/* Eats characters until '\n' and adds them to the comment. */
static void
eat_comment_line ()
{
for (;;)
{
int c = phase1_getc ();
if (c == '\n' || c == EOF)
{
comment_line_end (0);
break;
}
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
}
}
static int
phase2_getc ()
{
int c;
bool last_was_star;
c = phase1_getc ();
if (c != '/')
return c;
c = phase1_getc ();
switch (c)
{
default:
phase1_ungetc (c);
return '/';
case '*':
/* C comment. */
comment_start ();
last_was_star = false;
for (;;)
{
c = phase1_getc ();
if (c == EOF)
break;
/* We skip all leading white space, but not EOLs. */
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
switch (c)
{
case '\n':
comment_line_end (1);
comment_start ();
last_was_star = false;
continue;
case '*':
last_was_star = true;
continue;
case '/':
if (last_was_star)
{
comment_line_end (2);
break;
}
/* FALLTHROUGH */
default:
last_was_star = false;
continue;
}
break;
}
last_comment_line = line_number;
return ' ';
case '/':
/* C++ or ISO C 99 comment. */
comment_start ();
for (;;)
{
c = phase1_getc ();
if (c == '\n' || c == EOF)
break;
/* We skip all leading white space, but not EOLs. */
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
}
comment_line_end (0);
last_comment_line = line_number;
return '\n';
}
}
static int
phase3_getc ()
{
int lineno;
int c;
if (phase3_pushback_length)
return phase3_pushback[--phase3_pushback_length];
c = phase1_getc ();
if (c == '#')
{
/* sh comment. */
bool last_was_qmark = false;
comment_start ();
lineno = line_number;
for (;;)
{
c = phase1_getc ();
if (c == '\n' || c == EOF)
{
comment_line_end (0);
break;
}
if (last_was_qmark && c == '>')
{
comment_line_end (1);
skip_html ();
break;
}
/* We skip all leading white space, but not EOLs. */
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
last_was_qmark = (c == '?' || c == '%');
}
last_comment_line = lineno;
return '\n';
}
else if (c == '/')
{
c = phase1_getc ();
switch (c)
{
default:
phase1_ungetc (c);
return '/';
case '*':
{
/* C comment. */
bool last_was_star;
comment_start ();
lineno = line_number;
last_was_star = false;
for (;;)
{
c = phase1_getc ();
if (c == EOF)
break;
/* We skip all leading white space, but not EOLs. */
if (buflen == 0 && (c == ' ' || c == '\t'))
continue;
comment_add (c);
switch (c)
{
case '\n':
comment_line_end (1);
comment_start ();
lineno = line_number;
last_was_star = false;
continue;
case '*':
last_was_star = true;
continue;
case '/':
if (last_was_star)
{
comment_line_end (2);
break;
}
/* FALLTHROUGH */
default:
last_was_star = false;
continue;
}
break;
}
last_comment_line = lineno;
return ' ';
}
case '/':
{
/* C++ comment. */
bool last_was_qmark = false;
comment_start ();
lineno = line_number;
for (;;)
{
c = phase1_getc ();
if (c == '\n' || c == EOF)
{
comment_line_end (0);
break;
}
if (last_was_qmark && c == '>')
{
comment_line_end (1);
skip_html ();
break;
}
/* We skip all leading white space, but not EOLs. */
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
last_was_qmark = (c == '?' || c == '%');
}
last_comment_line = lineno;
return '\n';
}
}
}
else
return c;
}
static void
phase2_get (token_ty *tp)
{
static char *buffer;
static int bufmax;
int bufpos;
int c;
if (phase2_pushback_length)
{
*tp = phase2_pushback[--phase2_pushback_length];
return;
}
tp->string = NULL;
for (;;)
{
tp->line_number = line_number;
c = phase1_getc ();
switch (c)
{
case EOF:
tp->type = token_type_eof;
return;
case '"':
{
/* Comment. */
int lineno;
comment_start ();
lineno = line_number;
for (;;)
{
c = phase1_getc ();
if (c == '"' || c == EOF)
break;
if (c == '\n')
{
comment_line_end ();
comment_start ();
}
else
{
/* We skip all leading white space, but not EOLs. */
if (!(buflen == 0 && (c == ' ' || c == '\t')))
comment_add (c);
}
}
comment_line_end ();
last_comment_line = lineno;
continue;
}
case '\n':
if (last_non_comment_line > last_comment_line)
savable_comment_reset ();
/* FALLTHROUGH */
case ' ':
case '\t':
case '\r':
/* Ignore whitespace. */
continue;
}
last_non_comment_line = tp->line_number;
switch (c)
{
case '\'':
/* String literal. */
bufpos = 0;
for (;;)
{
c = phase1_getc ();
if (c == EOF)
break;
if (c == '\'')
{
c = phase1_getc ();
if (c != '\'')
{
phase1_ungetc (c);
break;
}
}
if (bufpos >= bufmax)
{
bufmax = 2 * bufmax + 10;
buffer = xrealloc (buffer, bufmax);
}
buffer[bufpos++] = c;
}
if (bufpos >= bufmax)
{
bufmax = 2 * bufmax + 10;
buffer = xrealloc (buffer, bufmax);
}
buffer[bufpos] = 0;
tp->type = token_type_string_literal;
tp->string = xstrdup (buffer);
return;
case '+':
case '-':
case '*':
case '/':
case '~':
case '|':
case ',':
case '<':
case '>':
case '=':
case '&':
case '@':
case '?':
case '%':
case '\\':
{
char *name;
int c2 = phase1_getc ();
switch (c2)
{
case '+':
case '-':
case '*':
case '/':
case '~':
case '|':
case ',':
case '<':
case '>':
case '=':
case '&':
case '@':
case '?':
case '%':
name = xmalloc (3);
name[0] = c;
name[1] = c2;
name[2] = '\0';
tp->type = token_type_symbol;
tp->string = name;
return;
default:
phase1_ungetc (c2);
break;
}
name = xmalloc (2);
name[0] = c;
name[1] = '\0';
tp->type = token_type_symbol;
tp->string = name;
return;
}
case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
case 'Y': case 'Z':
case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
case 's': case 't': case 'u': case 'v': case 'w': case 'x':
case 'y': case 'z':
/* Recognize id or id":"[id":"]* or id":"[id":"]*id. */
bufpos = 0;
for (;;)
{
if (bufpos >= bufmax)
{
bufmax = 2 * bufmax + 10;
buffer = xrealloc (buffer, bufmax);
}
buffer[bufpos++] = c;
c = phase1_getc ();
switch (c)
{
case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
case 'Y': case 'Z':
case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
case 's': case 't': case 'u': case 'v': case 'w': case 'x':
case 'y': case 'z':
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
continue;
case ':':
if (bufpos >= bufmax)
{
bufmax = 2 * bufmax + 10;
buffer = xrealloc (buffer, bufmax);
}
buffer[bufpos++] = c;
c = phase1_getc ();
switch (c)
{
case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
case 'G': case 'H': case 'I': case 'J': case 'K': case 'L':
case 'M': case 'N': case 'O': case 'P': case 'Q': case 'R':
case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':
case 'Y': case 'Z':
case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
case 'g': case 'h': case 'i': case 'j': case 'k': case 'l':
case 'm': case 'n': case 'o': case 'p': case 'q': case 'r':
case 's': case 't': case 'u': case 'v': case 'w': case 'x':
case 'y': case 'z':
continue;
default:
phase1_ungetc (c);
break;
}
break;
default:
phase1_ungetc (c);
break;
}
break;
}
if (bufpos >= bufmax)
{
bufmax = 2 * bufmax + 10;
buffer = xrealloc (buffer, bufmax);
}
buffer[bufpos] = '\0';
tp->string = xstrdup (buffer);
tp->type = token_type_symbol;
return;
case '#':
/* Uniquification operator. */
tp->type = token_type_uniq;
return;
case '$':
c = phase1_getc ();
tp->type = token_type_other;
return;
default:
tp->type = token_type_other;
return;
}
}
}
static int
phase2_getc ()
{
int c;
int lineno;
c = phase1_getc ();
if (c == '-')
{
c = phase1_getc ();
if (c == '-')
{
/* It starts with '--', so it must be either a short or a long
comment. */
c = phase1_getc ();
if (c == '[')
{
c = phase1_getc ();
int esigns = 0;
while (c == '=')
{
esigns++;
c = phase1_getc ();
}
if (c == '[')
{
/* Long comment. */
bool right_bracket = false;
bool end = false;
int esigns2 = 0;
lineno = line_number;
comment_start ();
while (!end)
{
c = phase1_getc ();
if (c == EOF)
break;
/* Ignore leading spaces and tabs. */
if (buflen == 0 && (c == ' ' || c == '\t'))
continue;
comment_add (c);
switch (c)
{
case ']':
if (!right_bracket)
{
right_bracket = true;
esigns2 = 0;
}
else
{
if (esigns2 == esigns)
{
comment_line_end (2 + esigns);
end = true;
}
}
break;
case '=':
if (right_bracket)
esigns2++;
break;
case '\n':
comment_line_end (1);
comment_start ();
lineno = line_number;
/* Intentionally not breaking. */
default:
right_bracket = false;
}
}
last_comment_line = lineno;
return ' ';
}
else
{
/* One line (short) comment, starting with '--[=...='. */
lineno = last_comment_line;
comment_start ();
comment_add ('[');
while (esigns--)
comment_add ('=');
phase1_ungetc (c);
eat_comment_line ();
last_comment_line = lineno;
return '\n';
}
}
else
{
/* One line (short) comment. */
lineno = line_number;
comment_start ();
phase1_ungetc (c);
eat_comment_line ();
last_comment_line = lineno;
return '\n';
}
}
else
{
/* Minus sign. */
phase1_ungetc (c);
return '-';
}
}
else
return c;
}
