1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
|
package cgo
// This file implements a parser of a subset of the C language, just enough to
// parse common #define statements to Go constant expressions.
import (
"fmt"
"go/ast"
"go/scanner"
"go/token"
"strings"
)
// parseConst parses the given string as a C constant.
func parseConst(pos token.Pos, fset *token.FileSet, value string) (ast.Expr, *scanner.Error) {
t := newTokenizer(pos, fset, value)
expr, err := parseConstExpr(t)
if t.token != token.EOF {
return nil, &scanner.Error{
Pos: t.fset.Position(t.pos),
Msg: "unexpected token " + t.token.String(),
}
}
return expr, err
}
// parseConstExpr parses a stream of C tokens to a Go expression.
func parseConstExpr(t *tokenizer) (ast.Expr, *scanner.Error) {
switch t.token {
case token.LPAREN:
lparen := t.pos
t.Next()
x, err := parseConstExpr(t)
if err != nil {
return nil, err
}
if t.token != token.RPAREN {
return nil, unexpectedToken(t, token.RPAREN)
}
expr := &ast.ParenExpr{
Lparen: lparen,
X: x,
Rparen: t.pos,
}
t.Next()
return expr, nil
case token.INT, token.FLOAT, token.STRING, token.CHAR:
expr := &ast.BasicLit{
ValuePos: t.pos,
Kind: t.token,
Value: t.value,
}
t.Next()
return expr, nil
case token.IDENT:
expr := &ast.Ident{
NamePos: t.pos,
Name: "C." + t.value,
}
t.Next()
return expr, nil
case token.EOF:
return nil, &scanner.Error{
Pos: t.fset.Position(t.pos),
Msg: "empty constant",
}
default:
return nil, &scanner.Error{
Pos: t.fset.Position(t.pos),
Msg: fmt.Sprintf("unexpected token %s", t.token),
}
}
}
// unexpectedToken returns an error of the form "unexpected token FOO, expected
// BAR".
func unexpectedToken(t *tokenizer, expected token.Token) *scanner.Error {
return &scanner.Error{
Pos: t.fset.Position(t.pos),
Msg: fmt.Sprintf("unexpected token %s, expected %s", t.token, expected),
}
}
// tokenizer reads C source code and converts it to Go tokens.
type tokenizer struct {
pos token.Pos
fset *token.FileSet
token token.Token
value string
buf string
}
// newTokenizer initializes a new tokenizer, positioned at the first token in
// the string.
func newTokenizer(start token.Pos, fset *token.FileSet, buf string) *tokenizer {
t := &tokenizer{
pos: start,
fset: fset,
buf: buf,
token: token.ILLEGAL,
}
t.Next() // Parse the first token.
return t
}
// Next consumes the next token in the stream. There is no return value, read
// the next token from the pos, token and value properties.
func (t *tokenizer) Next() {
t.pos += token.Pos(len(t.value))
for {
if len(t.buf) == 0 {
t.token = token.EOF
return
}
c := t.buf[0]
switch {
case c == ' ' || c == '\f' || c == '\n' || c == '\r' || c == '\t' || c == '\v':
// Skip whitespace.
// Based on this source, not sure whether it represents C whitespace:
// https://en.cppreference.com/w/cpp/string/byte/isspace
t.pos++
t.buf = t.buf[1:]
case c == '(' || c == ')':
// Single-character tokens.
switch c {
case '(':
t.token = token.LPAREN
case ')':
t.token = token.RPAREN
}
t.value = t.buf[:1]
t.buf = t.buf[1:]
return
case c >= '0' && c <= '9':
// Numeric constant (int, float, etc.).
// Find the last non-numeric character.
tokenLen := len(t.buf)
hasDot := false
for i, c := range t.buf {
if c == '.' {
hasDot = true
}
if c >= '0' && c <= '9' || c == '.' || c == '_' || c >= 'a' && c <= 'z' || c >= 'A' && c <= 'Z' {
tokenLen = i + 1
} else {
break
}
}
t.value = t.buf[:tokenLen]
t.buf = t.buf[tokenLen:]
if hasDot {
// Integer constants are more complicated than this but this is
// a close approximation.
// https://en.cppreference.com/w/cpp/language/integer_literal
t.token = token.FLOAT
t.value = strings.TrimRight(t.value, "f")
} else {
t.token = token.INT
t.value = strings.TrimRight(t.value, "uUlL")
}
return
case c >= 'A' && c <= 'Z' || c >= 'a' && c <= 'z' || c == '_':
// Identifier. Find all remaining tokens that are part of this
// identifier.
tokenLen := len(t.buf)
for i, c := range t.buf {
if c >= '0' && c <= '9' || c >= 'A' && c <= 'Z' || c >= 'a' && c <= 'z' || c == '_' {
tokenLen = i + 1
} else {
break
}
}
t.value = t.buf[:tokenLen]
t.buf = t.buf[tokenLen:]
t.token = token.IDENT
return
case c == '"':
// String constant. Find the first '"' character that is not
// preceded by a backslash.
escape := false
tokenLen := len(t.buf)
for i, c := range t.buf {
if i != 0 && c == '"' && !escape {
tokenLen = i + 1
break
}
if !escape {
escape = c == '\\'
}
}
t.token = token.STRING
t.value = t.buf[:tokenLen]
t.buf = t.buf[tokenLen:]
return
case c == '\'':
// Char (rune) constant. Find the first '\'' character that is not
// preceded by a backslash.
escape := false
tokenLen := len(t.buf)
for i, c := range t.buf {
if i != 0 && c == '\'' && !escape {
tokenLen = i + 1
break
}
if !escape {
escape = c == '\\'
}
}
t.token = token.CHAR
t.value = t.buf[:tokenLen]
t.buf = t.buf[tokenLen:]
return
default:
t.token = token.ILLEGAL
return
}
}
}
|
