aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorNeil Booth <neilb@earthling.net>2000-04-25 11:57:57 +0000
committerNeil Booth <neil@gcc.gnu.org>2000-04-25 11:57:57 +0000
commit05183359dab76cc72b7c149b44b79012e8234b88 (patch)
treef6054d71feac30674ea9d0da03b21c1292f844cb
parentac79db28f447f57b159b40fd7e05de089d204a55 (diff)
downloadgcc-05183359dab76cc72b7c149b44b79012e8234b88.zip
gcc-05183359dab76cc72b7c149b44b79012e8234b88.tar.gz
gcc-05183359dab76cc72b7c149b44b79012e8234b88.tar.bz2
* Revert previous patch until build failure cause determined.
From-SVN: r33411
-rw-r--r--gcc/ChangeLog5
-rw-r--r--gcc/cpphash.c11
-rw-r--r--gcc/cpphash.h1
-rw-r--r--gcc/cpplex.c1382
-rw-r--r--gcc/cpplib.c23
-rw-r--r--gcc/cpplib.h123
6 files changed, 75 insertions, 1470 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index ac5d46d..a09d530 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,8 @@
+2000-04-25 Neil Booth <NeilB@earthling.net>
+
+ * Revert my patch below until cause of build failures
+ determined.
+
2000-04-25 Richard Henderson <rth@cygnus.com>
* combine.c (combine_instructions): Add missing argument
diff --git a/gcc/cpphash.c b/gcc/cpphash.c
index ac8294d..5817e964 100644
--- a/gcc/cpphash.c
+++ b/gcc/cpphash.c
@@ -459,8 +459,7 @@ collect_objlike_expansion (pfile, list)
default:;
}
- if (i > 1 && !last_was_paste
- && (list->tokens[i].flags & PREV_WHITESPACE))
+ if (i > 1 && !last_was_paste && (list->tokens[i].flags & HSPACE_BEFORE))
CPP_PUTC (pfile, ' ');
CPP_PUTS (pfile,
@@ -572,10 +571,10 @@ collect_funlike_expansion (pfile, list, arglist, replacement)
}
if (last_token != PASTE && last_token != START
- && (list->tokens[i].flags & PREV_WHITESPACE))
+ && (list->tokens[i].flags & HSPACE_BEFORE))
CPP_PUTC (pfile, ' ');
if (last_token == ARG && CPP_TRADITIONAL (pfile)
- && !(list->tokens[i].flags & PREV_WHITESPACE))
+ && !(list->tokens[i].flags & HSPACE_BEFORE))
endpat->raw_after = 1;
switch (token)
@@ -617,7 +616,7 @@ collect_funlike_expansion (pfile, list, arglist, replacement)
{
int raw_before = (last_token == PASTE
|| (CPP_TRADITIONAL (pfile)
- && !(list->tokens[i].flags & PREV_WHITESPACE)));
+ && !(list->tokens[i].flags & HSPACE_BEFORE)));
add_pat (&pat, &endpat,
CPP_WRITTEN (pfile) - last /* nchars */, j /* argno */,
@@ -866,7 +865,7 @@ _cpp_create_definition (pfile, list, hp)
/* The macro is function-like only if the next character,
with no intervening whitespace, is '('. */
else if (list->tokens[1].type == CPP_OPEN_PAREN
- && ! (list->tokens[1].flags & PREV_WHITESPACE))
+ && ! (list->tokens[1].flags & HSPACE_BEFORE))
{
struct arglist args;
int replacement;
diff --git a/gcc/cpphash.h b/gcc/cpphash.h
index 2d2ea8d..78185f2 100644
--- a/gcc/cpphash.h
+++ b/gcc/cpphash.h
@@ -317,6 +317,5 @@ extern void _cpp_scan_line PARAMS ((cpp_reader *, cpp_toklist *));
/* In cpplib.c */
extern int _cpp_handle_directive PARAMS ((cpp_reader *));
extern void _cpp_handle_eof PARAMS ((cpp_reader *));
-extern void _cpp_check_directive PARAMS((cpp_toklist *, cpp_token *));
#endif
diff --git a/gcc/cpplex.c b/gcc/cpplex.c
index eceef7d..f46b638 100644
--- a/gcc/cpplex.c
+++ b/gcc/cpplex.c
@@ -4,7 +4,6 @@
Based on CCCP program by Paul Rubin, June 1986
Adapted to ANSI C, Richard Stallman, Jan 1987
Broken out to separate file, Zack Weinberg, Mar 2000
- Single-pass line tokenization by Neil Booth, April 2000
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
@@ -55,15 +54,12 @@ static void output_line_command PARAMS ((cpp_reader *, cpp_printer *,
unsigned int));
static void bump_column PARAMS ((cpp_printer *, unsigned int,
unsigned int));
-static void expand_name_space PARAMS ((cpp_toklist *, unsigned int));
+static void expand_name_space PARAMS ((cpp_toklist *));
static void expand_token_space PARAMS ((cpp_toklist *));
static void init_token_list PARAMS ((cpp_reader *, cpp_toklist *, int));
static void pedantic_whitespace PARAMS ((cpp_reader *, U_CHAR *,
unsigned int));
-#define auto_expand_name_space(list) \
- expand_name_space ((list), (list)->name_cap / 2)
-
/* Re-allocates PFILE->token_buffer so it will hold at least N more chars. */
void
@@ -435,12 +431,12 @@ cpp_file_buffer (pfile)
/* Expand a token list's string space. */
static void
-expand_name_space (list, len)
+expand_name_space (list)
cpp_toklist *list;
- unsigned int len;
-{
- list->name_cap += len;
- list->namebuf = (unsigned char *) xrealloc (list->namebuf, list->name_cap);
+{
+ list->name_cap *= 2;
+ list->namebuf = (unsigned char *) xrealloc (list->namebuf,
+ list->name_cap);
}
/* Expand the number of tokens in a list. */
@@ -453,38 +449,33 @@ expand_token_space (list)
xrealloc (list->tokens, list->tokens_cap * sizeof (cpp_token));
}
-/* Initialize a token list. We leave the first token unused, as this
- allows us to always peek at the previous token without worrying
- about underflowing the list. */
+/* Initialise a token list. */
static void
init_token_list (pfile, list, recycle)
cpp_reader *pfile;
cpp_toklist *list;
int recycle;
{
- /* Recycling a used list saves 3 free-malloc pairs. */
- if (!recycle)
+ /* Recycling a used list saves 2 free-malloc pairs. */
+ if (recycle)
{
- /* Initialize token space. */
- list->tokens_cap = 256; /* 4K's worth. */
+ list->tokens_used = 0;
+ list->name_used = 0;
+ }
+ else
+ {
+ /* Initialise token space. */
+ list->tokens_cap = 256; /* 4K on Intel. */
+ list->tokens_used = 0;
list->tokens = (cpp_token *)
xmalloc (list->tokens_cap * sizeof (cpp_token));
- /* Initialize name space. */
+ /* Initialise name space. */
list->name_cap = 1024;
+ list->name_used = 0;
list->namebuf = (unsigned char *) xmalloc (list->name_cap);
-
- /* Only create a comment space on demand. */
- list->comments_cap = 0;
- list->comments = 0;
}
- /* Put a dummy token at the start that will fail matches. */
- list->tokens[0].type = CPP_EOF;
-
- list->tokens_used = 1;
- list->name_used = 0;
- list->comments_used = 0;
if (pfile->buffer)
list->line = pfile->buffer->lineno;
list->dir_handler = 0;
@@ -531,7 +522,7 @@ _cpp_scan_line (pfile, list)
if (list->tokens_used >= list->tokens_cap)
expand_token_space (list);
if (list->name_used + len >= list->name_cap)
- auto_expand_name_space (list);
+ expand_name_space (list);
if (type == CPP_MACRO)
type = CPP_NAME;
@@ -539,7 +530,7 @@ _cpp_scan_line (pfile, list)
list->tokens_used++;
list->tokens[i].type = type;
list->tokens[i].col = col;
- list->tokens[i].flags = space_before ? PREV_WHITESPACE : 0;
+ list->tokens[i].flags = space_before ? HSPACE_BEFORE : 0;
if (type == CPP_VSPACE)
break;
@@ -2046,1334 +2037,3 @@ _cpp_init_input_buffer (pfile)
pfile->input_buffer = tmp;
pfile->input_buffer_len = 8192;
}
-
-#if 0
-
-static void expand_comment_space PARAMS ((cpp_toklist *));
-void init_trigraph_map PARAMS ((void));
-static unsigned char* trigraph_replace PARAMS ((cpp_reader *, unsigned char *,
- unsigned char *));
-static const unsigned char *backslash_start PARAMS ((cpp_reader *,
- const unsigned char *));
-static int skip_block_comment PARAMS ((cpp_reader *));
-static int skip_line_comment PARAMS ((cpp_reader *));
-static void skip_whitespace PARAMS ((cpp_reader *, int));
-static void parse_name PARAMS ((cpp_reader *, cpp_toklist *, cpp_name *));
-static void parse_number PARAMS ((cpp_reader *, cpp_toklist *, cpp_name *));
-static void parse_string PARAMS ((cpp_reader *, cpp_toklist *, cpp_name *,
- unsigned int));
-static int trigraph_ok PARAMS ((cpp_reader *, const unsigned char *));
-static void copy_comment PARAMS ((cpp_toklist *, const unsigned char *,
- unsigned int, unsigned int, unsigned int));
-void _cpp_lex_line PARAMS ((cpp_reader *, cpp_toklist *));
-
-static void _cpp_output_list PARAMS ((cpp_reader *, cpp_toklist *));
-
-unsigned int spell_char PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *token));
-unsigned int spell_string PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *token));
-unsigned int spell_comment PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *token));
-unsigned int spell_name PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *token));
-unsigned int spell_other PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *token));
-
-typedef unsigned int (* speller) PARAMS ((unsigned char *, cpp_toklist *,
- cpp_token *));
-
-/* Macros on a cpp_name. */
-#define INIT_NAME(list, name) \
- do {(name).len = 0; (name).offset = (list)->name_used;} while (0)
-
-/* Careful: 1-based. First token is NTH_TOKEN (1). */
-#define NTH_TOKEN(list, n) ((list)->tokens[n])
-#define IS_DIRECTIVE(list) (NTH_TOKEN(list, 1).type == CPP_HASH)
-#define COLUMN(cur) ((cur) - buffer->line_base)
-
-/* Maybe put these in the ISTABLE eventually. */
-#define IS_HSPACE(c) ((c) == ' ' || (c) == '\t')
-#define IS_NEWLINE(c) ((c) == '\n' || (c) == '\r')
-
-/* Handle LF, CR, CR-LF and LF-CR style newlines. Assumes next
- character, if any, is in buffer. */
-#define handle_newline(cur, limit, c) \
- do {\
- if ((cur) < (limit) && *(cur) == '\r' + '\n' - c) \
- (cur)++; \
- CPP_BUMP_LINE_CUR (pfile, (cur)); \
- } while (0)
-
-#define IMMED_TOKEN() (!(cur_token->flags & PREV_WHITESPACE))
-#define PREV_TOKEN_TYPE (cur_token[-1].type)
-
-#define SPELL_TEXT 0
-#define SPELL_HANDLER 1
-#define SPELL_NONE 2
-#define SPELL_EOL 3
-
-#define T(e, s) {SPELL_TEXT, s},
-#define H(e, s) {SPELL_HANDLER, s},
-#define N(e, s) {SPELL_NONE, s},
-#define E(e, s) {SPELL_EOL, s},
-
-static const struct token_spelling
-{
- char type;
- PTR speller;
-} token_spellings [N_TTYPES + 1] = {TTYPE_TABLE {0, 0} };
-
-#undef T
-#undef H
-#undef N
-#undef E
-
-static const unsigned char *digraph_spellings [] = {"%:", "%:%:", "<:",
- ":>", "<%", "%>"};
-
-static void
-expand_comment_space (list)
- cpp_toklist *list;
-{
- if (list->comments_cap == 0)
- {
- list->comments_cap = 10;
- list->comments = (cpp_token *)
- xmalloc (list->comments_cap * sizeof (cpp_token));
- }
- else
- {
- list->comments_cap *= 2;
- list->comments = (cpp_token *)
- xrealloc (list->comments, list->comments_cap);
- }
-}
-
-void
-cpp_free_token_list (list)
- cpp_toklist *list;
-{
- if (list->comments)
- free (list->comments);
- free (list->tokens);
- free (list->namebuf);
- free (list);
-}
-
-static char trigraph_map[256];
-
-void
-init_trigraph_map ()
-{
- trigraph_map['='] = '#';
- trigraph_map['('] = '[';
- trigraph_map[')'] = ']';
- trigraph_map['/'] = '\\';
- trigraph_map['\''] = '^';
- trigraph_map['<'] = '{';
- trigraph_map['>'] = '}';
- trigraph_map['!'] = '|';
- trigraph_map['-'] = '~';
-}
-
-/* Call when a trigraph is encountered. It warns if necessary, and
- returns true if the trigraph should be honoured. END is the third
- character of a trigraph in the input stream. */
-static int
-trigraph_ok (pfile, end)
- cpp_reader *pfile;
- const unsigned char *end;
-{
- int accept = CPP_OPTION (pfile, trigraphs);
-
- if (CPP_OPTION (pfile, warn_trigraphs))
- {
- unsigned int col = end - 1 - pfile->buffer->line_base;
- if (accept)
- cpp_warning_with_line (pfile, pfile->buffer->lineno, col,
- "trigraph ??%c converted to %c",
- (int) *end, (int) trigraph_map[*end]);
- else
- cpp_warning_with_line (pfile, pfile->buffer->lineno, col,
- "trigraph ??%c ignored", (int) *end);
- }
- return accept;
-}
-
-/* Scan a string for trigraphs, warning or replacing them inline as
- appropriate. When parsing a string, we must call this routine
- before processing a newline character (if trigraphs are enabled),
- since the newline might be escaped by a preceding backslash
- trigraph sequence. Returns a pointer to the end of the name after
- replacement. */
-
-static unsigned char*
-trigraph_replace (pfile, src, limit)
- cpp_reader *pfile;
- unsigned char *src;
- unsigned char* limit;
-{
- unsigned char *dest;
-
- /* Starting with src[1], find two consecutive '?'. The case of no
- trigraphs is streamlined. */
-
- for (; src + 1 < limit; src += 2)
- {
- if (src[0] != '?')
- continue;
-
- /* Make src point to the 1st (NOT 2nd) of two consecutive '?'s. */
- if (src[-1] == '?')
- src--;
- else if (src + 2 == limit || src[1] != '?')
- continue;
-
- /* Check if it really is a trigraph. */
- if (trigraph_map[src[2]] == 0)
- continue;
-
- dest = src;
- goto trigraph_found;
- }
- return limit;
-
- /* Now we have a trigraph, we need to scan the remaining buffer, and
- copy-shifting its contents left if replacement is enabled. */
- for (; src + 2 < limit; dest++, src++)
- if ((*dest = *src) == '?' && src[1] == '?' && trigraph_map[src[2]])
- {
- trigraph_found:
- src += 2;
- if (trigraph_ok (pfile, pfile->buffer->cur - (limit - src)))
- *dest = trigraph_map[*src];
- }
-
- /* Copy remaining (at most 2) characters. */
- while (src < limit)
- *dest++ = *src++;
- return dest;
-}
-
-/* If CUR is a backslash or the end of a trigraphed backslash, return
- a pointer to its beginning, otherwise NULL. We don't read beyond
- the buffer start, because there is the start of the comment in the
- buffer. */
-static const unsigned char *
-backslash_start (pfile, cur)
- cpp_reader *pfile;
- const unsigned char *cur;
-{
- if (cur[0] == '\\')
- return cur;
- if (cur[0] == '/' && cur[-1] == '?' && cur[-2] == '?'
- && trigraph_ok (pfile, cur))
- return cur - 2;
- return 0;
-}
-
-/* Skip a C-style block comment. This is probably the trickiest
- handler. We find the end of the comment by seeing if an asterisk
- is before every '/' we encounter. The nasty complication is that a
- previous asterisk may be separated by one or more escaped newlines.
- Returns non-zero if comment terminated by EOF, zero otherwise. */
-static int
-skip_block_comment (pfile)
- cpp_reader *pfile;
-{
- cpp_buffer *buffer = pfile->buffer;
- const unsigned char *char_after_star = 0;
- register const unsigned char *cur = buffer->cur;
- int seen_eof = 0;
-
- /* Inner loop would think the comment has ended if the first comment
- character is a '/'. Avoid this and keep the inner loop clean by
- skipping such a character. */
- if (cur < buffer->rlimit && cur[0] == '/')
- cur++;
-
- for (; cur < buffer->rlimit; )
- {
- unsigned char c = *cur++;
-
- /* People like decorating comments with '*', so check for
- '/' instead for efficiency. */
- if (c == '/')
- {
- if (cur[-2] == '*' || cur - 1 == char_after_star)
- goto out;
-
- /* Warn about potential nested comments, but not when
- the final character inside the comment is a '/'.
- Don't bother to get it right across escaped newlines. */
- if (CPP_OPTION (pfile, warn_comments) && cur + 1 < buffer->rlimit
- && cur[0] == '*' && cur[1] != '/')
- {
- buffer->cur = cur;
- cpp_warning (pfile, "'/*' within comment");
- }
- }
- else if (IS_NEWLINE(c))
- {
- const unsigned char* bslash = backslash_start (pfile, cur - 2);
-
- handle_newline (cur, buffer->rlimit, c);
- /* Work correctly if there is an asterisk before an
- arbirtrarily long sequence of escaped newlines. */
- if (bslash && (bslash[-1] == '*' || bslash == char_after_star))
- char_after_star = cur;
- else
- char_after_star = 0;
- }
- }
- seen_eof = 1;
-
- out:
- buffer->cur = cur;
- return seen_eof;
-}
-
-/* Skip a C++ or Chill line comment. Handles escaped newlines.
- Returns non-zero if a multiline comment. */
-static int
-skip_line_comment (pfile)
- cpp_reader *pfile;
-{
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
- int multiline = 0;
-
- for (; cur < buffer->rlimit; )
- {
- unsigned char c = *cur++;
-
- if (IS_NEWLINE (c))
- {
- /* Check for a (trigaph?) backslash escaping the newline. */
- if (!backslash_start (pfile, cur - 2))
- goto out;
- multiline = 1;
- handle_newline (cur, buffer->rlimit, c);
- }
- }
- cur++;
-
- out:
- buffer->cur = cur - 1; /* Leave newline for caller. */
- return multiline;
-}
-
-/* Skips whitespace, stopping at next non-whitespace character. */
-static void
-skip_whitespace (pfile, in_directive)
- cpp_reader *pfile;
- int in_directive;
-{
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
- unsigned short null_count = 0;
-
- for (; cur < buffer->rlimit; )
- {
- unsigned char c = *cur++;
-
- if (IS_HSPACE(c)) /* FIXME: Fix ISTABLE. */
- continue;
- if (!is_space(c) || IS_NEWLINE (c)) /* Main loop handles newlines. */
- goto out;
- if (c == '\0')
- null_count++;
- /* Mut be '\f' or '\v' */
- else if (in_directive && CPP_PEDANTIC (pfile))
- cpp_pedwarn (pfile, "%s in preprocessing directive",
- c == '\f' ? "formfeed" : "vertical tab");
- }
- cur++;
-
- out:
- buffer->cur = cur - 1;
- if (null_count)
- cpp_warning (pfile, null_count > 1 ? "embedded null characters ignored"
- : "embedded null character ignored");
-}
-
-/* Parse (append) an identifier. */
-static void
-parse_name (pfile, list, name)
- cpp_reader *pfile;
- cpp_toklist *list;
- cpp_name *name;
-{
- const unsigned char *name_limit;
- unsigned char *namebuf;
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
-
- expanded:
- name_limit = list->namebuf + list->name_cap;
- namebuf = list->namebuf + list->name_used;
-
- for (; cur < buffer->rlimit && namebuf < name_limit; )
- {
- unsigned char c = *namebuf = *cur; /* Copy a single char. */
-
- if (! is_idchar(c))
- goto out;
- namebuf++;
- cur++;
- if (c == '$' && CPP_PEDANTIC (pfile))
- {
- buffer->cur = cur;
- cpp_pedwarn (pfile, "'$' character in identifier");
- }
- }
-
- /* Run out of name space? */
- if (cur < buffer->rlimit)
- {
- list->name_used = namebuf - list->namebuf;
- auto_expand_name_space (list);
- goto expanded;
- }
-
- out:
- buffer->cur = cur;
- name->len = namebuf - (list->namebuf + name->offset);
- list->name_used = namebuf - list->namebuf;
-}
-
-/* Parse (append) a number. */
-
-#define VALID_SIGN(c, prevc) \
- (((c) == '+' || (c) == '-') && \
- ((prevc) == 'e' || (prevc) == 'E' \
- || (((prevc) == 'p' || (prevc) == 'P') && !CPP_OPTION (pfile, c89))))
-
-static void
-parse_number (pfile, list, name)
- cpp_reader *pfile;
- cpp_toklist *list;
- cpp_name *name;
-{
- const unsigned char *name_limit;
- unsigned char *namebuf;
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
-
- expanded:
- name_limit = list->namebuf + list->name_cap;
- namebuf = list->namebuf + list->name_used;
-
- for (; cur < buffer->rlimit && namebuf < name_limit; )
- {
- unsigned char c = *namebuf = *cur; /* Copy a single char. */
-
- /* Perhaps we should accept '$' here if we accept it for
- identifiers. We know namebuf[-1] is safe, because for c to
- be a sign we must have pushed at least one character. */
- if (!is_numchar (c) && c != '.' && ! VALID_SIGN (c, namebuf[-1]))
- goto out;
-
- namebuf++;
- cur++;
- }
-
- /* Run out of name space? */
- if (cur < buffer->rlimit)
- {
- list->name_used = namebuf - list->namebuf;
- auto_expand_name_space (list);
- goto expanded;
- }
-
- out:
- buffer->cur = cur;
- name->len = namebuf - (list->namebuf + name->offset);
- list->name_used = namebuf - list->namebuf;
-}
-
-/* Places a string terminated by an unescaped TERMINATOR into a
- cpp_name, which should be expandable and thus at the top of the
- list's stack. Handles embedded trigraphs, if necessary, and
- escaped newlines.
-
- Can be used for character constants (terminator = '\''), string
- constants ('"'), angled headers ('>') and assertions (')'). */
-
-static void
-parse_string (pfile, list, name, terminator)
- cpp_reader *pfile;
- cpp_toklist *list;
- cpp_name *name;
- unsigned int terminator;
-{
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
- const unsigned char *name_limit;
- unsigned char *namebuf;
- unsigned int null_count = 0;
- int trigraphed_len = 0;
-
- expanded:
- name_limit = list->namebuf + list->name_cap;
- namebuf = list->namebuf + list->name_used;
-
- for (; cur < buffer->rlimit && namebuf < name_limit; )
- {
- unsigned int c = *namebuf++ = *cur++; /* Copy a single char. */
-
- if (c == '\0')
- null_count++;
- else if (c == terminator || IS_NEWLINE (c))
- {
- unsigned char* name_start = list->namebuf + name->offset;
-
- /* Needed for trigraph_replace and multiline string warning. */
- buffer->cur = cur;
-
- /* Scan for trigraphs before checking if backslash-escaped. */
- if (CPP_OPTION (pfile, trigraphs)
- || CPP_OPTION (pfile, warn_trigraphs))
- {
- namebuf = trigraph_replace (pfile, name_start + trigraphed_len,
- namebuf);
- trigraphed_len = namebuf - 2 - (name_start + trigraphed_len);
- if (trigraphed_len < 0)
- trigraphed_len = 0;
- }
-
- namebuf--; /* Drop the newline / terminator from the name. */
- if (IS_NEWLINE (c))
- {
- /* Drop a backslash newline, and continue. */
- if (namebuf[-1] == '\\')
- {
- handle_newline (cur, buffer->rlimit, c);
- namebuf--;
- continue;
- }
-
- cur--;
-
- /* In Fortran and assembly language, silently terminate
- strings of either variety at end of line. This is a
- kludge around not knowing where comments are in these
- languages. */
- if (CPP_OPTION (pfile, lang_fortran)
- || CPP_OPTION (pfile, lang_asm))
- goto out;
-
- /* Character constants, headers and asserts may not
- extend over multiple lines. In Standard C, neither
- may strings. We accept multiline strings as an
- extension, but not in directives. */
- if (terminator != '"' || IS_DIRECTIVE (list))
- goto unterminated;
-
- cur++; /* Move forwards again. */
-
- if (pfile->multiline_string_line == 0)
- {
- pfile->multiline_string_line = list->line;
- if (CPP_PEDANTIC (pfile))
- cpp_pedwarn (pfile, "multi-line string constant");
- }
-
- *namebuf++ = '\n';
- handle_newline (cur, buffer->rlimit, c);
- }
- else
- {
- unsigned char *temp;
-
- /* An odd number of consecutive backslashes represents
- an escaped terminator. */
- temp = namebuf - 1;
- while (temp >= name_start && *temp == '\\')
- temp--;
-
- if ((namebuf - temp) & 1)
- goto out;
- namebuf++;
- }
- }
- }
-
- /* Run out of name space? */
- if (cur < buffer->rlimit)
- {
- list->name_used = namebuf - list->namebuf;
- auto_expand_name_space (list);
- goto expanded;
- }
-
- /* We may not have trigraph-replaced the input for this code path,
- but as the input is in error by being unterminated we don't
- bother. Prevent warnings about no newlines at EOF. */
- if (IS_NEWLINE(cur[-1]))
- cur--;
-
- unterminated:
- cpp_error (pfile, "missing terminating %c character", (int) terminator);
-
- if (terminator == '\"' && pfile->multiline_string_line != list->line
- && pfile->multiline_string_line != 0)
- {
- cpp_error_with_line (pfile, pfile->multiline_string_line, -1,
- "possible start of unterminated string literal");
- pfile->multiline_string_line = 0;
- }
-
- out:
- buffer->cur = cur;
- name->len = namebuf - (list->namebuf + name->offset);
- list->name_used = namebuf - list->namebuf;
-
- if (null_count > 0)
- cpp_warning (pfile, (null_count > 1 ? "null characters preserved"
- : "null character preserved"));
-}
-
-/* The character C helps us distinguish comment types: '*' = C style,
- '-' = Chill-style and '/' = C++ style. For code simplicity, the
- stored comment includes any C-style comment terminator. */
-static void
-copy_comment (list, from, len, tok_no, type)
- cpp_toklist *list;
- const unsigned char *from;
- unsigned int len;
- unsigned int tok_no;
- unsigned int type;
-{
- cpp_token *comment;
-
- if (list->comments_used == list->comments_cap)
- expand_comment_space (list);
-
- if (list->name_used + len > list->name_cap)
- expand_name_space (list, len);
-
- comment = &list->comments[list->comments_used++];
- comment->type = type;
- comment->aux = tok_no;
- comment->val.name.len = len;
- comment->val.name.offset = list->name_used;
-
- memcpy (list->namebuf + list->name_used, from, len);
- list->name_used += len;
-}
-
-/*
- * The tokenizer's main loop. Returns a token list, representing a
- * logical line in the input file, terminated with a CPP_VSPACE
- * token. On EOF, a token list containing the single CPP_EOF token
- * is returned.
- *
- * Implementation relies almost entirely on lookback, rather than
- * looking forwards. This means that tokenization requires just
- * a single pass of the file, even in the presence of trigraphs and
- * escaped newlines, providing significant performance benefits.
- * Trigraph overhead is negligible if they are disabled, and low
- * even when enabled.
- */
-
-#define PUSH_TOKEN(ttype) cur_token++->type = ttype
-#define REVISE_TOKEN(ttype) cur_token[-1].type = ttype
-#define BACKUP_TOKEN(ttype) (--cur_token)->type = ttype
-#define BACKUP_DIGRAPH(ttype) do { \
- BACKUP_TOKEN(ttype); cur_token->flags |= DIGRAPH;} while (0)
-
-void
-_cpp_lex_line (pfile, list)
- cpp_reader *pfile;
- cpp_toklist *list;
-{
- cpp_token *cur_token, *token_limit;
- cpp_buffer *buffer = pfile->buffer;
- register const unsigned char *cur = buffer->cur;
- unsigned char flags = 0;
-
- expanded:
- token_limit = list->tokens + list->tokens_cap;
- cur_token = list->tokens + list->tokens_used;
-
- for (; cur < buffer->rlimit && cur_token < token_limit;)
- {
- unsigned char c = *cur++;
-
- /* Optimize whitespace skipping, in particular the case of a
- single whitespace character, as every other token is probably
- whitespace. (' ' '\t' '\v' '\f' '\0'). */
- if (is_hspace ((unsigned int) c))
- {
- if (c == '\0' || (cur < buffer->rlimit && is_hspace (*cur)))
- {
- buffer->cur = cur - (c == '\0'); /* Get the null warning. */
- skip_whitespace (pfile, IS_DIRECTIVE (list));
- cur = buffer->cur;
- }
- flags = PREV_WHITESPACE;
- if (cur == buffer->rlimit)
- break;
- c = *cur++;
- }
-
- /* Initialize current token. Its type is set in the switch. */
- cur_token->col = COLUMN (cur);
- cur_token->flags = flags;
- flags = 0;
-
- switch (c)
- {
- case '0': case '1': case '2': case '3': case '4':
- case '5': case '6': case '7': case '8': case '9':
- /* Prepend an immediately previous CPP_DOT token. */
- if (PREV_TOKEN_TYPE == CPP_DOT && IMMED_TOKEN ())
- {
- cur_token--;
- if (list->name_cap == list->name_used)
- auto_expand_name_space (list);
-
- cur_token->val.name.len = 1;
- cur_token->val.name.offset = list->name_used;
- list->namebuf[list->name_used++] = '.';
- }
- else
- INIT_NAME (list, cur_token->val.name);
- cur--; /* Backup character. */
-
- continue_number:
- buffer->cur = cur;
- parse_number (pfile, list, &cur_token->val.name);
- cur = buffer->cur;
-
- PUSH_TOKEN (CPP_NUMBER); /* Number not yet interpreted. */
- break;
-
- letter:
- case '_':
- 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':
- INIT_NAME (list, cur_token->val.name);
- cur--; /* Backup character. */
- cur_token->type = CPP_NAME; /* Identifier, macro etc. */
-
- continue_name:
- buffer->cur = cur;
- parse_name (pfile, list, &cur_token->val.name);
- cur = buffer->cur;
-
- /* Find handler for newly created / extended directive. */
- if (IS_DIRECTIVE (list) && cur_token == &NTH_TOKEN (list, 2))
- _cpp_check_directive (list, cur_token);
- cur_token++;
- break;
-
- case '\'':
- /* Fall through. */
- case '\"':
- cur_token->type = c == '\'' ? CPP_CHAR : CPP_STRING;
- /* Do we have a wide string? */
- if (cur_token[-1].type == CPP_NAME && IMMED_TOKEN ()
- && cur_token[-1].val.name.len == 1
- && TOK_NAME (list, cur_token - 1)[0] == 'L'
- && !CPP_TRADITIONAL (pfile))
- {
- /* No need for 'L' any more. */
- list->name_used--;
- (--cur_token)->type = (c == '\'' ? CPP_WCHAR : CPP_WSTRING);
- }
-
- do_parse_string:
- /* Here c is one of ' " > or ). */
- INIT_NAME (list, cur_token->val.name);
- buffer->cur = cur;
- parse_string (pfile, list, &cur_token->val.name, c);
- cur = buffer->cur;
- cur_token++;
- break;
-
- case '/':
- cur_token->type = CPP_DIV;
- if (IMMED_TOKEN ())
- {
- if (PREV_TOKEN_TYPE == CPP_DIV)
- {
- /* We silently allow C++ comments in system headers,
- irrespective of conformance mode, because lots of
- broken systems do that and trying to clean it up
- in fixincludes is a nightmare. */
- if (buffer->system_header_p)
- goto do_line_comment;
- else if (CPP_OPTION (pfile, cplusplus_comments))
- {
- if (CPP_OPTION (pfile, c89) && CPP_PEDANTIC (pfile)
- && ! buffer->warned_cplusplus_comments)
- {
- buffer->cur = cur;
- cpp_pedwarn (pfile,
- "C++ style comments are not allowed in ISO C89");
- cpp_pedwarn (pfile,
- "(this will be reported only once per input file)");
- buffer->warned_cplusplus_comments = 1;
- }
- do_line_comment:
- buffer->cur = cur;
- if (cur[-2] != c)
- cpp_warning (pfile,
- "comment start split across lines");
- if (skip_line_comment (pfile))
- cpp_error_with_line (pfile, list->line,
- cur_token[-1].col,
- "multi-line comment");
- if (!CPP_OPTION (pfile, discard_comments))
- copy_comment (list, cur, buffer->cur - cur,
- cur_token - 1 - list->tokens, c == '/'
- ? CPP_CPP_COMMENT: CPP_CHILL_COMMENT);
- cur = buffer->cur;
-
- /* Back-up to first '-' or '/'. */
- cur_token -= 2;
- if (!CPP_OPTION (pfile, traditional))
- flags = PREV_WHITESPACE;
- }
- }
- }
- cur_token++;
- break;
-
- case '*':
- cur_token->type = CPP_MULT;
- if (IMMED_TOKEN ())
- {
- if (PREV_TOKEN_TYPE == CPP_DIV)
- {
- buffer->cur = cur;
- if (cur[-2] != '/')
- cpp_warning (pfile,
- "comment start '/*' split across lines");
- if (skip_block_comment (pfile))
- cpp_error_with_line (pfile, list->line, cur_token[-1].col,
- "unterminated comment");
- else if (buffer->cur[-2] != '*')
- cpp_warning (pfile,
- "comment end '*/' split across lines");
- if (!CPP_OPTION (pfile, discard_comments))
- copy_comment (list, cur, buffer->cur - cur,
- cur_token - 1 - list->tokens, CPP_C_COMMENT);
- cur = buffer->cur;
-
- cur_token -= 2;
- if (!CPP_OPTION (pfile, traditional))
- flags = PREV_WHITESPACE;
- }
- else if (CPP_OPTION (pfile, cplusplus))
- {
- /* In C++, there are .* and ->* operators. */
- if (PREV_TOKEN_TYPE == CPP_DEREF)
- BACKUP_TOKEN (CPP_DEREF_STAR);
- else if (PREV_TOKEN_TYPE == CPP_DOT)
- BACKUP_TOKEN (CPP_DOT_STAR);
- }
- }
- cur_token++;
- break;
-
- case '\n':
- case '\r':
- handle_newline (cur, buffer->rlimit, c);
- if (PREV_TOKEN_TYPE != CPP_BACKSLASH || !IMMED_TOKEN ())
- {
- if (PREV_TOKEN_TYPE == CPP_BACKSLASH)
- {
- buffer->cur = cur;
- cpp_warning (pfile,
- "backslash and newline separated by space");
- }
- PUSH_TOKEN (CPP_VSPACE);
- goto out;
- }
- /* Remove the escaped newline. Then continue to process
- any interrupted name or number. */
- cur_token--;
- if (IMMED_TOKEN ())
- {
- cur_token--;
- if (cur_token->type == CPP_NAME)
- goto continue_name;
- else if (cur_token->type == CPP_NUMBER)
- goto continue_number;
- cur_token++;
- }
- break;
-
- case '-':
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_MINUS)
- {
- if (CPP_OPTION (pfile, chill))
- goto do_line_comment;
- REVISE_TOKEN (CPP_MINUS_MINUS);
- }
- else
- PUSH_TOKEN (CPP_MINUS);
- break;
-
- /* The digraph flag checking ensures that ## and %:%:
- are interpreted as CPP_PASTE, but #%: and %:# are not. */
- make_hash:
- case '#':
- if (PREV_TOKEN_TYPE == CPP_HASH && IMMED_TOKEN ()
- && ((cur_token->flags ^ cur_token[-1].flags) & DIGRAPH) == 0)
- REVISE_TOKEN (CPP_PASTE);
- else
- PUSH_TOKEN (CPP_HASH);
- break;
-
- case ':':
- cur_token->type = CPP_COLON;
- if (IMMED_TOKEN ())
- {
- if (PREV_TOKEN_TYPE == CPP_COLON
- && CPP_OPTION (pfile, cplusplus))
- BACKUP_TOKEN (CPP_SCOPE);
- /* Digraph: "<:" is a '[' */
- else if (PREV_TOKEN_TYPE == CPP_LESS)
- BACKUP_DIGRAPH (CPP_OPEN_SQUARE);
- /* Digraph: "%:" is a '#' */
- else if (PREV_TOKEN_TYPE == CPP_MOD)
- {
- (--cur_token)->flags |= DIGRAPH;
- goto make_hash;
- }
- }
- cur_token++;
- break;
-
- case '&':
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_AND)
- REVISE_TOKEN (CPP_AND_AND);
- else
- PUSH_TOKEN (CPP_AND);
- break;
-
- make_or:
- case '|':
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_OR)
- REVISE_TOKEN (CPP_OR_OR);
- else
- PUSH_TOKEN (CPP_OR);
- break;
-
- case '+':
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_PLUS)
- REVISE_TOKEN (CPP_PLUS_PLUS);
- else
- PUSH_TOKEN (CPP_PLUS);
- break;
-
- case '=':
- /* This relies on equidistance of "?=" and "?" tokens. */
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE <= CPP_LAST_EQ)
- REVISE_TOKEN (PREV_TOKEN_TYPE + (CPP_EQ_EQ - CPP_EQ));
- else
- PUSH_TOKEN (CPP_EQ);
- break;
-
- case '>':
- cur_token->type = CPP_GREATER;
- if (IMMED_TOKEN ())
- {
- if (PREV_TOKEN_TYPE == CPP_GREATER)
- BACKUP_TOKEN (CPP_RSHIFT);
- else if (PREV_TOKEN_TYPE == CPP_MINUS)
- BACKUP_TOKEN (CPP_DEREF);
- /* Digraph: ":>" is a ']' */
- else if (PREV_TOKEN_TYPE == CPP_COLON)
- BACKUP_DIGRAPH (CPP_CLOSE_SQUARE);
- /* Digraph: "%>" is a '}' */
- else if (PREV_TOKEN_TYPE == CPP_MOD)
- BACKUP_DIGRAPH (CPP_CLOSE_BRACE);
- }
- cur_token++;
- break;
-
- case '<':
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_LESS)
- {
- REVISE_TOKEN (CPP_LSHIFT);
- break;
- }
- /* Is this the beginning of a header name? */
- if (list->dir_flags & SYNTAX_INCLUDE)
- {
- c = '>'; /* Terminator. */
- cur_token->type = CPP_HEADER_NAME;
- goto do_parse_string;
- }
- PUSH_TOKEN (CPP_LESS);
- break;
-
- case '%':
- /* Digraph: "<%" is a '{' */
- cur_token->type = CPP_MOD;
- if (IMMED_TOKEN () && PREV_TOKEN_TYPE == CPP_LESS)
- BACKUP_DIGRAPH (CPP_OPEN_BRACE);
- cur_token++;
- break;
-
- case ')':
- PUSH_TOKEN (CPP_CLOSE_PAREN);
- break;
-
- case '(':
- /* Is this the beginning of an assertion string? */
- if (list->dir_flags & SYNTAX_ASSERT)
- {
- c = ')'; /* Terminator. */
- cur_token->type = CPP_ASSERTION;
- goto do_parse_string;
- }
- PUSH_TOKEN (CPP_OPEN_PAREN);
- break;
-
- make_complement:
- case '~':
- PUSH_TOKEN (CPP_COMPL);
- break;
-
- case '?':
- if (cur + 1 < buffer->rlimit && *cur == '?'
- && trigraph_map[cur[1]] && trigraph_ok (pfile, cur + 1))
- {
- /* Handle trigraph. */
- cur++;
- switch (*cur++)
- {
- case '(': goto make_open_square;
- case ')': goto make_close_square;
- case '<': goto make_open_brace;
- case '>': goto make_close_brace;
- case '=': goto make_hash;
- case '!': goto make_or;
- case '-': goto make_complement;
- case '/': goto make_backslash;
- case '\'': goto make_xor;
- }
- }
- if (IMMED_TOKEN () && CPP_OPTION (pfile, cplusplus))
- {
- /* GNU C++ defines <? and >? operators. */
- if (PREV_TOKEN_TYPE == CPP_LESS)
- {
- REVISE_TOKEN (CPP_MIN);
- break;
- }
- else if (PREV_TOKEN_TYPE == CPP_GREATER)
- {
- REVISE_TOKEN (CPP_MAX);
- break;
- }
- }
- PUSH_TOKEN (CPP_QUERY);
- break;
-
- case '.':
- if (PREV_TOKEN_TYPE == CPP_DOT && cur_token[-2].type == CPP_DOT
- && IMMED_TOKEN ()
- && !(cur_token[-1].flags & PREV_WHITESPACE))
- {
- cur_token -= 2;
- PUSH_TOKEN (CPP_ELLIPSIS);
- }
- else
- PUSH_TOKEN (CPP_DOT);
- break;
-
- make_xor:
- case '^': PUSH_TOKEN (CPP_XOR); break;
- make_open_brace:
- case '{': PUSH_TOKEN (CPP_OPEN_BRACE); break;
- make_close_brace:
- case '}': PUSH_TOKEN (CPP_CLOSE_BRACE); break;
- make_open_square:
- case '[': PUSH_TOKEN (CPP_OPEN_SQUARE); break;
- make_close_square:
- case ']': PUSH_TOKEN (CPP_CLOSE_SQUARE); break;
- make_backslash:
- case '\\': PUSH_TOKEN (CPP_BACKSLASH); break;
- case '!': PUSH_TOKEN (CPP_NOT); break;
- case ',': PUSH_TOKEN (CPP_COMMA); break;
- case ';': PUSH_TOKEN (CPP_SEMICOLON); break;
-
- case '$':
- if (CPP_OPTION (pfile, dollars_in_ident))
- goto letter;
- /* Fall through */
- default:
- cur_token->aux = c;
- PUSH_TOKEN (CPP_OTHER);
- break;
- }
- }
-
- /* Run out of token space? */
- if (cur_token == token_limit)
- {
- list->tokens_used = cur_token - list->tokens;
- expand_token_space (list);
- goto expanded;
- }
-
- cur_token->type = CPP_EOF;
- cur_token->flags = flags;
-
- if (cur_token != &NTH_TOKEN (list, 1))
- {
- /* Next call back will get just a CPP_EOF. */
- buffer->cur = cur;
- cpp_warning (pfile, "no newline at end of file");
- PUSH_TOKEN (CPP_VSPACE);
- }
-
- out:
- buffer->cur = cur;
-
- list->tokens_used = cur_token - list->tokens;
-
- /* FIXME: take this check out and put it in the caller.
- list->directive == 0 indicates an unknown directive (but null
- directive is OK). This is the first time we can be sure the
- directive is invalid, and thus warn about it, because it might
- have been split by escaped newlines. Also, don't complain about
- invalid directives in assembly source, we don't know where the
- comments are, and # may introduce assembler pseudo-ops. */
-
- if (IS_DIRECTIVE (list) && list->dir_handler == 0
- && NTH_TOKEN (list, 2).type != CPP_VSPACE
- && !CPP_OPTION (pfile, lang_asm))
- cpp_error_with_line (pfile, list->line, NTH_TOKEN(list, 2).col,
- "invalid preprocessing directive");
-}
-
-/* Token spelling functions. Used for output of a preprocessed file,
- stringizing and token pasting. They all assume sufficient buffer
- is allocated, and return exactly how much they used. */
-
-/* Needs buffer of 3 + len. */
-unsigned int
-spell_char (buffer, list, token)
- unsigned char *buffer;
- cpp_toklist *list;
- cpp_token *token;
-{
- unsigned char* orig_buff = buffer;
- size_t len;
-
- if (token->type == CPP_WCHAR)
- *buffer++ = 'L';
- *buffer++ = '\'';
-
- len = token->val.name.len;
- memcpy (buffer, TOK_NAME (list, token), len);
- buffer += len;
- *buffer++ = '\'';
- return buffer - orig_buff;
-}
-
-/* Needs buffer of 3 + len. */
-unsigned int
-spell_string (buffer, list, token)
- unsigned char *buffer;
- cpp_toklist *list;
- cpp_token *token;
-{
- unsigned char* orig_buff = buffer;
- size_t len;
-
- if (token->type == CPP_WSTRING)
- *buffer++ = 'L';
- *buffer++ = '"';
-
- len = token->val.name.len;
- memcpy (buffer, TOK_NAME (list, token), len);
- buffer += len;
- *buffer++ = '"';
- return buffer - orig_buff;
-}
-
-/* Needs buffer of len + 2. */
-unsigned int
-spell_comment (buffer, list, token)
- unsigned char *buffer;
- cpp_toklist *list;
- cpp_token *token;
-{
- size_t len;
-
- if (token->type == CPP_C_COMMENT)
- {
- *buffer++ = '/';
- *buffer++ = '*';
- }
- else if (token->type == CPP_CPP_COMMENT)
- {
- *buffer++ = '/';
- *buffer++ = '/';
- }
- else
- {
- *buffer++ = '-';
- *buffer++ = '-';
- }
-
- len = token->val.name.len;
- memcpy (buffer, TOK_NAME (list, token), len);
-
- return len + 2;
-}
-
-/* Needs buffer of len. */
-unsigned int
-spell_name (buffer, list, token)
- unsigned char *buffer;
- cpp_toklist *list;
- cpp_token *token;
-{
- size_t len;
-
- len = token->val.name.len;
- memcpy (buffer, TOK_NAME (list, token), len);
- buffer += len;
-
- return len;
-}
-
-/* Needs buffer of 1. */
-unsigned int
-spell_other (buffer, list, token)
- unsigned char *buffer;
- cpp_toklist *list ATTRIBUTE_UNUSED;
- cpp_token *token;
-{
- *buffer++ = token->aux;
- return 1;
-}
-
-void
-_cpp_lex_file (pfile)
- cpp_reader* pfile;
-{
- int recycle;
- cpp_toklist* list;
-
- init_trigraph_map ();
- list = (cpp_toklist *) xmalloc (sizeof (cpp_toklist));
-
- for (recycle = 0; ;)
- {
- init_token_list (pfile, list, recycle);
- recycle = 1;
-
- _cpp_lex_line (pfile, list);
- if (NTH_TOKEN (list, 1).type == CPP_EOF)
- break;
-
- if (list->dir_handler)
- {
- if (list->dir_handler (pfile))
- {
- list = (cpp_toklist *) xmalloc (sizeof (cpp_toklist));
- recycle = 0;
- }
- }
- else
- _cpp_output_list (pfile, list);
- }
-}
-
-static void
-_cpp_output_list (pfile, list)
- cpp_reader *pfile;
- cpp_toklist *list;
-{
- unsigned int comment_no = 0;
- cpp_token *token, *comment_token = 0;
-
- if (list->comments_used > 0)
- comment_token = list->tokens + list->comments[0].aux;
-
- CPP_RESERVE (pfile, 2); /* Always have room for " \n". */
- for (token = &NTH_TOKEN(list, 1);; token++)
- {
- if (token->flags & PREV_WHITESPACE)
- {
- /* Output comments if -C. Otherwise a space will do. */
- if (token == comment_token)
- {
- cpp_token *comment = &list->comments[comment_no];
- do
- {
- /* Longest wrapper is 4. */
- CPP_RESERVE (pfile, 4 + 2 + comment->val.name.len);
- pfile->limit += spell_comment (pfile->limit, list, comment);
- comment_no++, comment++;
- if (comment_no == list->comments_used)
- break;
- comment_token = comment->aux + list->tokens;
- }
- while (comment_token == token);
- }
- else
- CPP_PUTC_Q (pfile, ' ');
- }
-
- switch (token_spellings[token->type].type)
- {
- case SPELL_TEXT:
- {
- const unsigned char *spelling;
- unsigned char c;
-
- CPP_RESERVE (pfile, 4 + 2); /* Longest is 4. */
- if (token->flags & DIGRAPH)
- spelling = digraph_spellings [token->type - CPP_FIRST_DIGRAPH];
- else
- spelling = token_spellings[token->type].speller;
-
- while ((c = *spelling++) != '\0')
- CPP_PUTC_Q (pfile, c);
- }
- break;
-
- case SPELL_HANDLER:
- {
- speller s;
-
- s = (speller) token_spellings[token->type].speller;
- /* Longest wrapper is 4. */
- CPP_RESERVE (pfile, 4 + 2 + token->val.name.len);
- pfile->limit += s (pfile->limit, list, token);
- }
- break;
-
- case SPELL_EOL:
- CPP_PUTC_Q (pfile, '\n');
- return;
-
- case SPELL_NONE:
- cpp_error (pfile, "Unwriteable token");
- break;
- }
- }
-}
-
-#endif
diff --git a/gcc/cpplib.c b/gcc/cpplib.c
index 2d466ff..5bb5162 100644
--- a/gcc/cpplib.c
+++ b/gcc/cpplib.c
@@ -150,29 +150,6 @@ DIRECTIVE_TABLE
#undef D
#undef DIRECTIVE_TABLE
-/* Check if a token's name matches that of a known directive. Put in
- this file to save exporting dtable and other unneeded information. */
-void
-_cpp_check_directive (list, token)
- cpp_toklist *list;
- cpp_token *token;
-{
- const char *name = list->namebuf + token->val.name.offset;
- size_t len = token->val.name.len;
- unsigned int i;
-
- list->dir_handler = 0;
- list->dir_flags = 0;
-
- for (i = 0; i < N_DIRECTIVES; i++)
- if (dtable[i].length == len && !strncmp (dtable[i].name, name, len))
- {
- list->dir_handler = dtable[i].func;
- list->dir_flags = dtable[i].flags;
- break;
- }
-}
-
/* Handle a possible # directive.
'#' has already been read. */
diff --git a/gcc/cpplib.h b/gcc/cpplib.h
index b3d8c09..764d8e6 100644
--- a/gcc/cpplib.h
+++ b/gcc/cpplib.h
@@ -34,26 +34,13 @@ typedef struct cpp_options cpp_options;
typedef struct cpp_printer cpp_printer;
typedef struct cpp_token cpp_token;
typedef struct cpp_toklist cpp_toklist;
-typedef struct cpp_name cpp_name;
-
-/* The first two groups, apart from '=', can appear in preprocessor
- expressions. This allows a lookup table to be implemented in
- _cpp_parse_expr.
-
- The first group, to CPP_LAST_EQ, can be immediately followed by an
- '='. The lexer needs operators ending in '=', like ">>=", to be in
- the same order as their counterparts without the '=', like ">>". */
-
-/* Positions in the table. */
-#define CPP_LAST_EQ CPP_LSHIFT
-#define CPP_FIRST_DIGRAPH CPP_HASH
+ /* Put operators that can appear in a preprocessor expression first.
+ This allows a lookup table to be implemented in _cpp_parse_expr.
+ Ordering within this group is currently not significant, apart
+ from those ending in '=' being at the end. */
#define TTYPE_TABLE \
- T(CPP_EQ = 0, "=") \
- T(CPP_NOT, "!") \
- T(CPP_GREATER, ">") /* compare */ \
- T(CPP_LESS, "<") \
- T(CPP_PLUS, "+") /* math */ \
+ T(CPP_PLUS = 0, "+") /* math */ \
T(CPP_MINUS, "-") \
T(CPP_MULT, "*") \
T(CPP_DIV, "/") \
@@ -64,19 +51,22 @@ typedef struct cpp_name cpp_name;
T(CPP_COMPL, "~") \
T(CPP_RSHIFT, ">>") \
T(CPP_LSHIFT, "<<") \
-\
- T(CPP_AND_AND, "&&") /* logical */ \
+ T(CPP_NOT, "!") /* logicals */ \
+ T(CPP_AND_AND, "&&") \
T(CPP_OR_OR, "||") \
T(CPP_QUERY, "?") \
T(CPP_COLON, ":") \
T(CPP_COMMA, ",") /* grouping */ \
T(CPP_OPEN_PAREN, "(") \
T(CPP_CLOSE_PAREN, ")") \
- T(CPP_EQ_EQ, "==") /* compare */ \
+ T(CPP_GREATER, ">") /* compare */ \
+ T(CPP_LESS, "<") \
+ T(CPP_EQ_EQ, "==") \
T(CPP_NOT_EQ, "!=") \
T(CPP_GREATER_EQ, ">=") \
T(CPP_LESS_EQ, "<=") \
\
+ /* The remainder of the punctuation. Order is not significant. */ \
T(CPP_PLUS_EQ, "+=") /* math */ \
T(CPP_MINUS_EQ, "-=") \
T(CPP_MULT_EQ, "*=") \
@@ -88,67 +78,55 @@ typedef struct cpp_name cpp_name;
T(CPP_COMPL_EQ, "~=") \
T(CPP_RSHIFT_EQ, ">>=") \
T(CPP_LSHIFT_EQ, "<<=") \
- /* Digraphs together, beginning with CPP_FIRST_DIGRAPH. */ \
- T(CPP_HASH, "#") /* digraphs */ \
- T(CPP_PASTE, "##") \
- T(CPP_OPEN_SQUARE, "[") \
- T(CPP_CLOSE_SQUARE, "]") \
- T(CPP_OPEN_BRACE, "{") \
- T(CPP_CLOSE_BRACE, "}") \
- /* The remainder of the punctuation. Order is not significant. */ \
- T(CPP_SEMICOLON, ";") /* structure */ \
- T(CPP_ELLIPSIS, "...") \
- T(CPP_BACKSLASH, "\\") \
+ T(CPP_EQ, "=") /* assign */ \
T(CPP_PLUS_PLUS, "++") /* increment */ \
T(CPP_MINUS_MINUS, "--") \
T(CPP_DEREF, "->") /* accessors */ \
T(CPP_DOT, ".") \
+ T(CPP_OPEN_SQUARE, "[") \
+ T(CPP_CLOSE_SQUARE, "]") \
T(CPP_SCOPE, "::") \
T(CPP_DEREF_STAR, "->*") \
T(CPP_DOT_STAR, ".*") \
+ T(CPP_OPEN_BRACE, "{") /* structure */ \
+ T(CPP_CLOSE_BRACE, "}") \
+ T(CPP_SEMICOLON, ";") \
+ T(CPP_ELLIPSIS, "...") \
+ T(CPP_HASH, "#") \
+ T(CPP_PASTE, "##") \
+ T(CPP_BACKSLASH, "\\") \
T(CPP_MIN, "<?") /* extension */ \
T(CPP_MAX, ">?") \
- H(CPP_OTHER, spell_other) /* stray punctuation */ \
+ T(CPP_OTHER, spell_other) /* stray punctuation */ \
\
- H(CPP_NAME, spell_name) /* word */ \
- N(CPP_INT, 0) /* 23 */ \
- N(CPP_FLOAT, 0) /* 3.14159 */ \
- H(CPP_NUMBER, spell_name) /* 34_be+ta */ \
- H(CPP_CHAR, spell_char) /* 'char' */ \
- H(CPP_WCHAR, spell_char) /* L'char' */ \
- H(CPP_STRING, spell_string) /* "string" */ \
- H(CPP_WSTRING, spell_string) /* L"string" */ \
+ T(CPP_NAME, spell_name) /* word */ \
+ T(CPP_INT, 0) /* 23 */ \
+ T(CPP_FLOAT, 0) /* 3.14159 */ \
+ T(CPP_NUMBER, spell_name) /* 34_be+ta */ \
+ T(CPP_CHAR, spell_char) /* 'char' */ \
+ T(CPP_WCHAR, spell_char) /* L'char' */ \
+ T(CPP_STRING, spell_string) /* "string" */ \
+ T(CPP_WSTRING, spell_string) /* L"string" */ \
\
- H(CPP_C_COMMENT, spell_comment) /* Only if output comments. */ \
- H(CPP_CPP_COMMENT, spell_comment) /* Only if output comments. */ \
- H(CPP_CHILL_COMMENT, spell_comment) /* Only if output comments. */ \
- N(CPP_MACRO_ARG, 0) /* Macro argument. */ \
- N(CPP_SUBLIST, 0) /* Sublist. */ \
- E(CPP_VSPACE, "\n") /* End of line. */ \
- N(CPP_EOF, 0) /* End of file. */ \
- N(CPP_HEADER_NAME, 0) /* <stdio.h> in #include */ \
- N(CPP_ASSERTION, 0) /* (...) in #assert */ \
+ T(CPP_COMMENT, spell_comment) /* Only if output comments. */ \
+ T(CPP_VSPACE, "\n") /* End of line. */ \
+ T(CPP_EOF, 0) /* End of file. */ \
+ T(CPP_HEADER_NAME, 0) /* <stdio.h> in #include */ \
+ T(CPP_ASSERTION, 0) /* (...) in #assert */ \
\
/* Obsolete - will be removed when no code uses them still. */ \
- H(CPP_COMMENT, 0) /* Only if output comments. */ \
- N(CPP_HSPACE, 0) /* Horizontal white space. */ \
- N(CPP_POP, 0) /* End of buffer. */ \
- N(CPP_DIRECTIVE, 0) /* #define and the like */ \
- N(CPP_MACRO, 0) /* Like a NAME, but expanded. */
+ T(CPP_HSPACE, 0) /* Horizontal white space. */ \
+ T(CPP_POP, 0) /* End of buffer. */ \
+ T(CPP_DIRECTIVE, 0) /* #define and the like */ \
+ T(CPP_MACRO, 0) /* Like a NAME, but expanded. */
#define T(e, s) e,
-#define H(e, s) e,
-#define N(e, s) e,
-#define E(e, s) e,
enum cpp_ttype
{
TTYPE_TABLE
N_TTYPES
};
#undef T
-#undef H
-#undef N
-#undef E
/* Payload of a NAME, NUMBER, FLOAT, STRING, or COMMENT token. */
struct cpp_name
@@ -157,12 +135,8 @@ struct cpp_name
unsigned int offset; /* from list->namebuf */
};
-#define TOK_NAME(list, token) ((list)->namebuf + (token)->val.name.offset)
-
-/* Flags for the cpp_token structure. */
-#define PREV_WHITESPACE 1 /* If whitespace before this token. */
-#define DIGRAPH 2 /* If it was a digraph. */
-#define UNSIGNED_INT 4 /* If int preprocessing token unsigned. */
+/* Per token flags. */
+#define HSPACE_BEFORE (1 << 0) /* token preceded by hspace */
/* A preprocessing token.
This has been carefully packed and should occupy 16 bytes on
@@ -176,9 +150,8 @@ struct cpp_token
unsigned char type;
#endif
unsigned char flags; /* flags - see above */
- unsigned int aux; /* CPP_OTHER character. Hash of a
- NAME, or something - see uses
- in the code */
+ unsigned int aux; /* hash of a NAME, or something -
+ see uses in the code */
union
{
struct cpp_name name; /* a string */
@@ -195,7 +168,7 @@ typedef int (*parse_cleanup_t) PARAMS ((cpp_buffer *, cpp_reader *));
struct cpp_toklist
{
- cpp_token *tokens; /* actual tokens as an array */
+ struct cpp_token *tokens; /* actual tokens as an array */
unsigned int tokens_used; /* tokens used */
unsigned int tokens_cap; /* tokens allocated */
@@ -205,11 +178,6 @@ struct cpp_toklist
unsigned int line; /* starting line number */
- /* Comment copying. */
- cpp_token *comments; /* comment tokens. */
- unsigned int comments_used; /* comment tokens used. */
- unsigned int comments_cap; /* comment token capacity. */
-
/* Only used if tokens[0].type == CPP_DIRECTIVE. This is the
handler to call after lexing the rest of this line. The flags
indicate whether the rest of the line gets special treatment
@@ -593,7 +561,6 @@ struct cpp_printer
/* Name under which this program was invoked. */
extern const char *progname;
-extern void _cpp_lex_file PARAMS((cpp_reader *));
extern int cpp_handle_options PARAMS ((cpp_reader *, int, char **));
extern enum cpp_ttype cpp_get_token PARAMS ((cpp_reader *));
extern enum cpp_ttype cpp_get_non_space_token PARAMS ((cpp_reader *));
@@ -613,8 +580,6 @@ extern void cpp_assert PARAMS ((cpp_reader *, const char *));
extern void cpp_undef PARAMS ((cpp_reader *, const char *));
extern void cpp_unassert PARAMS ((cpp_reader *, const char *));
-extern void cpp_free_token_list PARAMS ((cpp_toklist *));
-
/* N.B. The error-message-printer prototypes have not been nicely
formatted because exgettext needs to see 'msgid' on the same line
as the name of the function in order to work properly. Only the