1 /* Mainly the interface between cpplib and the C front ends.
2 Copyright (C) 1987, 1988, 1989, 1992, 1994, 1995, 1996, 1997
3 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2007, 2008, 2009, 2010,
4 2011 Free Software Foundation, Inc.
6 This file is part of GCC.
8 GCC is free software; you can redistribute it and/or modify it under
9 the terms of the GNU General Public License as published by the Free
10 Software Foundation; either version 3, or (at your option) any later
13 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
14 WARRANTY; without even the implied warranty of MERCHANTABILITY or
15 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
18 You should have received a copy of the GNU General Public License
19 along with GCC; see the file COPYING3. If not see
20 <http://www.gnu.org/licenses/>. */
24 #include "coretypes.h"
36 #include "splay-tree.h"
40 /* We may keep statistics about how long which files took to compile. */
41 static int header_time, body_time;
42 static splay_tree file_info_tree;
44 int pending_lang_change; /* If we need to switch languages - C++ only */
45 int c_header_level; /* depth in C headers - C++ only */
47 static tree interpret_integer (const cpp_token *, unsigned int);
48 static tree interpret_float (const cpp_token *, unsigned int, const char *);
49 static tree interpret_fixed (const cpp_token *, unsigned int);
50 static enum integer_type_kind narrowest_unsigned_type
51 (unsigned HOST_WIDE_INT, unsigned HOST_WIDE_INT, unsigned int);
52 static enum integer_type_kind narrowest_signed_type
53 (unsigned HOST_WIDE_INT, unsigned HOST_WIDE_INT, unsigned int);
54 static enum cpp_ttype lex_string (const cpp_token *, tree *, bool, bool);
55 static tree lex_charconst (const cpp_token *);
56 static void update_header_times (const char *);
57 static int dump_one_header (splay_tree_node, void *);
58 static void cb_line_change (cpp_reader *, const cpp_token *, int);
59 static void cb_ident (cpp_reader *, unsigned int, const cpp_string *);
60 static void cb_def_pragma (cpp_reader *, unsigned int);
61 static void cb_define (cpp_reader *, unsigned int, cpp_hashnode *);
62 static void cb_undef (cpp_reader *, unsigned int, cpp_hashnode *);
67 struct cpp_callbacks *cb;
68 struct c_fileinfo *toplevel;
70 /* The get_fileinfo data structure must be initialized before
71 cpp_read_main_file is called. */
72 toplevel = get_fileinfo ("<top level>");
73 if (flag_detailed_statistics)
76 body_time = get_run_time ();
77 toplevel->time = body_time;
80 cb = cpp_get_callbacks (parse_in);
82 cb->line_change = cb_line_change;
84 cb->def_pragma = cb_def_pragma;
85 cb->valid_pch = c_common_valid_pch;
86 cb->read_pch = c_common_read_pch;
88 /* Set the debug callbacks if we can use them. */
89 if ((debug_info_level == DINFO_LEVEL_VERBOSE
90 && (write_symbols == DWARF2_DEBUG
91 || write_symbols == VMS_AND_DWARF2_DEBUG))
92 || flag_dump_go_spec != NULL)
94 cb->define = cb_define;
100 get_fileinfo (const char *name)
103 struct c_fileinfo *fi;
106 file_info_tree = splay_tree_new ((splay_tree_compare_fn) strcmp,
108 (splay_tree_delete_value_fn) free);
110 n = splay_tree_lookup (file_info_tree, (splay_tree_key) name);
112 return (struct c_fileinfo *) n->value;
114 fi = XNEW (struct c_fileinfo);
116 fi->interface_only = 0;
117 fi->interface_unknown = 1;
118 splay_tree_insert (file_info_tree, (splay_tree_key) name,
119 (splay_tree_value) fi);
124 update_header_times (const char *name)
126 /* Changing files again. This means currently collected time
127 is charged against header time, and body time starts back at 0. */
128 if (flag_detailed_statistics)
130 int this_time = get_run_time ();
131 struct c_fileinfo *file = get_fileinfo (name);
132 header_time += this_time - body_time;
133 file->time += this_time - body_time;
134 body_time = this_time;
139 dump_one_header (splay_tree_node n, void * ARG_UNUSED (dummy))
141 print_time ((const char *) n->key,
142 ((struct c_fileinfo *) n->value)->time);
147 dump_time_statistics (void)
149 struct c_fileinfo *file = get_fileinfo (input_filename);
150 int this_time = get_run_time ();
151 file->time += this_time - body_time;
153 fprintf (stderr, "\n******\n");
154 print_time ("header files (total)", header_time);
155 print_time ("main file (total)", this_time - body_time);
156 fprintf (stderr, "ratio = %g : 1\n",
157 (double) header_time / (double) (this_time - body_time));
158 fprintf (stderr, "\n******\n");
160 splay_tree_foreach (file_info_tree, dump_one_header, 0);
164 cb_ident (cpp_reader * ARG_UNUSED (pfile),
165 unsigned int ARG_UNUSED (line),
166 const cpp_string * ARG_UNUSED (str))
168 #ifdef ASM_OUTPUT_IDENT
171 /* Convert escapes in the string. */
172 cpp_string cstr = { 0, 0 };
173 if (cpp_interpret_string (pfile, str, 1, &cstr, CPP_STRING))
175 ASM_OUTPUT_IDENT (asm_out_file, (const char *) cstr.text);
176 free (CONST_CAST (unsigned char *, cstr.text));
182 /* Called at the start of every non-empty line. TOKEN is the first
183 lexed token on the line. Used for diagnostic line numbers. */
185 cb_line_change (cpp_reader * ARG_UNUSED (pfile), const cpp_token *token,
188 if (token->type != CPP_EOF && !parsing_args)
189 input_location = token->src_loc;
193 fe_file_change (const struct line_map *new_map)
198 if (new_map->reason == LC_ENTER)
200 /* Don't stack the main buffer on the input stack;
201 we already did in compile_file. */
202 if (!MAIN_FILE_P (new_map))
204 unsigned int included_at = LAST_SOURCE_LINE_LOCATION (new_map - 1);
206 if (included_at > BUILTINS_LOCATION)
207 line = SOURCE_LINE (new_map - 1, included_at);
209 input_location = new_map->start_location;
210 (*debug_hooks->start_source_file) (line, LINEMAP_FILE (new_map));
211 #ifndef NO_IMPLICIT_EXTERN_C
214 else if (LINEMAP_SYSP (new_map) == 2)
217 ++pending_lang_change;
222 else if (new_map->reason == LC_LEAVE)
224 #ifndef NO_IMPLICIT_EXTERN_C
225 if (c_header_level && --c_header_level == 0)
227 if (LINEMAP_SYSP (new_map) == 2)
228 warning (0, "badly nested C headers from preprocessor");
229 --pending_lang_change;
232 input_location = new_map->start_location;
234 (*debug_hooks->end_source_file) (LINEMAP_LINE (new_map));
237 update_header_times (LINEMAP_FILE (new_map));
238 input_location = new_map->start_location;
242 cb_def_pragma (cpp_reader *pfile, source_location loc)
244 /* Issue a warning message if we have been asked to do so. Ignore
245 unknown pragmas in system headers unless an explicit
246 -Wunknown-pragmas has been given. */
247 if (warn_unknown_pragmas > in_system_header)
249 const unsigned char *space, *name;
251 location_t fe_loc = loc;
253 space = name = (const unsigned char *) "";
254 s = cpp_get_token (pfile);
255 if (s->type != CPP_EOF)
257 space = cpp_token_as_text (pfile, s);
258 s = cpp_get_token (pfile);
259 if (s->type == CPP_NAME)
260 name = cpp_token_as_text (pfile, s);
263 warning_at (fe_loc, OPT_Wunknown_pragmas, "ignoring #pragma %s %s",
268 /* #define callback for DWARF and DWARF2 debug info. */
270 cb_define (cpp_reader *pfile, source_location loc, cpp_hashnode *node)
272 const struct line_map *map = linemap_lookup (line_table, loc);
273 (*debug_hooks->define) (SOURCE_LINE (map, loc),
274 (const char *) cpp_macro_definition (pfile, node));
277 /* #undef callback for DWARF and DWARF2 debug info. */
279 cb_undef (cpp_reader * ARG_UNUSED (pfile), source_location loc,
282 const struct line_map *map = linemap_lookup (line_table, loc);
283 (*debug_hooks->undef) (SOURCE_LINE (map, loc),
284 (const char *) NODE_NAME (node));
287 /* Read a token and return its type. Fill *VALUE with its value, if
288 applicable. Fill *CPP_FLAGS with the token's flags, if it is
292 c_lex_with_flags (tree *value, location_t *loc, unsigned char *cpp_flags,
295 static bool no_more_pch;
296 const cpp_token *tok;
298 unsigned char add_flags = 0;
300 timevar_push (TV_CPP);
302 tok = cpp_get_token_with_location (parse_in, loc);
312 *value = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node.node));
317 const char *suffix = NULL;
318 unsigned int flags = cpp_classify_number (parse_in, tok, &suffix);
320 switch (flags & CPP_N_CATEGORY)
323 /* cpplib has issued an error. */
324 *value = error_mark_node;
328 /* C++ uses '0' to mark virtual functions as pure.
329 Set PURE_ZERO to pass this information to the C++ parser. */
330 if (tok->val.str.len == 1 && *tok->val.str.text == '0')
331 add_flags = PURE_ZERO;
332 *value = interpret_integer (tok, flags);
336 *value = interpret_float (tok, flags, suffix);
343 if (flags & CPP_N_USERDEF)
345 tree suffix_id = get_identifier (suffix);
346 int len = tok->val.str.len - strlen (suffix);
347 /* If this is going to be used as a C string to pass to a
348 raw literal operator, we need to add a trailing NUL. */
349 tree num_string = build_string (len + 1,
350 (const char *) tok->val.str.text);
351 TREE_TYPE (num_string) = char_array_type_node;
352 num_string = fix_string_type (num_string);
353 char *str = CONST_CAST (char *, TREE_STRING_POINTER (num_string));
355 tree literal = build_userdef_literal (suffix_id, *value,
363 /* An @ may give the next token special significance in Objective-C. */
364 if (c_dialect_objc ())
366 location_t atloc = *loc;
370 tok = cpp_get_token_with_location (parse_in, &newloc);
382 type = lex_string (tok, value, true, true);
386 *value = HT_IDENT_TO_GCC_IDENT (HT_NODE (tok->val.node.node));
387 if (OBJC_IS_AT_KEYWORD (C_RID_CODE (*value))
388 || OBJC_IS_CXX_KEYWORD (C_RID_CODE (*value)))
391 /* Note the complication: if we found an OBJC_CXX
392 keyword, for example, 'class', we will be
393 returning a token of type CPP_AT_NAME and rid
394 code RID_CLASS (not RID_AT_CLASS). The language
395 parser needs to convert that to RID_AT_CLASS.
403 error_at (atloc, "stray %<@%> in program");
414 unsigned char name[8];
416 *cpp_spell_token (parse_in, tok, name, true) = 0;
418 error ("stray %qs in program", name);
425 cppchar_t c = tok->val.str.text[0];
427 if (c == '"' || c == '\'')
428 error ("missing terminating %c character", (int) c);
429 else if (ISGRAPH (c))
430 error ("stray %qc in program", (int) c);
432 error ("stray %<\\%o%> in program", (int) c);
436 case CPP_CHAR_USERDEF:
437 case CPP_WCHAR_USERDEF:
438 case CPP_CHAR16_USERDEF:
439 case CPP_CHAR32_USERDEF:
442 cpp_token temp_tok = *tok;
443 const char *suffix = cpp_get_userdef_suffix (tok);
444 temp_tok.val.str.len -= strlen (suffix);
445 temp_tok.type = cpp_userdef_char_remove_type (type);
446 literal = build_userdef_literal (get_identifier (suffix),
447 lex_charconst (&temp_tok), NULL_TREE);
456 *value = lex_charconst (tok);
459 case CPP_STRING_USERDEF:
460 case CPP_WSTRING_USERDEF:
461 case CPP_STRING16_USERDEF:
462 case CPP_STRING32_USERDEF:
463 case CPP_UTF8STRING_USERDEF:
465 tree literal, string;
466 const char *suffix = cpp_get_userdef_suffix (tok);
467 string = build_string (tok->val.str.len - strlen (suffix),
468 (const char *) tok->val.str.text);
469 literal = build_userdef_literal (get_identifier (suffix),
480 if ((lex_flags & C_LEX_STRING_NO_JOIN) == 0)
482 type = lex_string (tok, value, false,
483 (lex_flags & C_LEX_STRING_NO_TRANSLATE) == 0);
486 *value = build_string (tok->val.str.len, (const char *) tok->val.str.text);
490 *value = build_int_cst (integer_type_node, tok->val.pragma);
493 /* These tokens should not be visible outside cpplib. */
494 case CPP_HEADER_NAME:
498 /* CPP_COMMENT will appear when compiling with -C and should be
509 *cpp_flags = tok->flags | add_flags;
514 c_common_no_more_pch ();
517 timevar_pop (TV_CPP);
522 /* Returns the narrowest C-visible unsigned type, starting with the
523 minimum specified by FLAGS, that can fit HIGH:LOW, or itk_none if
526 static enum integer_type_kind
527 narrowest_unsigned_type (unsigned HOST_WIDE_INT low,
528 unsigned HOST_WIDE_INT high,
533 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
534 itk = itk_unsigned_int;
535 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
536 itk = itk_unsigned_long;
538 itk = itk_unsigned_long_long;
540 for (; itk < itk_none; itk += 2 /* skip unsigned types */)
544 if (integer_types[itk] == NULL_TREE)
546 upper = TYPE_MAX_VALUE (integer_types[itk]);
548 if ((unsigned HOST_WIDE_INT) TREE_INT_CST_HIGH (upper) > high
549 || ((unsigned HOST_WIDE_INT) TREE_INT_CST_HIGH (upper) == high
550 && TREE_INT_CST_LOW (upper) >= low))
551 return (enum integer_type_kind) itk;
557 /* Ditto, but narrowest signed type. */
558 static enum integer_type_kind
559 narrowest_signed_type (unsigned HOST_WIDE_INT low,
560 unsigned HOST_WIDE_INT high, unsigned int flags)
564 if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
566 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
572 for (; itk < itk_none; itk += 2 /* skip signed types */)
576 if (integer_types[itk] == NULL_TREE)
578 upper = TYPE_MAX_VALUE (integer_types[itk]);
580 if ((unsigned HOST_WIDE_INT) TREE_INT_CST_HIGH (upper) > high
581 || ((unsigned HOST_WIDE_INT) TREE_INT_CST_HIGH (upper) == high
582 && TREE_INT_CST_LOW (upper) >= low))
583 return (enum integer_type_kind) itk;
589 /* Interpret TOKEN, an integer with FLAGS as classified by cpplib. */
591 interpret_integer (const cpp_token *token, unsigned int flags)
594 enum integer_type_kind itk;
596 cpp_options *options = cpp_get_options (parse_in);
598 integer = cpp_interpret_integer (parse_in, token, flags);
599 integer = cpp_num_sign_extend (integer, options->precision);
601 /* The type of a constant with a U suffix is straightforward. */
602 if (flags & CPP_N_UNSIGNED)
603 itk = narrowest_unsigned_type (integer.low, integer.high, flags);
606 /* The type of a potentially-signed integer constant varies
607 depending on the base it's in, the standard in use, and the
609 enum integer_type_kind itk_u
610 = narrowest_unsigned_type (integer.low, integer.high, flags);
611 enum integer_type_kind itk_s
612 = narrowest_signed_type (integer.low, integer.high, flags);
614 /* In both C89 and C99, octal and hex constants may be signed or
615 unsigned, whichever fits tighter. We do not warn about this
616 choice differing from the traditional choice, as the constant
617 is probably a bit pattern and either way will work. */
618 if ((flags & CPP_N_RADIX) != CPP_N_DECIMAL)
619 itk = MIN (itk_u, itk_s);
622 /* In C99, decimal constants are always signed.
623 In C89, decimal constants that don't fit in long have
624 undefined behavior; we try to make them unsigned long.
625 In GCC's extended C89, that last is true of decimal
626 constants that don't fit in long long, too. */
629 if (itk_s > itk_u && itk_s > itk_long)
633 if (itk_u < itk_unsigned_long)
634 itk_u = itk_unsigned_long;
636 warning (0, "this decimal constant is unsigned only in ISO C90");
639 warning (OPT_Wtraditional,
640 "this decimal constant would be unsigned in ISO C90");
646 /* cpplib has already issued a warning for overflow. */
647 type = ((flags & CPP_N_UNSIGNED)
648 ? widest_unsigned_literal_type_node
649 : widest_integer_literal_type_node);
652 type = integer_types[itk];
653 if (itk > itk_unsigned_long
654 && (flags & CPP_N_WIDTH) != CPP_N_LARGE)
656 ((c_dialect_cxx () ? cxx_dialect == cxx98 : !flag_isoc99)
657 ? DK_PEDWARN : DK_WARNING,
658 input_location, OPT_Wlong_long,
659 (flags & CPP_N_UNSIGNED)
660 ? "integer constant is too large for %<unsigned long%> type"
661 : "integer constant is too large for %<long%> type");
664 value = build_int_cst_wide (type, integer.low, integer.high);
666 /* Convert imaginary to a complex type. */
667 if (flags & CPP_N_IMAGINARY)
668 value = build_complex (NULL_TREE, build_int_cst (type, 0), value);
673 /* Interpret TOKEN, a floating point number with FLAGS as classified
674 by cpplib. For C++0X SUFFIX may contain a user-defined literal suffix. */
676 interpret_float (const cpp_token *token, unsigned int flags,
682 REAL_VALUE_TYPE real;
683 REAL_VALUE_TYPE real_trunc;
687 /* Default (no suffix) depends on whether the FLOAT_CONST_DECIMAL64
688 pragma has been used and is either double or _Decimal64. Types
689 that are not allowed with decimal float default to double. */
690 if (flags & CPP_N_DEFAULT)
692 flags ^= CPP_N_DEFAULT;
693 flags |= CPP_N_MEDIUM;
695 if (((flags & CPP_N_HEX) == 0) && ((flags & CPP_N_IMAGINARY) == 0))
697 warning (OPT_Wunsuffixed_float_constants,
698 "unsuffixed float constant");
699 if (float_const_decimal64_p ())
700 flags |= CPP_N_DFLOAT;
704 /* Decode _Fract and _Accum. */
705 if (flags & CPP_N_FRACT || flags & CPP_N_ACCUM)
706 return interpret_fixed (token, flags);
708 /* Decode type based on width and properties. */
709 if (flags & CPP_N_DFLOAT)
710 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
711 type = dfloat128_type_node;
712 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
713 type = dfloat32_type_node;
715 type = dfloat64_type_node;
717 if (flags & CPP_N_WIDTH_MD)
720 enum machine_mode mode;
722 if ((flags & CPP_N_WIDTH_MD) == CPP_N_MD_W)
727 mode = targetm.c.mode_for_suffix (suffix);
728 if (mode == VOIDmode)
730 error ("unsupported non-standard suffix on floating constant");
732 return error_mark_node;
735 pedwarn (input_location, OPT_pedantic, "non-standard suffix on floating constant");
737 type = c_common_type_for_mode (mode, 0);
740 else if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
741 type = long_double_type_node;
742 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL
743 || flag_single_precision_constant)
744 type = float_type_node;
746 type = double_type_node;
748 const_type = excess_precision_type (type);
752 /* Copy the constant to a nul-terminated buffer. If the constant
753 has any suffixes, cut them off; REAL_VALUE_ATOF/ REAL_VALUE_HTOF
754 can't handle them. */
755 copylen = token->val.str.len;
756 if (flags & CPP_N_USERDEF)
757 copylen -= strlen (suffix);
758 else if (flags & CPP_N_DFLOAT)
762 if ((flags & CPP_N_WIDTH) != CPP_N_MEDIUM)
763 /* Must be an F or L or machine defined suffix. */
765 if (flags & CPP_N_IMAGINARY)
770 copy = (char *) alloca (copylen + 1);
771 memcpy (copy, token->val.str.text, copylen);
772 copy[copylen] = '\0';
774 real_from_string3 (&real, copy, TYPE_MODE (const_type));
775 if (const_type != type)
776 /* Diagnosing if the result of converting the value with excess
777 precision to the semantic type would overflow (with associated
778 double rounding) is more appropriate than diagnosing if the
779 result of converting the string directly to the semantic type
781 real_convert (&real_trunc, TYPE_MODE (type), &real);
783 /* Both C and C++ require a diagnostic for a floating constant
784 outside the range of representable values of its type. Since we
785 have __builtin_inf* to produce an infinity, this is now a
786 mandatory pedwarn if the target does not support infinities. */
787 if (REAL_VALUE_ISINF (real)
788 || (const_type != type && REAL_VALUE_ISINF (real_trunc)))
790 if (!MODE_HAS_INFINITIES (TYPE_MODE (type)))
791 pedwarn (input_location, 0, "floating constant exceeds range of %qT", type);
793 warning (OPT_Woverflow, "floating constant exceeds range of %qT", type);
795 /* We also give a warning if the value underflows. */
796 else if (REAL_VALUES_EQUAL (real, dconst0)
797 || (const_type != type && REAL_VALUES_EQUAL (real_trunc, dconst0)))
799 REAL_VALUE_TYPE realvoidmode;
800 int overflow = real_from_string (&realvoidmode, copy);
801 if (overflow < 0 || !REAL_VALUES_EQUAL (realvoidmode, dconst0))
802 warning (OPT_Woverflow, "floating constant truncated to zero");
805 /* Create a node with determined type and value. */
806 value = build_real (const_type, real);
807 if (flags & CPP_N_IMAGINARY)
809 value = build_complex (NULL_TREE, convert (const_type,
810 integer_zero_node), value);
811 if (type != const_type)
813 const_type = TREE_TYPE (value);
814 type = build_complex_type (type);
818 if (type != const_type)
819 value = build1 (EXCESS_PRECISION_EXPR, type, value);
824 /* Interpret TOKEN, a fixed-point number with FLAGS as classified
828 interpret_fixed (const cpp_token *token, unsigned int flags)
832 FIXED_VALUE_TYPE fixed;
836 copylen = token->val.str.len;
838 if (flags & CPP_N_FRACT) /* _Fract. */
840 if (flags & CPP_N_UNSIGNED) /* Unsigned _Fract. */
842 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
844 type = unsigned_long_long_fract_type_node;
847 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
849 type = unsigned_long_fract_type_node;
852 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
854 type = unsigned_short_fract_type_node;
859 type = unsigned_fract_type_node;
863 else /* Signed _Fract. */
865 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
867 type = long_long_fract_type_node;
870 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
872 type = long_fract_type_node;
875 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
877 type = short_fract_type_node;
882 type = fract_type_node;
889 if (flags & CPP_N_UNSIGNED) /* Unsigned _Accum. */
891 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
893 type = unsigned_long_long_accum_type_node;
896 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
898 type = unsigned_long_accum_type_node;
901 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
903 type = unsigned_short_accum_type_node;
908 type = unsigned_accum_type_node;
912 else /* Signed _Accum. */
914 if ((flags & CPP_N_WIDTH) == CPP_N_LARGE)
916 type = long_long_accum_type_node;
919 else if ((flags & CPP_N_WIDTH) == CPP_N_MEDIUM)
921 type = long_accum_type_node;
924 else if ((flags & CPP_N_WIDTH) == CPP_N_SMALL)
926 type = short_accum_type_node;
931 type = accum_type_node;
937 copy = (char *) alloca (copylen + 1);
938 memcpy (copy, token->val.str.text, copylen);
939 copy[copylen] = '\0';
941 fixed_from_string (&fixed, copy, TYPE_MODE (type));
943 /* Create a node with determined type and value. */
944 value = build_fixed (type, fixed);
949 /* Convert a series of STRING, WSTRING, STRING16, STRING32 and/or
950 UTF8STRING tokens into a tree, performing string constant
951 concatenation. TOK is the first of these. VALP is the location to
952 write the string into. OBJC_STRING indicates whether an '@' token
953 preceded the incoming token (in that case, the strings can either
954 be ObjC strings, preceded by a single '@', or normal strings, not
955 preceded by '@'. The result will be a CPP_OBJC_STRING). Returns
956 the CPP token type of the result (CPP_STRING, CPP_WSTRING,
957 CPP_STRING32, CPP_STRING16, CPP_UTF8STRING, or CPP_OBJC_STRING).
959 This is unfortunately more work than it should be. If any of the
960 strings in the series has an L prefix, the result is a wide string
961 (6.4.5p4). Whether or not the result is a wide string affects the
962 meaning of octal and hexadecimal escapes (6.4.4.4p6,9). But escape
963 sequences do not continue across the boundary between two strings in
964 a series (6.4.5p7), so we must not lose the boundaries. Therefore
965 cpp_interpret_string takes a vector of cpp_string structures, which
966 we must arrange to provide. */
968 static enum cpp_ttype
969 lex_string (const cpp_token *tok, tree *valp, bool objc_string, bool translate)
973 struct obstack str_ob;
975 enum cpp_ttype type = tok->type;
977 /* Try to avoid the overhead of creating and destroying an obstack
978 for the common case of just one string. */
979 cpp_string str = tok->val.str;
980 cpp_string *strs = &str;
982 /* objc_at_sign_was_seen is only used when doing Objective-C string
983 concatenation. It is 'true' if we have seen an '@' before the
984 current string, and 'false' if not. We must see exactly one or
985 zero '@' before each string. */
986 bool objc_at_sign_was_seen = false;
989 tok = cpp_get_token (parse_in);
997 if (objc_at_sign_was_seen)
998 error ("repeated %<@%> before Objective-C string");
1000 objc_at_sign_was_seen = true;
1011 case CPP_UTF8STRING:
1012 if (type != tok->type)
1014 if (type == CPP_STRING)
1017 error ("unsupported non-standard concatenation of string literals");
1023 gcc_obstack_init (&str_ob);
1024 obstack_grow (&str_ob, &str, sizeof (cpp_string));
1028 obstack_grow (&str_ob, &tok->val.str, sizeof (cpp_string));
1030 objc_at_sign_was_seen = false;
1034 /* It is an error if we saw a '@' with no following string. */
1035 if (objc_at_sign_was_seen)
1036 error ("stray %<@%> in program");
1038 /* We have read one more token than we want. */
1039 _cpp_backup_tokens (parse_in, 1);
1041 strs = XOBFINISH (&str_ob, cpp_string *);
1043 if (concats && !objc_string && !in_system_header)
1044 warning (OPT_Wtraditional,
1045 "traditional C rejects string constant concatenation");
1048 ? cpp_interpret_string : cpp_interpret_string_notranslate)
1049 (parse_in, strs, concats + 1, &istr, type))
1051 value = build_string (istr.len, (const char *) istr.text);
1052 free (CONST_CAST (unsigned char *, istr.text));
1056 /* Callers cannot generally handle error_mark_node in this context,
1057 so return the empty string instead. cpp_interpret_string has
1063 case CPP_UTF8STRING:
1064 value = build_string (1, "");
1067 value = build_string (TYPE_PRECISION (char16_type_node)
1068 / TYPE_PRECISION (char_type_node),
1069 "\0"); /* char16_t is 16 bits */
1072 value = build_string (TYPE_PRECISION (char32_type_node)
1073 / TYPE_PRECISION (char_type_node),
1074 "\0\0\0"); /* char32_t is 32 bits */
1077 value = build_string (TYPE_PRECISION (wchar_type_node)
1078 / TYPE_PRECISION (char_type_node),
1079 "\0\0\0"); /* widest supported wchar_t
1089 case CPP_UTF8STRING:
1090 TREE_TYPE (value) = char_array_type_node;
1093 TREE_TYPE (value) = char16_array_type_node;
1096 TREE_TYPE (value) = char32_array_type_node;
1099 TREE_TYPE (value) = wchar_array_type_node;
1101 *valp = fix_string_type (value);
1104 obstack_free (&str_ob, 0);
1106 return objc_string ? CPP_OBJC_STRING : type;
1109 /* Converts a (possibly wide) character constant token into a tree. */
1111 lex_charconst (const cpp_token *token)
1115 unsigned int chars_seen;
1118 result = cpp_interpret_charconst (parse_in, token,
1119 &chars_seen, &unsignedp);
1121 if (token->type == CPP_WCHAR)
1122 type = wchar_type_node;
1123 else if (token->type == CPP_CHAR32)
1124 type = char32_type_node;
1125 else if (token->type == CPP_CHAR16)
1126 type = char16_type_node;
1127 /* In C, a character constant has type 'int'.
1128 In C++ 'char', but multi-char charconsts have type 'int'. */
1129 else if (!c_dialect_cxx () || chars_seen > 1)
1130 type = integer_type_node;
1132 type = char_type_node;
1134 /* Cast to cppchar_signed_t to get correct sign-extension of RESULT
1135 before possibly widening to HOST_WIDE_INT for build_int_cst. */
1136 if (unsignedp || (cppchar_signed_t) result >= 0)
1137 value = build_int_cst_wide (type, result, 0);
1139 value = build_int_cst_wide (type, (cppchar_signed_t) result, -1);