PulkoMandy | 17fc759 | 2022-07-28 18:27:54 +0200 | [diff] [blame] | 1 | /* |
| 2 | * (c) Thomas Pornin 1999, 2000 |
| 3 | * |
| 4 | * Redistribution and use in source and binary forms, with or without |
| 5 | * modification, are permitted provided that the following conditions |
| 6 | * are met: |
| 7 | * 1. Redistributions of source code must retain the above copyright |
| 8 | * notice, this list of conditions and the following disclaimer. |
| 9 | * 2. Redistributions in binary form must reproduce the above copyright |
| 10 | * notice, this list of conditions and the following disclaimer in the |
| 11 | * documentation and/or other materials provided with the distribution. |
| 12 | * 4. The name of the authors may not be used to endorse or promote |
| 13 | * products derived from this software without specific prior written |
| 14 | * permission. |
| 15 | * |
| 16 | * THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR |
| 17 | * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED |
| 18 | * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
| 19 | * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE |
| 20 | * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR |
| 21 | * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT |
| 22 | * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR |
| 23 | * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, |
| 24 | * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE |
| 25 | * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, |
| 26 | * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 27 | * |
| 28 | */ |
| 29 | |
| 30 | #ifndef UCPP__CPP__ |
| 31 | #define UCPP__CPP__ |
| 32 | |
| 33 | /* |
| 34 | * Uncomment the following if you want ucpp to use externally provided |
| 35 | * error-reporting functions (ucpp_warning(), ucpp_error() and ucpp_ouch()) |
| 36 | */ |
| 37 | /* #define NO_UCPP_ERROR_FUNCTIONS */ |
| 38 | |
| 39 | /* |
| 40 | * Tokens (do not change the order unless checking operators_name[] in cpp.c) |
| 41 | * |
| 42 | * It is important that the token NONE is 0 |
| 43 | * Check the STRING_TOKEN macro |
| 44 | */ |
| 45 | #define CPPERR 512 |
| 46 | enum { |
| 47 | NONE, /* whitespace */ |
| 48 | NEWLINE, /* newline */ |
| 49 | COMMENT, /* comment */ |
| 50 | NUMBER, /* number constant */ |
| 51 | NAME, /* identifier */ |
| 52 | BUNCH, /* non-C characters */ |
| 53 | PRAGMA, /* a #pragma directive */ |
| 54 | CONTEXT, /* new file or #line */ |
| 55 | STRING, /* constant "xxx" */ |
| 56 | CHAR, /* constant 'xxx' */ |
| 57 | SLASH, /* / */ |
| 58 | ASSLASH, /* /= */ |
| 59 | MINUS, /* - */ |
| 60 | MMINUS, /* -- */ |
| 61 | ASMINUS, /* -= */ |
| 62 | ARROW, /* -> */ |
| 63 | PLUS, /* + */ |
| 64 | PPLUS, /* ++ */ |
| 65 | ASPLUS, /* += */ |
| 66 | LT, /* < */ |
| 67 | LEQ, /* <= */ |
| 68 | LSH, /* << */ |
| 69 | ASLSH, /* <<= */ |
| 70 | GT, /* > */ |
| 71 | GEQ, /* >= */ |
| 72 | RSH, /* >> */ |
| 73 | ASRSH, /* >>= */ |
| 74 | ASGN, /* = */ |
| 75 | SAME, /* == */ |
| 76 | #ifdef CAST_OP |
| 77 | CAST, /* => */ |
| 78 | #endif |
| 79 | NOT, /* ~ */ |
| 80 | NEQ, /* != */ |
| 81 | AND, /* & */ |
| 82 | LAND, /* && */ |
| 83 | ASAND, /* &= */ |
| 84 | OR, /* | */ |
| 85 | LOR, /* || */ |
| 86 | ASOR, /* |= */ |
| 87 | PCT, /* % */ |
| 88 | ASPCT, /* %= */ |
| 89 | STAR, /* * */ |
| 90 | ASSTAR, /* *= */ |
| 91 | CIRC, /* ^ */ |
| 92 | ASCIRC, /* ^= */ |
| 93 | LNOT, /* ! */ |
| 94 | ASNOT, /* ~= */ |
| 95 | LBRA, /* { */ |
| 96 | RBRA, /* } */ |
| 97 | LBRK, /* [ */ |
| 98 | RBRK, /* ] */ |
| 99 | LPAR, /* ( */ |
| 100 | RPAR, /* ) */ |
| 101 | COMMA, /* , */ |
| 102 | QUEST, /* ? */ |
| 103 | SEMIC, /* ; */ |
| 104 | COLON, /* : */ |
| 105 | DOT, /* . */ |
| 106 | MDOTS, /* ... */ |
| 107 | SHARP, /* # */ |
| 108 | DSHARP, /* ## */ |
| 109 | |
| 110 | OPT_NONE, /* optional space to separate tokens in text output */ |
| 111 | |
| 112 | DIGRAPH_TOKENS, /* there begin digraph tokens */ |
| 113 | |
| 114 | /* for DIG_*, do not change order, unless checking undig() in cpp.c */ |
| 115 | DIG_LBRK, /* <: */ |
| 116 | DIG_RBRK, /* :> */ |
| 117 | DIG_LBRA, /* <% */ |
| 118 | DIG_RBRA, /* %> */ |
| 119 | DIG_SHARP, /* %: */ |
| 120 | DIG_DSHARP, /* %:%: */ |
| 121 | |
| 122 | DIGRAPH_TOKENS_END, /* digraph tokens end here */ |
| 123 | |
| 124 | LAST_MEANINGFUL_TOKEN, /* reserved words will go there */ |
| 125 | |
| 126 | MACROARG, /* special token for representing macro arguments */ |
| 127 | |
| 128 | UPLUS = CPPERR, /* unary + */ |
| 129 | UMINUS /* unary - */ |
| 130 | }; |
| 131 | |
| 132 | #include <stdio.h> |
| 133 | #include <setjmp.h> |
| 134 | #include "tune.h" |
| 135 | |
| 136 | struct token { |
| 137 | int type; |
| 138 | long line; |
| 139 | char *name; |
| 140 | }; |
| 141 | |
| 142 | struct token_fifo { |
| 143 | struct token *t; |
| 144 | size_t nt, art; |
| 145 | }; |
| 146 | |
| 147 | struct lexer_state { |
| 148 | /* input control */ |
| 149 | FILE *input; |
| 150 | #ifndef NO_UCPP_BUF |
| 151 | unsigned char *input_buf; |
| 152 | #ifdef UCPP_MMAP |
| 153 | int from_mmap; |
| 154 | unsigned char *input_buf_sav; |
| 155 | #endif |
| 156 | #endif |
| 157 | unsigned char *input_string; |
| 158 | size_t ebuf; |
| 159 | size_t pbuf; |
| 160 | int lka[2]; |
| 161 | int nlka; |
| 162 | int macfile; |
| 163 | int last; |
| 164 | int discard; |
| 165 | unsigned long utf8; |
| 166 | unsigned char copy_line[COPY_LINE_LENGTH]; |
| 167 | int cli; |
| 168 | |
| 169 | /* output control */ |
| 170 | |
| 171 | FILE *output; |
| 172 | struct token_fifo *output_fifo, *toplevel_of; |
| 173 | #ifndef NO_UCPP_BUF |
| 174 | unsigned char *output_buf; |
| 175 | #endif |
| 176 | size_t sbuf; |
| 177 | |
| 178 | /* token control */ |
| 179 | struct token *ctok; |
| 180 | struct token *save_ctok; |
| 181 | int tknl; |
| 182 | int ltwnl; |
| 183 | int pending_token; |
| 184 | #ifdef INMACRO_FLAG |
| 185 | int inmacro; |
| 186 | long macro_count; |
| 187 | #endif |
| 188 | |
| 189 | /* lexer options */ |
| 190 | long line; |
| 191 | long oline; |
| 192 | unsigned long flags; |
| 193 | long count_trigraphs; |
| 194 | struct garbage_fifo *gf; |
| 195 | int ifnest; |
| 196 | int condnest; |
| 197 | int condcomp; |
| 198 | int condmet; |
| 199 | unsigned long condf[2]; |
| 200 | }; |
| 201 | |
| 202 | /* |
| 203 | * Flags for struct lexer_state |
| 204 | */ |
| 205 | /* warning flags */ |
| 206 | #define WARN_STANDARD 0x000001UL /* emit standard warnings */ |
| 207 | #define WARN_ANNOYING 0x000002UL /* emit annoying warnings */ |
| 208 | #define WARN_TRIGRAPHS 0x000004UL /* warn when trigraphs are used */ |
| 209 | #define WARN_TRIGRAPHS_MORE 0x000008UL /* extra-warn for trigraphs */ |
| 210 | #define WARN_PRAGMA 0x000010UL /* warn for pragmas in non-lexer mode */ |
| 211 | |
| 212 | /* error flags */ |
| 213 | #define FAIL_SHARP 0x000020UL /* emit errors on rogue '#' */ |
| 214 | #define CCHARSET 0x000040UL /* emit errors on non-C characters */ |
| 215 | |
| 216 | /* emission flags */ |
| 217 | #define DISCARD_COMMENTS 0x000080UL /* discard comments from text output */ |
| 218 | #define CPLUSPLUS_COMMENTS 0x000100UL /* understand C++-like comments */ |
| 219 | #define LINE_NUM 0x000200UL /* emit #line directives in output */ |
| 220 | #define GCC_LINE_NUM 0x000400UL /* same as #line, with gcc-syntax */ |
| 221 | |
| 222 | /* language flags */ |
| 223 | #define HANDLE_ASSERTIONS 0x000800UL /* understand assertions */ |
| 224 | #define HANDLE_PRAGMA 0x001000UL /* emit PRAGMA tokens in lexer mode */ |
| 225 | #define MACRO_VAARG 0x002000UL /* understand macros with '...' */ |
| 226 | #define UTF8_SOURCE 0x004000UL /* identifiers are in UTF8 encoding */ |
| 227 | #define HANDLE_TRIGRAPHS 0x008000UL /* handle trigraphs */ |
| 228 | |
| 229 | /* global ucpp behaviour */ |
| 230 | #define LEXER 0x010000UL /* behave as a lexer */ |
| 231 | #define KEEP_OUTPUT 0x020000UL /* emit the result of preprocessing */ |
| 232 | #define COPY_LINE 0x040000UL /* make a copy of the parsed line */ |
| 233 | |
| 234 | /* internal flags */ |
| 235 | #define READ_AGAIN 0x080000UL /* emit again the last token */ |
| 236 | #define TEXT_OUTPUT 0x100000UL /* output text */ |
| 237 | |
| 238 | /* |
| 239 | * Public function prototypes |
| 240 | */ |
| 241 | |
| 242 | #ifndef NO_UCPP_BUF |
| 243 | void flush_output(struct lexer_state *); |
| 244 | #endif |
| 245 | |
| 246 | void init_assertions(void); |
| 247 | int make_assertion(char *); |
| 248 | int destroy_assertion(char *); |
| 249 | void print_assertions(void); |
| 250 | |
| 251 | void init_macros(void); |
| 252 | int define_macro(struct lexer_state *, char *); |
| 253 | int undef_macro(struct lexer_state *, char *); |
| 254 | void print_defines(void); |
| 255 | |
| 256 | void set_init_filename(char *, int); |
| 257 | void init_cpp(void); |
| 258 | void init_include_path(char *[]); |
| 259 | void init_lexer_state(struct lexer_state *); |
| 260 | void init_lexer_mode(struct lexer_state *); |
| 261 | void free_lexer_state(struct lexer_state *); |
| 262 | int lex(struct lexer_state *); |
| 263 | int check_cpp_errors(struct lexer_state *); |
| 264 | void add_incpath(char *); |
| 265 | void init_tables(int); |
| 266 | void enter_file(struct lexer_state *, unsigned long); |
| 267 | int cpp(struct lexer_state *); |
| 268 | |
| 269 | #ifdef UCPP_MMAP |
| 270 | FILE *fopen_mmap_file(char *); |
| 271 | void set_input_file(struct lexer_state *, FILE *); |
| 272 | #endif |
| 273 | |
| 274 | struct stack_context { |
| 275 | char *long_name, *name; |
| 276 | long line; |
| 277 | }; |
| 278 | struct stack_context *report_context(void); |
| 279 | |
| 280 | extern int no_special_macros, system_macros, |
| 281 | emit_dependencies, emit_defines, emit_assertions; |
| 282 | extern int c99_compliant, c99_hosted; |
| 283 | extern FILE *emit_output; |
| 284 | extern char *current_filename, *current_long_filename; |
| 285 | extern char *operators_name[]; |
| 286 | extern struct protect { |
| 287 | char *macro; |
| 288 | int state; |
| 289 | struct found_file *ff; |
| 290 | } protect_detect; |
| 291 | |
| 292 | void ucpp_ouch(char *, ...); |
| 293 | void ucpp_error(long, char *, ...); |
| 294 | void ucpp_warning(long, char *, ...); |
| 295 | |
| 296 | extern int *transient_characters; |
| 297 | |
| 298 | /* |
| 299 | * Errors from CPPERR_EOF and above are not real erros, only show-stoppers. |
| 300 | * Errors below CPPERR_EOF are real ones. |
| 301 | */ |
| 302 | #define CPPERR_NEST 900 |
| 303 | #define CPPERR_EOF 1000 |
| 304 | |
| 305 | /* |
| 306 | * This macro tells whether the name field of a given token type is |
| 307 | * relevant, or not. Irrelevant name field means that it might point |
| 308 | * to outerspace. |
| 309 | */ |
| 310 | #ifdef SEMPER_FIDELIS |
| 311 | #define STRING_TOKEN(x) ((x) == NONE || ((x) >= COMMENT && (x) <= CHAR)) |
| 312 | #else |
| 313 | #define STRING_TOKEN(x) ((x) >= NUMBER && (x) <= CHAR) |
| 314 | #endif |
| 315 | |
| 316 | #endif |