, :expr_mid then
|
1339
|
-
|
1312
|
+
token = :tLPAREN
|
1340
1313
|
elsif space_seen then
|
1341
1314
|
if in_lex_state? :expr_cmdarg then
|
1342
|
-
|
1315
|
+
token = :tLPAREN_ARG
|
1343
1316
|
elsif in_lex_state? :expr_arg then
|
1344
|
-
self.tern.push false
|
1345
1317
|
warning "don't put space before argument parentheses"
|
1346
1318
|
end
|
1347
1319
|
else
|
1348
|
-
|
1320
|
+
# not a ternary -- do nothing?
|
1349
1321
|
end
|
1350
1322
|
|
1351
|
-
|
1323
|
+
token
|
1352
1324
|
end
|
1353
1325
|
|
1354
1326
|
def yylex_paren19
|
@@ -1361,146 +1333,7 @@ class RubyLexer
|
|
1361
1333
|
end
|
1362
1334
|
end
|
1363
1335
|
|
1364
|
-
def
|
1365
|
-
in_lex_state? :expr_arg, :expr_cmdarg
|
1366
|
-
end
|
1367
|
-
|
1368
|
-
def is_end?
|
1369
|
-
in_lex_state? :expr_end, :expr_endarg, :expr_endfn
|
1370
|
-
end
|
1371
|
-
|
1372
|
-
def is_beg?
|
1373
|
-
in_lex_state? :expr_beg, :expr_value, :expr_mid, :expr_class
|
1374
|
-
end
|
1375
|
-
|
1376
|
-
# TODO #define IS_AFTER_OPERATOR() IS_lex_state(EXPR_FNAME | EXPR_DOT)
|
1377
|
-
|
1378
|
-
def is_space_arg? c = "x"
|
1379
|
-
is_arg? and space_seen and c !~ /\s/
|
1380
|
-
end
|
1381
|
-
|
1382
|
-
def is_label_possible? command_state
|
1383
|
-
(in_lex_state?(:expr_beg) && !command_state) || is_arg?
|
1384
|
-
end
|
1385
|
-
|
1386
|
-
def process_token(command_state)
|
1387
|
-
token << src.matched if token =~ IDENT_RE && src.scan(/[\!\?](?!=)/)
|
1388
|
-
|
1389
|
-
result = nil
|
1390
|
-
last_state = lex_state
|
1391
|
-
|
1392
|
-
case token
|
1393
|
-
when /^\$/ then
|
1394
|
-
self.lex_state, result = :expr_end, :tGVAR
|
1395
|
-
when /^@@/ then
|
1396
|
-
self.lex_state, result = :expr_end, :tCVAR
|
1397
|
-
when /^@/ then
|
1398
|
-
self.lex_state, result = :expr_end, :tIVAR
|
1399
|
-
else
|
1400
|
-
if token =~ /[!?]$/ then
|
1401
|
-
result = :tFID
|
1402
|
-
else
|
1403
|
-
if in_lex_state? :expr_fname then
|
1404
|
-
# ident=, not =~ => == or followed by =>
|
1405
|
-
# TODO test lexing of a=>b vs a==>b
|
1406
|
-
if src.scan(/=(?:(?![~>=])|(?==>))/) then
|
1407
|
-
result = :tIDENTIFIER
|
1408
|
-
token << src.matched
|
1409
|
-
end
|
1410
|
-
end
|
1411
|
-
|
1412
|
-
result ||= if token =~ /^[A-Z]/ then
|
1413
|
-
:tCONSTANT
|
1414
|
-
else
|
1415
|
-
:tIDENTIFIER
|
1416
|
-
end
|
1417
|
-
end
|
1418
|
-
|
1419
|
-
unless ruby18
|
1420
|
-
if is_label_possible? command_state then
|
1421
|
-
colon = src.scan(/:/)
|
1422
|
-
|
1423
|
-
if colon && src.peek(1) != ":" then
|
1424
|
-
self.lex_state = :expr_beg
|
1425
|
-
self.yacc_value = [token, src.lineno]
|
1426
|
-
return :tLABEL
|
1427
|
-
end
|
1428
|
-
|
1429
|
-
src.unscan if colon
|
1430
|
-
end
|
1431
|
-
end
|
1432
|
-
|
1433
|
-
unless in_lex_state? :expr_dot then
|
1434
|
-
# See if it is a reserved word.
|
1435
|
-
keyword = if ruby18 then # REFACTOR need 18/19 lexer subclasses
|
1436
|
-
RubyParserStuff::Keyword.keyword18 token
|
1437
|
-
else
|
1438
|
-
RubyParserStuff::Keyword.keyword19 token
|
1439
|
-
end
|
1440
|
-
|
1441
|
-
if keyword then
|
1442
|
-
state = lex_state
|
1443
|
-
self.lex_state = keyword.state
|
1444
|
-
self.yacc_value = [token, src.lineno]
|
1445
|
-
|
1446
|
-
if state == :expr_fname then
|
1447
|
-
self.yacc_value = keyword.name
|
1448
|
-
return keyword.id0
|
1449
|
-
end
|
1450
|
-
|
1451
|
-
self.command_start = true if lex_state == :expr_beg
|
1452
|
-
|
1453
|
-
if keyword.id0 == :kDO then
|
1454
|
-
if lpar_beg && lpar_beg == paren_nest then
|
1455
|
-
self.lpar_beg = nil
|
1456
|
-
self.paren_nest -= 1
|
1457
|
-
|
1458
|
-
return :kDO_LAMBDA
|
1459
|
-
end
|
1460
|
-
|
1461
|
-
return :kDO_COND if cond.is_in_state
|
1462
|
-
return :kDO_BLOCK if cmdarg.is_in_state && state != :expr_cmdarg
|
1463
|
-
return :kDO_BLOCK if [:expr_beg, :expr_endarg].include? state
|
1464
|
-
return :kDO
|
1465
|
-
end
|
1466
|
-
|
1467
|
-
return keyword.id0 if [:expr_beg, :expr_value].include? state
|
1468
|
-
|
1469
|
-
self.lex_state = :expr_beg if keyword.id0 != keyword.id1
|
1470
|
-
|
1471
|
-
return keyword.id1
|
1472
|
-
end
|
1473
|
-
end
|
1474
|
-
|
1475
|
-
# TODO:
|
1476
|
-
# if (mb == ENC_CODERANGE_7BIT && lex_state != EXPR_DOT) {
|
1477
|
-
|
1478
|
-
self.lex_state =
|
1479
|
-
if is_beg? || is_arg? || in_lex_state?(:expr_dot) then
|
1480
|
-
if command_state then
|
1481
|
-
:expr_cmdarg
|
1482
|
-
else
|
1483
|
-
:expr_arg
|
1484
|
-
end
|
1485
|
-
elsif !ruby18 && in_lex_state?(:expr_fname) then
|
1486
|
-
:expr_endfn
|
1487
|
-
else
|
1488
|
-
:expr_end
|
1489
|
-
end
|
1490
|
-
|
1491
|
-
end
|
1492
|
-
|
1493
|
-
self.yacc_value = token
|
1494
|
-
|
1495
|
-
if (![:expr_dot, :expr_fname].include?(last_state) &&
|
1496
|
-
self.parser.env[token.to_sym] == :lvar) then
|
1497
|
-
self.lex_state = :expr_end
|
1498
|
-
end
|
1499
|
-
|
1500
|
-
return result
|
1501
|
-
end
|
1502
|
-
|
1503
|
-
def yylex_string # 23 lines
|
1336
|
+
def yylex_string # TODO: rewrite / remove
|
1504
1337
|
token = if lex_strterm[0] == :heredoc then
|
1505
1338
|
self.heredoc lex_strterm
|
1506
1339
|
else
|