[407] | 1 | %{ |
---|
| 2 | /* |
---|
| 3 | * Bill's ABNF parser. |
---|
| 4 | * Copyright 2002-2007 William C. Fenner <fenner@fenron.com> |
---|
| 5 | * All rights reserved. |
---|
| 6 | * |
---|
| 7 | * Redistribution and use in source and binary forms, with or without |
---|
| 8 | * modification, are permitted provided that the following conditions |
---|
| 9 | * are met: |
---|
| 10 | * 1. Redistributions of source code must retain the above copyright |
---|
| 11 | * notice, this list of conditions and the following disclaimer. |
---|
| 12 | * 2. Redistributions in binary form must reproduce the above copyright |
---|
| 13 | * notice, this list of conditions and the following disclaimer in the |
---|
| 14 | * documentation and/or other materials provided with the distribution. |
---|
| 15 | * 3. Neither the name of the author nor the names of contributors |
---|
| 16 | * may be used to endorse or promote products derived from this software |
---|
| 17 | * without specific prior written permission. |
---|
| 18 | * |
---|
| 19 | * THIS SOFTWARE IS PROVIDED BY WILLIAM C. FENNER ``AS IS'' AND |
---|
| 20 | * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, |
---|
| 21 | * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR |
---|
| 22 | * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL WILLIAM C. FENNER OR HIS |
---|
| 23 | * BROTHER B1FF BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, |
---|
| 24 | * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, |
---|
| 25 | * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR |
---|
| 26 | * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY |
---|
| 27 | * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
---|
| 28 | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE |
---|
| 29 | * USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH |
---|
| 30 | * DAMAGE. |
---|
| 31 | */ |
---|
| 32 | |
---|
| 33 | #include "config.h" |
---|
| 34 | #include <stdlib.h> |
---|
| 35 | #include <string.h> |
---|
| 36 | #include "common.h" |
---|
| 37 | #include "y.tab.h" |
---|
| 38 | |
---|
| 39 | static const char rcsid[] = |
---|
| 40 | "$Id: scanner.l,v 1.1 2008-05-28 12:01:58 jre Exp $"; |
---|
| 41 | |
---|
| 42 | int yycolumn = 0; |
---|
| 43 | int yyerrors = 0; |
---|
| 44 | extern int permissive; |
---|
| 45 | int indent = -1; |
---|
| 46 | |
---|
| 47 | char badchar; |
---|
| 48 | |
---|
| 49 | static void scanrange(char *, int, struct range *); |
---|
| 50 | static char *scanstr(char *, int); |
---|
| 51 | static void gotcr(void); |
---|
| 52 | |
---|
| 53 | %} |
---|
| 54 | |
---|
| 55 | bit [01] |
---|
| 56 | digit [0-9] |
---|
| 57 | hexdig [0-9A-Fa-f] |
---|
| 58 | |
---|
| 59 | rulename [A-Za-z][-0-9A-Za-z]* |
---|
| 60 | wsp [ \t] |
---|
| 61 | |
---|
| 62 | /* *(%x20-21 / %x23-7E) */ |
---|
| 63 | charval [ !#-~]* |
---|
| 64 | |
---|
| 65 | /* *(%x20-3D / %x3F-7E) */ |
---|
| 66 | proseval [ -=?-~]* |
---|
| 67 | |
---|
| 68 | mycrlf (\n\r|\r\n|\r|\n) |
---|
| 69 | |
---|
| 70 | /* %x may be a flex feature; %s works but sometimes results in |
---|
| 71 | * confusing error messages */ |
---|
| 72 | %x SKIP |
---|
| 73 | |
---|
| 74 | /* line number isn't quite being updated properly. |
---|
| 75 | suspect unterminated charval and prosevals. */ |
---|
| 76 | %% |
---|
| 77 | <SKIP>.*{mycrlf} { char q = (badchar == '\'') ? '"' : '\''; |
---|
| 78 | mywarn(MYERROR, "Illegal character %c%c%c - skipping to end of line", q, badchar, q); |
---|
| 79 | gotcr(); |
---|
| 80 | BEGIN(INITIAL); |
---|
| 81 | return CRLF; } |
---|
| 82 | \"{charval}["\r\n] { |
---|
| 83 | char *p; |
---|
| 84 | yycolumn += strlen(yytext); |
---|
| 85 | yylval.string = strdup(yytext + 1); |
---|
| 86 | p = &yylval.string[strlen(yylval.string) - 1]; |
---|
| 87 | if (*p != '"') { |
---|
| 88 | mywarn(MYERROR, "unterminated char-val"); |
---|
| 89 | unput(*p); /* put the cr or lf back */ |
---|
| 90 | } |
---|
| 91 | *p = '\0'; |
---|
| 92 | if (*yylval.string == '\0') |
---|
| 93 | mywarn(MYWARNING, "zero-length char-val"); |
---|
| 94 | return CHARVAL; |
---|
| 95 | } |
---|
| 96 | \<{proseval}[>\r\n] { |
---|
| 97 | char *p; |
---|
| 98 | yycolumn += strlen(yytext); |
---|
| 99 | yylval.string = strdup(yytext + 1); |
---|
| 100 | p = &yylval.string[strlen(yylval.string) - 1]; |
---|
| 101 | if (*p != '>') { |
---|
| 102 | mywarn(MYERROR, "unterminated prose-val"); |
---|
| 103 | unput(*p); /* put the cr or lf back */ |
---|
| 104 | } |
---|
| 105 | *p = '\0'; |
---|
| 106 | return PROSEVAL; |
---|
| 107 | } |
---|
| 108 | {rulename} { |
---|
| 109 | /* record the indentation of the first rule name. */ |
---|
| 110 | if (indent == -1) |
---|
| 111 | indent = yycolumn; |
---|
| 112 | yycolumn += strlen(yytext); |
---|
| 113 | yylval.string = strdup(yytext); |
---|
| 114 | return RULENAME; |
---|
| 115 | } |
---|
| 116 | %[Bb]{bit}+(-|\.\.){bit}+ { |
---|
| 117 | yycolumn += strlen(yytext); |
---|
| 118 | scanrange(yytext + 2, 2, &yylval.range); |
---|
| 119 | return BINVALRANGE; |
---|
| 120 | } |
---|
| 121 | %[Bb]{bit}+(\.{bit}+)* { |
---|
| 122 | yycolumn += strlen(yytext); |
---|
| 123 | yylval.string = scanstr(yytext + 2, 2); |
---|
| 124 | return BINVAL; |
---|
| 125 | } |
---|
| 126 | %[Bb]. { mywarn(MYERROR, "bad bit value"); |
---|
| 127 | badchar = yytext[2]; BEGIN(SKIP); } |
---|
| 128 | %[Dd]{digit}+(-|\.\.){digit}+ { |
---|
| 129 | yycolumn += strlen(yytext); |
---|
| 130 | scanrange(yytext + 2, 10, &yylval.range); |
---|
| 131 | return DECVALRANGE; |
---|
| 132 | } |
---|
| 133 | %[Dd]{digit}+(\.{digit}+)* { |
---|
| 134 | yycolumn += strlen(yytext); |
---|
| 135 | yylval.string = scanstr(yytext + 2, 10); |
---|
| 136 | return DECVAL; |
---|
| 137 | } |
---|
| 138 | %[Dd]. { mywarn(MYERROR, "bad decimal value"); |
---|
| 139 | badchar = yytext[2]; BEGIN(SKIP); } |
---|
| 140 | %[Xx]{hexdig}+(-|\.\.){hexdig}+ { |
---|
| 141 | yycolumn += strlen(yytext); |
---|
| 142 | scanrange(yytext + 2, 16, &yylval.range); |
---|
| 143 | return HEXVALRANGE; |
---|
| 144 | } |
---|
| 145 | %[Xx]{hexdig}+(\.{hexdig}+)* { |
---|
| 146 | yycolumn += strlen(yytext); |
---|
| 147 | yylval.string = scanstr(yytext + 2, 16); |
---|
| 148 | return HEXVAL; |
---|
| 149 | } |
---|
| 150 | %[Xx]. { mywarn(MYERROR, "bad hex value"); |
---|
| 151 | badchar = yytext[2]; BEGIN(SKIP); } |
---|
| 152 | {digit}*\*{digit}* { |
---|
| 153 | char *ep; |
---|
| 154 | |
---|
| 155 | yycolumn += strlen(yytext); |
---|
| 156 | yylval.range.lo = strtoul(yytext, &ep, 10); |
---|
| 157 | if (*ep != '*') { |
---|
| 158 | mywarn(MYERROR, "internal scanner error 1"); |
---|
| 159 | yylval.range.hi = -1; |
---|
| 160 | } else { |
---|
| 161 | yylval.range.hi = strtoul(ep + 1, &ep, 10); |
---|
| 162 | if (*ep) { |
---|
| 163 | mywarn(MYERROR, "internal scanner error 2"); |
---|
| 164 | yylval.range.hi = -1; |
---|
| 165 | } else if (yylval.range.hi == 0) |
---|
| 166 | yylval.range.hi = -1; |
---|
| 167 | } |
---|
| 168 | return REPEAT; |
---|
| 169 | } |
---|
| 170 | {digit}*#{digit}* { |
---|
| 171 | char *ep; |
---|
| 172 | |
---|
| 173 | yycolumn += strlen(yytext); |
---|
| 174 | yylval.range.lo = strtoul(yytext, &ep, 10); |
---|
| 175 | if (*ep != '#') { |
---|
| 176 | mywarn(MYERROR, "internal scanner error 1"); |
---|
| 177 | yylval.range.hi = -1; |
---|
| 178 | } else { |
---|
| 179 | yylval.range.hi = strtoul(ep + 1, &ep, 10); |
---|
| 180 | if (*ep) { |
---|
| 181 | mywarn(MYERROR, "internal scanner error 2"); |
---|
| 182 | yylval.range.hi = -1; |
---|
| 183 | } else if (yylval.range.hi == 0) |
---|
| 184 | yylval.range.hi = -1; |
---|
| 185 | } |
---|
| 186 | return LIST; |
---|
| 187 | } |
---|
| 188 | {digit}+ { |
---|
| 189 | char *ep; |
---|
| 190 | |
---|
| 191 | yycolumn += strlen(yytext); |
---|
| 192 | yylval.range.hi = yylval.range.lo = strtoul(yytext, &ep, 10); |
---|
| 193 | if (*ep) { |
---|
| 194 | mywarn(MYERROR, "internal scanner error 3"); |
---|
| 195 | yylval.range.hi = yylval.range.lo = 42; |
---|
| 196 | } |
---|
| 197 | return REPEAT; |
---|
| 198 | } |
---|
| 199 | =\/ { yycolumn += 2; return EQSLASH; } |
---|
| 200 | ({wsp}+|(;[^\r\n]*)|{mycrlf}{wsp}+)+ { char *p = yytext; |
---|
| 201 | while (*p) { |
---|
| 202 | /* TO DO: |
---|
| 203 | * deal with indent if we |
---|
| 204 | * have one set - if a blank |
---|
| 205 | * line or a comment is indented |
---|
| 206 | * less than enough, we warn |
---|
| 207 | * about it. */ |
---|
| 208 | if (*p == '\r') { |
---|
| 209 | gotcr(); |
---|
| 210 | if (*(++p) == '\n') |
---|
| 211 | p++; |
---|
| 212 | continue; |
---|
| 213 | } |
---|
| 214 | if (*p == '\n') { |
---|
| 215 | gotcr(); |
---|
| 216 | if (*(++p) == '\r') |
---|
| 217 | p++; |
---|
| 218 | continue; |
---|
| 219 | } |
---|
| 220 | p++; |
---|
| 221 | yycolumn++; |
---|
| 222 | } |
---|
| 223 | /* If we don't know the indent yet, then just |
---|
| 224 | ignore leading whitespace. */ |
---|
| 225 | if (indent == -1) |
---|
| 226 | continue; |
---|
| 227 | /* If there is more whitespace than |
---|
| 228 | the initial indent, then tell the parser |
---|
| 229 | about the leading whitespace. */ |
---|
| 230 | if (yycolumn > indent) |
---|
| 231 | return CWSP; |
---|
| 232 | if (yycolumn < indent) { |
---|
| 233 | indent = yycolumn; |
---|
| 234 | mywarn(MYERROR, "adjusting indentation"); |
---|
| 235 | } |
---|
| 236 | /* Since we didn't have more whitespace than |
---|
| 237 | indent, tell the parser it was just |
---|
| 238 | a CR. */ |
---|
| 239 | return CRLF; } |
---|
| 240 | {mycrlf} { gotcr(); return CRLF; } |
---|
| 241 | [][()=/] { yycolumn++; return yytext[0]; } |
---|
| 242 | \| { yycolumn++; |
---|
| 243 | if (!permissive) { |
---|
| 244 | badchar = yytext[0]; |
---|
| 245 | BEGIN(SKIP); |
---|
| 246 | } |
---|
| 247 | return yytext[0]; } |
---|
| 248 | . { yycolumn++; badchar = yytext[0]; BEGIN(SKIP); } |
---|
| 249 | %% |
---|
| 250 | |
---|
| 251 | static void |
---|
| 252 | scanrange(char *p, int base, struct range *r) |
---|
| 253 | { |
---|
| 254 | char *ep; |
---|
| 255 | |
---|
| 256 | r->lo = strtoul(p, &ep, base); |
---|
| 257 | if (*ep != '-' && *ep != '.') { |
---|
| 258 | mywarn(MYERROR, "internal scanner error 4"); |
---|
| 259 | r->hi = r->lo; |
---|
| 260 | return; |
---|
| 261 | } |
---|
| 262 | if (*ep == '.') { |
---|
| 263 | if (!permissive) { |
---|
| 264 | badchar = '.'; |
---|
| 265 | BEGIN(SKIP); |
---|
| 266 | } |
---|
| 267 | mywarn(MYERROR, "Ranges use \"-\", not \"..\"."); |
---|
| 268 | ep++; |
---|
| 269 | } |
---|
| 270 | r->hi = strtoul(ep + 1, &ep, base); |
---|
| 271 | if (*ep) { |
---|
| 272 | mywarn(MYERROR, "internal scanner error 5"); |
---|
| 273 | } |
---|
| 274 | if (r->hi < r->lo) { |
---|
| 275 | mywarn(MYERROR, "inverted range"); |
---|
| 276 | } |
---|
| 277 | return; |
---|
| 278 | } |
---|
| 279 | |
---|
| 280 | static char * |
---|
| 281 | scanstr(char *p, int base) |
---|
| 282 | { |
---|
| 283 | char *ep; |
---|
| 284 | char buf[512]; /*XXX*/ |
---|
| 285 | char *b = buf; |
---|
| 286 | int i; |
---|
| 287 | |
---|
| 288 | do { |
---|
| 289 | i = strtoul(p, &ep, base); |
---|
| 290 | if (i > 255) { /* XXX */ |
---|
| 291 | mywarn(MYWARNING, "I can't handle this legal ABNF char value"); |
---|
| 292 | i = 255; |
---|
| 293 | } |
---|
| 294 | if (i == 0) { |
---|
| 295 | mywarn(MYERROR, "This parser will truncate strings at %%x00"); |
---|
| 296 | } |
---|
| 297 | *b++ = i; |
---|
| 298 | p = ep + 1; |
---|
| 299 | } while (*ep == '.'); |
---|
| 300 | if (*ep) |
---|
| 301 | mywarn(MYERROR, "internal scanner error 6"); |
---|
| 302 | *b++ = '\0'; |
---|
| 303 | return strdup(buf); |
---|
| 304 | } |
---|
| 305 | |
---|
| 306 | static void |
---|
| 307 | gotcr(void) |
---|
| 308 | { |
---|
| 309 | yylineno++; |
---|
| 310 | yycolumn = 0; |
---|
| 311 | } |
---|
| 312 | |
---|
| 313 | void |
---|
| 314 | scanreset(void) { |
---|
| 315 | yylineno = 0; |
---|
| 316 | yycolumn = 0; |
---|
| 317 | indent = -1; |
---|
| 318 | } |
---|