1991-04-04 18:19:53 +00:00
|
|
|
/* This is the Assembler Pre-Processor
|
1994-01-04 23:35:56 +00:00
|
|
|
Copyright (C) 1987, 1990, 1991, 1992, 1994 Free Software Foundation, Inc.
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1992-02-13 08:33:54 +00:00
|
|
|
This file is part of GAS, the GNU Assembler.
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1992-02-13 08:33:54 +00:00
|
|
|
GAS is free software; you can redistribute it and/or modify
|
|
|
|
it under the terms of the GNU General Public License as published by
|
|
|
|
the Free Software Foundation; either version 2, or (at your option)
|
|
|
|
any later version.
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1992-02-13 08:33:54 +00:00
|
|
|
GAS is distributed in the hope that it will be useful,
|
|
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
GNU General Public License for more details.
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1992-02-13 08:33:54 +00:00
|
|
|
You should have received a copy of the GNU General Public License
|
|
|
|
along with GAS; see the file COPYING. If not, write to
|
|
|
|
the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
1993-09-10 16:01:07 +00:00
|
|
|
/* Modified by Allen Wirfs-Brock, Instantiations Inc 2/90 */
|
1991-04-04 18:19:53 +00:00
|
|
|
/* App, the assembler pre-processor. This pre-processor strips out excess
|
|
|
|
spaces, turns single-quoted characters into a decimal constant, and turns
|
1993-05-27 19:29:38 +00:00
|
|
|
# <number> <filename> <garbage> into a .line <number>\n.file <filename>
|
1992-12-15 22:55:01 +00:00
|
|
|
pair. This needs better error-handling.
|
1992-02-13 08:33:54 +00:00
|
|
|
*/
|
1991-04-04 18:19:53 +00:00
|
|
|
|
|
|
|
#include <stdio.h>
|
1992-11-23 20:39:57 +00:00
|
|
|
#include "as.h" /* For BAD_CASE() only */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
1994-01-04 23:35:56 +00:00
|
|
|
#if (__STDC__ != 1)
|
|
|
|
#ifndef const
|
|
|
|
#define const /* empty */
|
|
|
|
#endif
|
1991-04-04 18:19:53 +00:00
|
|
|
#endif
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
static char lex[256];
|
1992-12-29 23:53:52 +00:00
|
|
|
static const char symbol_chars[] =
|
1992-11-23 20:39:57 +00:00
|
|
|
"$._ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789";
|
1991-04-04 18:19:53 +00:00
|
|
|
|
|
|
|
#define LEX_IS_SYMBOL_COMPONENT 1
|
|
|
|
#define LEX_IS_WHITESPACE 2
|
|
|
|
#define LEX_IS_LINE_SEPARATOR 3
|
|
|
|
#define LEX_IS_COMMENT_START 4
|
|
|
|
#define LEX_IS_LINE_COMMENT_START 5
|
|
|
|
#define LEX_IS_TWOCHAR_COMMENT_1ST 6
|
|
|
|
#define LEX_IS_TWOCHAR_COMMENT_2ND 7
|
|
|
|
#define LEX_IS_STRINGQUOTE 8
|
|
|
|
#define LEX_IS_COLON 9
|
|
|
|
#define LEX_IS_NEWLINE 10
|
|
|
|
#define LEX_IS_ONECHAR_QUOTE 11
|
1992-02-13 08:33:54 +00:00
|
|
|
#define IS_SYMBOL_COMPONENT(c) (lex[c] == LEX_IS_SYMBOL_COMPONENT)
|
|
|
|
#define IS_WHITESPACE(c) (lex[c] == LEX_IS_WHITESPACE)
|
|
|
|
#define IS_LINE_SEPARATOR(c) (lex[c] == LEX_IS_LINE_SEPARATOR)
|
|
|
|
#define IS_COMMENT(c) (lex[c] == LEX_IS_COMMENT_START)
|
|
|
|
#define IS_LINE_COMMENT(c) (lex[c] == LEX_IS_LINE_COMMENT_START)
|
|
|
|
#define IS_NEWLINE(c) (lex[c] == LEX_IS_NEWLINE)
|
|
|
|
|
1993-11-28 06:54:04 +00:00
|
|
|
static int process_escape PARAMS ((int));
|
|
|
|
|
1992-02-13 08:33:54 +00:00
|
|
|
/* FIXME-soon: The entire lexer/parser thingy should be
|
|
|
|
built statically at compile time rather than dynamically
|
|
|
|
each and every time the assembler is run. xoxorich. */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
void
|
|
|
|
do_scrub_begin ()
|
|
|
|
{
|
|
|
|
const char *p;
|
|
|
|
|
|
|
|
lex[' '] = LEX_IS_WHITESPACE;
|
|
|
|
lex['\t'] = LEX_IS_WHITESPACE;
|
|
|
|
lex['\n'] = LEX_IS_NEWLINE;
|
|
|
|
lex[';'] = LEX_IS_LINE_SEPARATOR;
|
|
|
|
lex['"'] = LEX_IS_STRINGQUOTE;
|
1993-09-10 16:01:07 +00:00
|
|
|
#ifndef TC_HPPA
|
1992-11-23 20:39:57 +00:00
|
|
|
lex['\''] = LEX_IS_ONECHAR_QUOTE;
|
1993-09-10 16:01:07 +00:00
|
|
|
#endif
|
1992-11-23 20:39:57 +00:00
|
|
|
lex[':'] = LEX_IS_COLON;
|
1992-06-24 18:02:38 +00:00
|
|
|
|
1992-12-15 22:55:01 +00:00
|
|
|
|
|
|
|
|
|
|
|
#ifdef SINGLE_QUOTE_STRINGS
|
|
|
|
lex['\''] = LEX_IS_STRINGQUOTE;
|
1992-06-24 18:02:38 +00:00
|
|
|
#endif
|
1992-12-15 22:55:01 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
/* Note that these override the previous defaults, e.g. if ';'
|
1992-12-15 22:55:01 +00:00
|
|
|
|
1991-04-04 18:19:53 +00:00
|
|
|
is a comment char, then it isn't a line separator. */
|
1992-11-23 20:39:57 +00:00
|
|
|
for (p = symbol_chars; *p; ++p)
|
|
|
|
{
|
1993-09-10 16:01:07 +00:00
|
|
|
lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT;
|
1992-11-23 20:39:57 +00:00
|
|
|
} /* declare symbol characters */
|
|
|
|
|
|
|
|
for (p = comment_chars; *p; p++)
|
|
|
|
{
|
1993-09-10 16:01:07 +00:00
|
|
|
lex[(unsigned char) *p] = LEX_IS_COMMENT_START;
|
1992-11-23 20:39:57 +00:00
|
|
|
} /* declare comment chars */
|
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
for (p = line_comment_chars; *p; p++)
|
|
|
|
{
|
1993-09-10 16:01:07 +00:00
|
|
|
lex[(unsigned char) *p] = LEX_IS_LINE_COMMENT_START;
|
1993-05-27 19:29:38 +00:00
|
|
|
} /* declare line comment chars */
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
for (p = line_separator_chars; *p; p++)
|
|
|
|
{
|
1993-09-10 16:01:07 +00:00
|
|
|
lex[(unsigned char) *p] = LEX_IS_LINE_SEPARATOR;
|
1992-11-23 20:39:57 +00:00
|
|
|
} /* declare line separators */
|
|
|
|
|
|
|
|
/* Only allow slash-star comments if slash is not in use */
|
|
|
|
if (lex['/'] == 0)
|
|
|
|
{
|
|
|
|
lex['/'] = LEX_IS_TWOCHAR_COMMENT_1ST;
|
|
|
|
}
|
|
|
|
/* FIXME-soon. This is a bad hack but otherwise, we
|
1992-02-13 08:33:54 +00:00
|
|
|
can't do c-style comments when '/' is a line
|
|
|
|
comment char. xoxorich. */
|
1992-11-23 20:39:57 +00:00
|
|
|
if (lex['*'] == 0)
|
|
|
|
{
|
|
|
|
lex['*'] = LEX_IS_TWOCHAR_COMMENT_2ND;
|
|
|
|
}
|
|
|
|
} /* do_scrub_begin() */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
|
|
|
FILE *scrub_file;
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
int
|
|
|
|
scrub_from_file ()
|
|
|
|
{
|
|
|
|
return getc (scrub_file);
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
void
|
|
|
|
scrub_to_file (ch)
|
|
|
|
int ch;
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
ungetc (ch, scrub_file);
|
|
|
|
} /* scrub_to_file() */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
|
|
|
char *scrub_string;
|
|
|
|
char *scrub_last_string;
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
int
|
|
|
|
scrub_from_string ()
|
|
|
|
{
|
|
|
|
return scrub_string == scrub_last_string ? EOF : *scrub_string++;
|
|
|
|
} /* scrub_from_string() */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
void
|
|
|
|
scrub_to_string (ch)
|
|
|
|
int ch;
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
*--scrub_string = ch;
|
|
|
|
} /* scrub_to_string() */
|
1991-04-04 18:19:53 +00:00
|
|
|
|
|
|
|
/* Saved state of the scrubber */
|
|
|
|
static int state;
|
|
|
|
static int old_state;
|
|
|
|
static char *out_string;
|
|
|
|
static char out_buf[20];
|
|
|
|
static int add_newlines = 0;
|
|
|
|
|
|
|
|
/* Data structure for saving the state of app across #include's. Note that
|
|
|
|
app is called asynchronously to the parsing of the .include's, so our
|
|
|
|
state at the time .include is interpreted is completely unrelated.
|
|
|
|
That's why we have to save it all. */
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
struct app_save
|
|
|
|
{
|
|
|
|
int state;
|
|
|
|
int old_state;
|
|
|
|
char *out_string;
|
|
|
|
char out_buf[sizeof (out_buf)];
|
|
|
|
int add_newlines;
|
|
|
|
char *scrub_string;
|
|
|
|
char *scrub_last_string;
|
|
|
|
FILE *scrub_file;
|
|
|
|
};
|
|
|
|
|
|
|
|
char *
|
|
|
|
app_push ()
|
|
|
|
{
|
1992-06-24 18:02:38 +00:00
|
|
|
register struct app_save *saved;
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
saved = (struct app_save *) xmalloc (sizeof (*saved));
|
|
|
|
saved->state = state;
|
|
|
|
saved->old_state = old_state;
|
|
|
|
saved->out_string = out_string;
|
1993-09-10 16:01:07 +00:00
|
|
|
memcpy (saved->out_buf, out_buf, sizeof (out_buf));
|
1992-11-23 20:39:57 +00:00
|
|
|
saved->add_newlines = add_newlines;
|
|
|
|
saved->scrub_string = scrub_string;
|
1992-06-24 18:02:38 +00:00
|
|
|
saved->scrub_last_string = scrub_last_string;
|
1992-11-23 20:39:57 +00:00
|
|
|
saved->scrub_file = scrub_file;
|
1992-06-24 18:02:38 +00:00
|
|
|
|
|
|
|
/* do_scrub_begin() is not useful, just wastes time. */
|
1992-11-23 20:39:57 +00:00
|
|
|
return (char *) saved;
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
void
|
|
|
|
app_pop (arg)
|
|
|
|
char *arg;
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
register struct app_save *saved = (struct app_save *) arg;
|
|
|
|
|
|
|
|
/* There is no do_scrub_end (). */
|
|
|
|
state = saved->state;
|
|
|
|
old_state = saved->old_state;
|
|
|
|
out_string = saved->out_string;
|
1993-09-10 16:01:07 +00:00
|
|
|
memcpy (out_buf, saved->out_buf, sizeof (out_buf));
|
1992-11-23 20:39:57 +00:00
|
|
|
add_newlines = saved->add_newlines;
|
|
|
|
scrub_string = saved->scrub_string;
|
|
|
|
scrub_last_string = saved->scrub_last_string;
|
|
|
|
scrub_file = saved->scrub_file;
|
|
|
|
|
|
|
|
free (arg);
|
|
|
|
} /* app_pop() */
|
|
|
|
|
1992-12-29 23:53:52 +00:00
|
|
|
/* @@ This assumes that \n &c are the same on host and target. This is not
|
|
|
|
necessarily true. */
|
1993-11-28 06:54:04 +00:00
|
|
|
static int
|
1992-11-23 20:39:57 +00:00
|
|
|
process_escape (ch)
|
1993-11-28 06:54:04 +00:00
|
|
|
int ch;
|
1992-06-24 18:02:38 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
switch (ch)
|
|
|
|
{
|
|
|
|
case 'b':
|
|
|
|
return '\b';
|
|
|
|
case 'f':
|
|
|
|
return '\f';
|
|
|
|
case 'n':
|
|
|
|
return '\n';
|
|
|
|
case 'r':
|
|
|
|
return '\r';
|
|
|
|
case 't':
|
|
|
|
return '\t';
|
|
|
|
case '\'':
|
|
|
|
return '\'';
|
|
|
|
case '"':
|
1992-12-29 23:53:52 +00:00
|
|
|
return '\"';
|
1992-11-23 20:39:57 +00:00
|
|
|
default:
|
|
|
|
return ch;
|
|
|
|
}
|
1992-06-24 18:02:38 +00:00
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
int
|
|
|
|
do_scrub_next_char (get, unget)
|
|
|
|
int (*get) ();
|
|
|
|
void (*unget) ();
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
/*State 0: beginning of normal line
|
1992-02-13 08:33:54 +00:00
|
|
|
1: After first whitespace on line (flush more white)
|
|
|
|
2: After first non-white (opcode) on line (keep 1white)
|
|
|
|
3: after second white on line (into operands) (flush white)
|
|
|
|
4: after putting out a .line, put out digits
|
|
|
|
5: parsing a string, then go to old-state
|
|
|
|
6: putting out \ escape in a "d string.
|
1993-05-27 19:29:38 +00:00
|
|
|
7: After putting out a .appfile, put out string.
|
|
|
|
8: After putting out a .appfile string, flush until newline.
|
1993-03-18 00:52:37 +00:00
|
|
|
9: After seeing symbol char in state 3 (keep 1white after symchar)
|
1993-05-27 19:29:38 +00:00
|
|
|
10: After seeing whitespace in state 9 (keep white before symchar)
|
1992-02-13 08:33:54 +00:00
|
|
|
-1: output string in out_string and go to the state in old_state
|
|
|
|
-2: flush text until a '*' '/' is seen, then go to state old_state
|
|
|
|
*/
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
/* I added states 9 and 10 because the MIPS ECOFF assembler uses
|
|
|
|
constructs like ``.loc 1 20''. This was turning into ``.loc
|
|
|
|
120''. States 9 and 10 ensure that a space is never dropped in
|
|
|
|
between characters which could appear in a identifier. Ian
|
|
|
|
Taylor, ian@cygnus.com. */
|
1993-03-18 00:52:37 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
register int ch, ch2 = 0;
|
1993-11-28 06:54:04 +00:00
|
|
|
int not_cpp_line = 0;
|
1992-11-23 20:39:57 +00:00
|
|
|
|
|
|
|
switch (state)
|
|
|
|
{
|
|
|
|
case -1:
|
|
|
|
ch = *out_string++;
|
|
|
|
if (*out_string == 0)
|
|
|
|
{
|
|
|
|
state = old_state;
|
|
|
|
old_state = 3;
|
|
|
|
}
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
case -2:
|
|
|
|
for (;;)
|
|
|
|
{
|
|
|
|
do
|
|
|
|
{
|
|
|
|
ch = (*get) ();
|
|
|
|
}
|
|
|
|
while (ch != EOF && ch != '\n' && ch != '*');
|
|
|
|
if (ch == '\n' || ch == EOF)
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
/* At this point, ch must be a '*' */
|
|
|
|
while ((ch = (*get) ()) == '*')
|
|
|
|
{
|
|
|
|
;
|
|
|
|
}
|
|
|
|
if (ch == EOF || ch == '/')
|
|
|
|
break;
|
|
|
|
(*unget) (ch);
|
|
|
|
}
|
|
|
|
state = old_state;
|
|
|
|
return ' ';
|
|
|
|
|
|
|
|
case 4:
|
|
|
|
ch = (*get) ();
|
|
|
|
if (ch == EOF || (ch >= '0' && ch <= '9'))
|
|
|
|
return ch;
|
|
|
|
else
|
|
|
|
{
|
|
|
|
while (ch != EOF && IS_WHITESPACE (ch))
|
|
|
|
ch = (*get) ();
|
|
|
|
if (ch == '"')
|
|
|
|
{
|
|
|
|
(*unget) (ch);
|
1993-12-29 19:33:31 +00:00
|
|
|
out_string = "\n\t.appfile ";
|
1992-11-23 20:39:57 +00:00
|
|
|
old_state = 7;
|
|
|
|
state = -1;
|
|
|
|
return *out_string++;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
while (ch != EOF && ch != '\n')
|
|
|
|
ch = (*get) ();
|
1993-09-10 16:01:07 +00:00
|
|
|
state = 0;
|
1992-11-23 20:39:57 +00:00
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
case 5:
|
|
|
|
ch = (*get) ();
|
|
|
|
if (lex[ch] == LEX_IS_STRINGQUOTE)
|
|
|
|
{
|
|
|
|
state = old_state;
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
else if (ch == '\\')
|
|
|
|
{
|
|
|
|
state = 6;
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
else if (ch == EOF)
|
|
|
|
{
|
|
|
|
as_warn ("End of file in string: inserted '\"'");
|
|
|
|
state = old_state;
|
|
|
|
(*unget) ('\n');
|
|
|
|
return '"';
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
|
|
|
|
case 6:
|
|
|
|
state = 5;
|
|
|
|
ch = (*get) ();
|
|
|
|
switch (ch)
|
|
|
|
{
|
1992-12-29 23:53:52 +00:00
|
|
|
/* Handle strings broken across lines, by turning '\n' into
|
|
|
|
'\\' and 'n'. */
|
1992-11-23 20:39:57 +00:00
|
|
|
case '\n':
|
|
|
|
(*unget) ('n');
|
|
|
|
add_newlines++;
|
|
|
|
return '\\';
|
|
|
|
|
|
|
|
case '"':
|
|
|
|
case '\\':
|
|
|
|
case 'b':
|
|
|
|
case 'f':
|
|
|
|
case 'n':
|
|
|
|
case 'r':
|
|
|
|
case 't':
|
1991-04-04 18:19:53 +00:00
|
|
|
#ifdef BACKSLASH_V
|
1992-11-23 20:39:57 +00:00
|
|
|
case 'v':
|
1991-04-04 18:19:53 +00:00
|
|
|
#endif /* BACKSLASH_V */
|
1993-11-28 06:54:04 +00:00
|
|
|
case 'x':
|
|
|
|
case 'X':
|
1992-11-23 20:39:57 +00:00
|
|
|
case '0':
|
|
|
|
case '1':
|
|
|
|
case '2':
|
|
|
|
case '3':
|
|
|
|
case '4':
|
|
|
|
case '5':
|
|
|
|
case '6':
|
|
|
|
case '7':
|
|
|
|
break;
|
1992-06-24 18:02:38 +00:00
|
|
|
#if defined(IGNORE_NONSTANDARD_ESCAPES) | defined(ONLY_STANDARD_ESCAPES)
|
1992-11-23 20:39:57 +00:00
|
|
|
default:
|
|
|
|
as_warn ("Unknown escape '\\%c' in string: Ignored", ch);
|
|
|
|
break;
|
1991-04-04 18:19:53 +00:00
|
|
|
#else /* ONLY_STANDARD_ESCAPES */
|
1992-11-23 20:39:57 +00:00
|
|
|
default:
|
|
|
|
/* Accept \x as x for any x */
|
|
|
|
break;
|
1991-04-04 18:19:53 +00:00
|
|
|
#endif /* ONLY_STANDARD_ESCAPES */
|
1992-06-24 18:02:38 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
case EOF:
|
|
|
|
as_warn ("End of file in string: '\"' inserted");
|
|
|
|
return '"';
|
|
|
|
}
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
case 7:
|
|
|
|
ch = (*get) ();
|
|
|
|
state = 5;
|
|
|
|
old_state = 8;
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
case 8:
|
|
|
|
do
|
|
|
|
ch = (*get) ();
|
|
|
|
while (ch != '\n');
|
|
|
|
state = 0;
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
/* OK, we are somewhere in states 0 through 4 or 9 through 10 */
|
1992-11-23 20:39:57 +00:00
|
|
|
|
|
|
|
/* flushchar: */
|
|
|
|
ch = (*get) ();
|
|
|
|
recycle:
|
|
|
|
if (ch == EOF)
|
|
|
|
{
|
|
|
|
if (state != 0)
|
|
|
|
as_warn ("End of file not at end of a line: Newline inserted.");
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (lex[ch])
|
|
|
|
{
|
|
|
|
case LEX_IS_WHITESPACE:
|
|
|
|
do
|
1993-11-28 06:54:04 +00:00
|
|
|
/* Preserve a single whitespace character at the beginning of
|
|
|
|
a line. */
|
|
|
|
if (state == 0)
|
|
|
|
{
|
|
|
|
state = 1;
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
ch = (*get) ();
|
1992-11-23 20:39:57 +00:00
|
|
|
while (ch != EOF && IS_WHITESPACE (ch));
|
|
|
|
if (ch == EOF)
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
if (IS_COMMENT (ch) || (state == 0 && IS_LINE_COMMENT (ch)) || ch == '/' || IS_LINE_SEPARATOR (ch))
|
|
|
|
{
|
1993-11-28 06:54:04 +00:00
|
|
|
/* cpp never outputs a leading space before the #, so try to
|
|
|
|
avoid being confused. */
|
|
|
|
not_cpp_line = 1;
|
1992-11-23 20:39:57 +00:00
|
|
|
goto recycle;
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
1994-01-04 23:35:56 +00:00
|
|
|
#ifdef MRI
|
|
|
|
(*unget) (ch); /* Put back */
|
|
|
|
return ' '; /* Always return one space at start of line */
|
|
|
|
#endif
|
1992-11-23 20:39:57 +00:00
|
|
|
|
|
|
|
/* If we're in state 2, we've seen a non-white
|
1992-12-29 23:53:52 +00:00
|
|
|
character followed by whitespace. If the next
|
|
|
|
character is ':', this is whitespace after a label
|
|
|
|
name which we can ignore. */
|
1992-11-23 20:39:57 +00:00
|
|
|
if (state == 2 && lex[ch] == LEX_IS_COLON)
|
|
|
|
{
|
|
|
|
state = 0;
|
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (state)
|
|
|
|
{
|
|
|
|
case 0:
|
|
|
|
state++;
|
|
|
|
goto recycle; /* Punted leading sp */
|
|
|
|
case 1:
|
1993-11-28 06:54:04 +00:00
|
|
|
/* We can arrive here if we leave a leading whitespace character
|
|
|
|
at the beginning of a line. */
|
|
|
|
goto recycle;
|
1992-11-23 20:39:57 +00:00
|
|
|
case 2:
|
1993-03-18 00:52:37 +00:00
|
|
|
state = 3;
|
1992-11-23 20:39:57 +00:00
|
|
|
(*unget) (ch);
|
|
|
|
return ' '; /* Sp after opco */
|
|
|
|
case 3:
|
|
|
|
goto recycle; /* Sp in operands */
|
1993-05-27 19:29:38 +00:00
|
|
|
case 9:
|
|
|
|
case 10:
|
|
|
|
state = 10; /* Sp after symbol char */
|
|
|
|
goto recycle;
|
1992-11-23 20:39:57 +00:00
|
|
|
default:
|
|
|
|
BAD_CASE (state);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case LEX_IS_TWOCHAR_COMMENT_1ST:
|
|
|
|
ch2 = (*get) ();
|
|
|
|
if (ch2 != EOF && lex[ch2] == LEX_IS_TWOCHAR_COMMENT_2ND)
|
|
|
|
{
|
|
|
|
for (;;)
|
|
|
|
{
|
|
|
|
do
|
|
|
|
{
|
|
|
|
ch2 = (*get) ();
|
|
|
|
if (ch2 != EOF && IS_NEWLINE (ch2))
|
|
|
|
add_newlines++;
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
while (ch2 != EOF &&
|
|
|
|
(lex[ch2] != LEX_IS_TWOCHAR_COMMENT_2ND));
|
|
|
|
|
|
|
|
while (ch2 != EOF &&
|
|
|
|
(lex[ch2] == LEX_IS_TWOCHAR_COMMENT_2ND))
|
|
|
|
{
|
|
|
|
ch2 = (*get) ();
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
|
|
|
|
if (ch2 == EOF
|
|
|
|
|| lex[ch2] == LEX_IS_TWOCHAR_COMMENT_1ST)
|
1991-04-04 18:19:53 +00:00
|
|
|
break;
|
1992-11-23 20:39:57 +00:00
|
|
|
(*unget) (ch);
|
|
|
|
}
|
|
|
|
if (ch2 == EOF)
|
|
|
|
as_warn ("End of file in multiline comment");
|
|
|
|
|
|
|
|
ch = ' ';
|
|
|
|
goto recycle;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
if (ch2 != EOF)
|
|
|
|
(*unget) (ch2);
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 9 || state == 10)
|
|
|
|
state = 3;
|
1992-11-23 20:39:57 +00:00
|
|
|
return ch;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case LEX_IS_STRINGQUOTE:
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 9 || state == 10)
|
|
|
|
old_state = 3;
|
|
|
|
else
|
|
|
|
old_state = state;
|
1992-11-23 20:39:57 +00:00
|
|
|
state = 5;
|
|
|
|
return ch;
|
|
|
|
#ifndef MRI
|
1992-02-13 08:33:54 +00:00
|
|
|
#ifndef IEEE_STYLE
|
1992-11-23 20:39:57 +00:00
|
|
|
case LEX_IS_ONECHAR_QUOTE:
|
|
|
|
ch = (*get) ();
|
|
|
|
if (ch == EOF)
|
|
|
|
{
|
|
|
|
as_warn ("End-of-file after a one-character quote; \\000 inserted");
|
|
|
|
ch = 0;
|
|
|
|
}
|
|
|
|
if (ch == '\\')
|
|
|
|
{
|
|
|
|
ch = (*get) ();
|
|
|
|
ch = process_escape (ch);
|
|
|
|
}
|
|
|
|
sprintf (out_buf, "%d", (int) (unsigned char) ch);
|
1992-06-24 18:02:38 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
/* None of these 'x constants for us. We want 'x'. */
|
1992-11-23 20:39:57 +00:00
|
|
|
if ((ch = (*get) ()) != '\'')
|
|
|
|
{
|
1991-04-04 18:19:53 +00:00
|
|
|
#ifdef REQUIRE_CHAR_CLOSE_QUOTE
|
1992-11-23 20:39:57 +00:00
|
|
|
as_warn ("Missing close quote: (assumed)");
|
1991-04-04 18:19:53 +00:00
|
|
|
#else
|
1992-11-23 20:39:57 +00:00
|
|
|
(*unget) (ch);
|
1991-04-04 18:19:53 +00:00
|
|
|
#endif
|
1992-11-23 20:39:57 +00:00
|
|
|
}
|
|
|
|
if (strlen (out_buf) == 1)
|
|
|
|
{
|
|
|
|
return out_buf[0];
|
|
|
|
}
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 9 || state == 10)
|
|
|
|
old_state = 3;
|
|
|
|
else
|
|
|
|
old_state = state;
|
1992-11-23 20:39:57 +00:00
|
|
|
state = -1;
|
|
|
|
out_string = out_buf;
|
|
|
|
return *out_string++;
|
1992-06-24 18:02:38 +00:00
|
|
|
#endif
|
1992-02-13 08:33:54 +00:00
|
|
|
#endif
|
1992-11-23 20:39:57 +00:00
|
|
|
case LEX_IS_COLON:
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 9 || state == 10)
|
|
|
|
state = 3;
|
|
|
|
else if (state != 3)
|
1992-11-23 20:39:57 +00:00
|
|
|
state = 0;
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
case LEX_IS_NEWLINE:
|
|
|
|
/* Roll out a bunch of newlines from inside comments, etc. */
|
|
|
|
if (add_newlines)
|
|
|
|
{
|
|
|
|
--add_newlines;
|
|
|
|
(*unget) (ch);
|
|
|
|
}
|
|
|
|
/* fall thru into... */
|
|
|
|
|
|
|
|
case LEX_IS_LINE_SEPARATOR:
|
|
|
|
state = 0;
|
|
|
|
return ch;
|
|
|
|
|
|
|
|
case LEX_IS_LINE_COMMENT_START:
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 0) /* Only comment at start of line. */
|
1992-11-23 20:39:57 +00:00
|
|
|
{
|
1993-05-27 19:29:38 +00:00
|
|
|
/* FIXME-someday: The two character comment stuff was badly
|
|
|
|
thought out. On i386, we want '/' as line comment start
|
|
|
|
AND we want C style comments. hence this hack. The
|
|
|
|
whole lexical process should be reworked. xoxorich. */
|
|
|
|
if (ch == '/')
|
1993-03-18 00:52:37 +00:00
|
|
|
{
|
1993-05-27 19:29:38 +00:00
|
|
|
ch2 = (*get) ();
|
|
|
|
if (ch2 == '*')
|
|
|
|
{
|
|
|
|
state = -2;
|
|
|
|
return (do_scrub_next_char (get, unget));
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
(*unget) (ch2);
|
|
|
|
}
|
|
|
|
} /* bad hack */
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1993-11-28 06:54:04 +00:00
|
|
|
if (ch != '#')
|
|
|
|
not_cpp_line = 1;
|
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
do
|
1992-11-23 20:39:57 +00:00
|
|
|
ch = (*get) ();
|
1993-05-27 19:29:38 +00:00
|
|
|
while (ch != EOF && IS_WHITESPACE (ch));
|
1992-11-23 20:39:57 +00:00
|
|
|
if (ch == EOF)
|
1993-05-27 19:29:38 +00:00
|
|
|
{
|
|
|
|
as_warn ("EOF in comment: Newline inserted");
|
|
|
|
return '\n';
|
|
|
|
}
|
1993-11-28 06:54:04 +00:00
|
|
|
if (ch < '0' || ch > '9' || not_cpp_line)
|
1993-05-27 19:29:38 +00:00
|
|
|
{
|
|
|
|
/* Non-numerics: Eat whole comment line */
|
|
|
|
while (ch != EOF && !IS_NEWLINE (ch))
|
|
|
|
ch = (*get) ();
|
|
|
|
if (ch == EOF)
|
|
|
|
as_warn ("EOF in Comment: Newline inserted");
|
|
|
|
state = 0;
|
|
|
|
return '\n';
|
|
|
|
}
|
|
|
|
/* Numerics begin comment. Perhaps CPP `# 123 "filename"' */
|
|
|
|
(*unget) (ch);
|
|
|
|
old_state = 4;
|
|
|
|
state = -1;
|
1993-12-29 19:33:31 +00:00
|
|
|
out_string = "\t.appline ";
|
1993-05-27 19:29:38 +00:00
|
|
|
return *out_string++;
|
1992-11-23 20:39:57 +00:00
|
|
|
}
|
|
|
|
|
1993-05-27 19:29:38 +00:00
|
|
|
/* We have a line comment character which is not at the start of
|
|
|
|
a line. If this is also a normal comment character, fall
|
|
|
|
through. Otherwise treat it as a default character. */
|
|
|
|
if (strchr (comment_chars, ch) == NULL)
|
|
|
|
goto de_fault;
|
|
|
|
/* Fall through. */
|
1992-11-23 20:39:57 +00:00
|
|
|
case LEX_IS_COMMENT_START:
|
|
|
|
do
|
|
|
|
ch = (*get) ();
|
|
|
|
while (ch != EOF && !IS_NEWLINE (ch));
|
|
|
|
if (ch == EOF)
|
|
|
|
as_warn ("EOF in comment: Newline inserted");
|
|
|
|
state = 0;
|
|
|
|
return '\n';
|
|
|
|
|
1993-03-18 00:52:37 +00:00
|
|
|
case LEX_IS_SYMBOL_COMPONENT:
|
1993-05-27 19:29:38 +00:00
|
|
|
if (state == 10)
|
|
|
|
{
|
|
|
|
/* This is a symbol character following another symbol
|
|
|
|
character, with whitespace in between. We skipped the
|
|
|
|
whitespace earlier, so output it now. */
|
|
|
|
(*unget) (ch);
|
|
|
|
state = 3;
|
|
|
|
return ' ';
|
|
|
|
}
|
1993-03-18 00:52:37 +00:00
|
|
|
if (state == 3)
|
|
|
|
state = 9;
|
|
|
|
/* Fall through. */
|
1992-11-23 20:39:57 +00:00
|
|
|
default:
|
|
|
|
de_fault:
|
|
|
|
/* Some relatively `normal' character. */
|
|
|
|
if (state == 0)
|
|
|
|
{
|
|
|
|
state = 2; /* Now seeing opcode */
|
|
|
|
return ch;
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
else if (state == 1)
|
|
|
|
{
|
|
|
|
state = 2; /* Ditto */
|
|
|
|
return ch;
|
|
|
|
}
|
1993-03-18 00:52:37 +00:00
|
|
|
else if (state == 9)
|
|
|
|
{
|
|
|
|
if (lex[ch] != LEX_IS_SYMBOL_COMPONENT)
|
|
|
|
state = 3;
|
|
|
|
return ch;
|
|
|
|
}
|
1993-05-27 19:29:38 +00:00
|
|
|
else if (state == 10)
|
|
|
|
{
|
|
|
|
state = 3;
|
|
|
|
return ch;
|
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
else
|
|
|
|
{
|
|
|
|
return ch; /* Opcode or operands already */
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return -1;
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef TEST
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
const char comment_chars[] = "|";
|
|
|
|
const char line_comment_chars[] = "#";
|
1991-04-04 18:19:53 +00:00
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
main ()
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
int ch;
|
|
|
|
|
|
|
|
app_begin ();
|
|
|
|
while ((ch = do_scrub_next_char (stdin)) != EOF)
|
|
|
|
putc (ch, stdout);
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
|
|
|
|
1992-11-23 20:39:57 +00:00
|
|
|
as_warn (str)
|
|
|
|
char *str;
|
1991-04-04 18:19:53 +00:00
|
|
|
{
|
1992-11-23 20:39:57 +00:00
|
|
|
fputs (str, stderr);
|
|
|
|
putc ('\n', stderr);
|
1991-04-04 18:19:53 +00:00
|
|
|
}
|
1992-11-23 20:39:57 +00:00
|
|
|
|
1991-04-04 18:19:53 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
/* end of app.c */
|