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
|
import io::reader_util;
import io::println;//XXXXXXXXxxx
import util::interner;
import lexer::{ string_reader, bump, is_eof, nextch,
is_whitespace, get_str_from, string_reader_as_reader };
export cmnt;
export lit;
export cmnt_style;
export gather_comments_and_literals;
enum cmnt_style {
isolated, // No code on either side of each line of the comment
trailing, // Code exists to the left of the comment
mixed, // Code before /* foo */ and after the comment
blank_line, // Just a manual blank line "\n\n", for layout
}
type cmnt = {style: cmnt_style, lines: [str]/~, pos: uint};
fn read_to_eol(rdr: string_reader) -> str {
let mut val = "";
while rdr.curr != '\n' && !is_eof(rdr) {
str::push_char(val, rdr.curr);
bump(rdr);
}
if rdr.curr == '\n' { bump(rdr); }
ret val;
}
fn read_one_line_comment(rdr: string_reader) -> str {
let val = read_to_eol(rdr);
assert ((val[0] == '/' as u8 && val[1] == '/' as u8) ||
(val[0] == '#' as u8 && val[1] == '!' as u8));
ret val;
}
fn consume_non_eol_whitespace(rdr: string_reader) {
while is_whitespace(rdr.curr) && rdr.curr != '\n' && !is_eof(rdr) {
bump(rdr);
}
}
fn push_blank_line_comment(rdr: string_reader, &comments: [cmnt]/~) {
#debug(">>> blank-line comment");
let v: [str]/~ = []/~;
comments += [{style: blank_line, lines: v, pos: rdr.chpos}]/~;
}
fn consume_whitespace_counting_blank_lines(rdr: string_reader,
&comments: [cmnt]/~) {
while is_whitespace(rdr.curr) && !is_eof(rdr) {
if rdr.col == 0u && rdr.curr == '\n' {
push_blank_line_comment(rdr, comments);
}
bump(rdr);
}
}
fn read_shebang_comment(rdr: string_reader, code_to_the_left: bool) -> cmnt {
#debug(">>> shebang comment");
let p = rdr.chpos;
#debug("<<< shebang comment");
ret {style: if code_to_the_left { trailing } else { isolated },
lines: [read_one_line_comment(rdr)]/~,
pos: p};
}
fn read_line_comments(rdr: string_reader, code_to_the_left: bool) -> cmnt {
#debug(">>> line comments");
let p = rdr.chpos;
let mut lines: [str]/~ = []/~;
while rdr.curr == '/' && nextch(rdr) == '/' {
let line = read_one_line_comment(rdr);
log(debug, line);
lines += [line]/~;
consume_non_eol_whitespace(rdr);
}
#debug("<<< line comments");
ret {style: if code_to_the_left { trailing } else { isolated },
lines: lines,
pos: p};
}
fn all_whitespace(s: str, begin: uint, end: uint) -> bool {
let mut i: uint = begin;
while i != end { if !is_whitespace(s[i] as char) { ret false; } i += 1u; }
ret true;
}
fn trim_whitespace_prefix_and_push_line(&lines: [str]/~,
s: str, col: uint) unsafe {
let mut s1;
let len = str::len(s);
if all_whitespace(s, 0u, uint::min(len, col)) {
if col < len {
s1 = str::slice(s, col, len);
} else { s1 = ""; }
} else { s1 = s; }
log(debug, "pushing line: " + s1);
lines += [s1]/~;
}
fn read_block_comment(rdr: string_reader, code_to_the_left: bool) -> cmnt {
#debug(">>> block comment");
let p = rdr.chpos;
let mut lines: [str]/~ = []/~;
let mut col: uint = rdr.col;
bump(rdr);
bump(rdr);
let mut curr_line = "/*";
let mut level: int = 1;
while level > 0 {
#debug("=== block comment level %d", level);
if is_eof(rdr) {(rdr as reader).fatal("unterminated block comment");}
if rdr.curr == '\n' {
trim_whitespace_prefix_and_push_line(lines, curr_line, col);
curr_line = "";
bump(rdr);
} else {
str::push_char(curr_line, rdr.curr);
if rdr.curr == '/' && nextch(rdr) == '*' {
bump(rdr);
bump(rdr);
curr_line += "*";
level += 1;
} else {
if rdr.curr == '*' && nextch(rdr) == '/' {
bump(rdr);
bump(rdr);
curr_line += "/";
level -= 1;
} else { bump(rdr); }
}
}
}
if str::len(curr_line) != 0u {
trim_whitespace_prefix_and_push_line(lines, curr_line, col);
}
let mut style = if code_to_the_left { trailing } else { isolated };
consume_non_eol_whitespace(rdr);
if !is_eof(rdr) && rdr.curr != '\n' && vec::len(lines) == 1u {
style = mixed;
}
#debug("<<< block comment");
ret {style: style, lines: lines, pos: p};
}
fn peeking_at_comment(rdr: string_reader) -> bool {
ret ((rdr.curr == '/' && nextch(rdr) == '/') ||
(rdr.curr == '/' && nextch(rdr) == '*')) ||
(rdr.curr == '#' && nextch(rdr) == '!');
}
fn consume_comment(rdr: string_reader, code_to_the_left: bool,
&comments: [cmnt]/~) {
#debug(">>> consume comment");
if rdr.curr == '/' && nextch(rdr) == '/' {
comments += [read_line_comments(rdr, code_to_the_left)]/~;
} else if rdr.curr == '/' && nextch(rdr) == '*' {
comments += [read_block_comment(rdr, code_to_the_left)]/~;
} else if rdr.curr == '#' && nextch(rdr) == '!' {
comments += [read_shebang_comment(rdr, code_to_the_left)]/~;
} else { fail; }
#debug("<<< consume comment");
}
type lit = {lit: str, pos: uint};
fn gather_comments_and_literals(span_diagnostic: diagnostic::span_handler,
path: str,
srdr: io::reader) ->
{cmnts: [cmnt]/~, lits: [lit]/~} {
let src = @str::from_bytes(srdr.read_whole_stream());
let itr = @interner::mk::<@str>(
{|x|str::hash(*x)},
{|x,y|str::eq(*x, *y)}
);
let rdr = lexer::new_low_level_string_reader
(span_diagnostic, codemap::new_filemap(path, src, 0u, 0u), itr);
let mut comments: [cmnt]/~ = []/~;
let mut literals: [lit]/~ = []/~;
let mut first_read: bool = true;
while !is_eof(rdr) {
loop {
let mut code_to_the_left = !first_read;
consume_non_eol_whitespace(rdr);
if rdr.curr == '\n' {
code_to_the_left = false;
consume_whitespace_counting_blank_lines(rdr, comments);
}
while peeking_at_comment(rdr) {
consume_comment(rdr, code_to_the_left, comments);
consume_whitespace_counting_blank_lines(rdr, comments);
}
break;
}
let bstart = rdr.pos;
rdr.next_token();
//discard, and look ahead; we're working with internal state
let {tok: tok, sp: sp} = rdr.peek();
if token::is_lit(tok) {
let s = get_str_from(rdr, bstart);
vec::push(literals, {lit: s, pos: sp.lo});
log(debug, "tok lit: " + s);
} else {
log(debug, "tok: " + token::to_str(*rdr.interner, tok));
}
first_read = false;
}
ret {cmnts: comments, lits: literals};
}
|
