1 /* $Vendor-Id: man_validate.c,v 1.47 2010/07/22 23:03:15 kristaps Exp $ */ 2 /* 3 * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@bsd.lv> 4 * 5 * Permission to use, copy, modify, and distribute this software for any 6 * purpose with or without fee is hereby granted, provided that the above 7 * copyright notice and this permission notice appear in all copies. 8 * 9 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES 10 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF 11 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR 12 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES 13 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN 14 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF 15 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. 16 */ 17 #ifdef HAVE_CONFIG_H 18 #include "config.h" 19 #endif 20 21 #include <sys/types.h> 22 23 #include <assert.h> 24 #include <ctype.h> 25 #include <errno.h> 26 #include <limits.h> 27 #include <stdarg.h> 28 #include <stdlib.h> 29 #include <string.h> 30 31 #include "mandoc.h" 32 #include "libman.h" 33 #include "libmandoc.h" 34 35 #define CHKARGS struct man *m, struct man_node *n 36 37 typedef int (*v_check)(CHKARGS); 38 39 struct man_valid { 40 v_check *pres; 41 v_check *posts; 42 }; 43 44 static int check_bline(CHKARGS); 45 static int check_eq0(CHKARGS); 46 static int check_le1(CHKARGS); 47 static int check_ge2(CHKARGS); 48 static int check_le5(CHKARGS); 49 static int check_par(CHKARGS); 50 static int check_part(CHKARGS); 51 static int check_root(CHKARGS); 52 static int check_sec(CHKARGS); 53 static int check_text(CHKARGS); 54 static int check_title(CHKARGS); 55 56 static v_check posts_eq0[] = { check_eq0, NULL }; 57 static v_check posts_th[] = { check_ge2, check_le5, check_title, NULL }; 58 static v_check posts_par[] = { check_par, NULL }; 59 static v_check posts_part[] = { check_part, NULL }; 60 static v_check posts_sec[] = { check_sec, NULL }; 61 static v_check posts_le1[] = { check_le1, NULL }; 62 static v_check pres_bline[] = { check_bline, NULL }; 63 64 static const struct man_valid man_valids[MAN_MAX] = { 65 { NULL, posts_eq0 }, /* br */ 66 { pres_bline, posts_th }, /* TH */ 67 { pres_bline, posts_sec }, /* SH */ 68 { pres_bline, posts_sec }, /* SS */ 69 { pres_bline, posts_par }, /* TP */ 70 { pres_bline, posts_par }, /* LP */ 71 { pres_bline, posts_par }, /* PP */ 72 { pres_bline, posts_par }, /* P */ 73 { pres_bline, posts_par }, /* IP */ 74 { pres_bline, posts_par }, /* HP */ 75 { NULL, NULL }, /* SM */ 76 { NULL, NULL }, /* SB */ 77 { NULL, NULL }, /* BI */ 78 { NULL, NULL }, /* IB */ 79 { NULL, NULL }, /* BR */ 80 { NULL, NULL }, /* RB */ 81 { NULL, NULL }, /* R */ 82 { NULL, NULL }, /* B */ 83 { NULL, NULL }, /* I */ 84 { NULL, NULL }, /* IR */ 85 { NULL, NULL }, /* RI */ 86 { NULL, posts_eq0 }, /* na */ /* FIXME: should warn only. */ 87 { NULL, NULL }, /* i */ 88 { NULL, posts_le1 }, /* sp */ /* FIXME: should warn only. */ 89 { pres_bline, posts_eq0 }, /* nf */ 90 { pres_bline, posts_eq0 }, /* fi */ 91 { NULL, NULL }, /* r */ 92 { NULL, NULL }, /* RE */ 93 { NULL, posts_part }, /* RS */ 94 { NULL, NULL }, /* DT */ 95 { NULL, NULL }, /* UC */ 96 { NULL, NULL }, /* PD */ 97 { NULL, posts_le1 }, /* Sp */ /* FIXME: should warn only. */ 98 { pres_bline, posts_le1 }, /* Vb */ /* FIXME: should warn only. */ 99 { pres_bline, posts_eq0 }, /* Ve */ 100 { NULL, NULL }, /* AT */ 101 { NULL, NULL }, /* in */ 102 }; 103 104 105 int 106 man_valid_pre(struct man *m, struct man_node *n) 107 { 108 v_check *cp; 109 110 if (MAN_TEXT == n->type) 111 return(1); 112 if (MAN_ROOT == n->type) 113 return(1); 114 115 if (NULL == (cp = man_valids[n->tok].pres)) 116 return(1); 117 for ( ; *cp; cp++) 118 if ( ! (*cp)(m, n)) 119 return(0); 120 return(1); 121 } 122 123 124 int 125 man_valid_post(struct man *m) 126 { 127 v_check *cp; 128 129 if (MAN_VALID & m->last->flags) 130 return(1); 131 m->last->flags |= MAN_VALID; 132 133 switch (m->last->type) { 134 case (MAN_TEXT): 135 return(check_text(m, m->last)); 136 case (MAN_ROOT): 137 return(check_root(m, m->last)); 138 default: 139 break; 140 } 141 142 if (NULL == (cp = man_valids[m->last->tok].posts)) 143 return(1); 144 for ( ; *cp; cp++) 145 if ( ! (*cp)(m, m->last)) 146 return(0); 147 148 return(1); 149 } 150 151 152 static int 153 check_root(CHKARGS) 154 { 155 156 if (MAN_BLINE & m->flags) 157 return(man_nmsg(m, n, MANDOCERR_SCOPEEXIT)); 158 if (MAN_ELINE & m->flags) 159 return(man_nmsg(m, n, MANDOCERR_SCOPEEXIT)); 160 161 m->flags &= ~MAN_BLINE; 162 m->flags &= ~MAN_ELINE; 163 164 if (NULL == m->first->child) { 165 man_nmsg(m, n, MANDOCERR_NODOCBODY); 166 return(0); 167 } else if (NULL == m->meta.title) { 168 if ( ! man_nmsg(m, n, MANDOCERR_NOTITLE)) 169 return(0); 170 /* 171 * If a title hasn't been set, do so now (by 172 * implication, date and section also aren't set). 173 * 174 * FIXME: this should be in man_action.c. 175 */ 176 m->meta.title = mandoc_strdup("unknown"); 177 m->meta.date = time(NULL); 178 m->meta.msec = mandoc_strdup("1"); 179 } 180 181 return(1); 182 } 183 184 185 static int 186 check_title(CHKARGS) 187 { 188 const char *p; 189 190 assert(n->child); 191 /* FIXME: is this sufficient? */ 192 if ('\0' == *n->child->string) { 193 man_nmsg(m, n, MANDOCERR_SYNTARGCOUNT); 194 return(0); 195 } 196 197 for (p = n->child->string; '\0' != *p; p++) 198 if (isalpha((u_char)*p) && ! isupper((u_char)*p)) 199 if ( ! man_nmsg(m, n, MANDOCERR_UPPERCASE)) 200 return(0); 201 202 return(1); 203 } 204 205 206 static int 207 check_text(CHKARGS) 208 { 209 char *p; 210 int pos, c; 211 size_t sz; 212 213 for (p = n->string, pos = n->pos + 1; *p; p++, pos++) { 214 sz = strcspn(p, "\t\\"); 215 p += (int)sz; 216 217 if ('\0' == *p) 218 break; 219 220 pos += (int)sz; 221 222 if ('\t' == *p) { 223 if (MAN_LITERAL & m->flags) 224 continue; 225 if (man_pmsg(m, n->line, pos, MANDOCERR_BADTAB)) 226 continue; 227 return(0); 228 } 229 230 /* Check the special character. */ 231 232 c = mandoc_special(p); 233 if (c) { 234 p += c - 1; 235 pos += c - 1; 236 continue; 237 } 238 239 c = man_pmsg(m, n->line, pos, MANDOCERR_BADESCAPE); 240 if ( ! (MAN_IGN_ESCAPE & m->pflags) && ! c) 241 return(c); 242 } 243 244 return(1); 245 } 246 247 248 #define INEQ_DEFINE(x, ineq, name) \ 249 static int \ 250 check_##name(CHKARGS) \ 251 { \ 252 if (n->nchild ineq (x)) \ 253 return(1); \ 254 man_vmsg(m, MANDOCERR_SYNTARGCOUNT, n->line, n->pos, \ 255 "line arguments %s %d (have %d)", \ 256 #ineq, (x), n->nchild); \ 257 return(0); \ 258 } 259 260 INEQ_DEFINE(0, ==, eq0) 261 INEQ_DEFINE(1, <=, le1) 262 INEQ_DEFINE(2, >=, ge2) 263 INEQ_DEFINE(5, <=, le5) 264 265 266 static int 267 check_sec(CHKARGS) 268 { 269 270 if (MAN_HEAD == n->type && 0 == n->nchild) { 271 man_nmsg(m, n, MANDOCERR_SYNTARGCOUNT); 272 return(0); 273 } else if (MAN_BODY == n->type && 0 == n->nchild) 274 return(man_nmsg(m, n, MANDOCERR_NOBODY)); 275 276 return(1); 277 } 278 279 280 static int 281 check_part(CHKARGS) 282 { 283 284 if (MAN_BODY == n->type && 0 == n->nchild) 285 return(man_nmsg(m, n, MANDOCERR_NOBODY)); 286 return(1); 287 } 288 289 290 static int 291 check_par(CHKARGS) 292 { 293 294 if (MAN_BODY == n->type) 295 switch (n->tok) { 296 case (MAN_IP): 297 /* FALLTHROUGH */ 298 case (MAN_HP): 299 /* FALLTHROUGH */ 300 case (MAN_TP): 301 /* Body-less lists are ok. */ 302 break; 303 default: 304 if (n->nchild) 305 break; 306 return(man_nmsg(m, n, MANDOCERR_NOBODY)); 307 } 308 if (MAN_HEAD == n->type) 309 switch (n->tok) { 310 case (MAN_PP): 311 /* FALLTHROUGH */ 312 case (MAN_P): 313 /* FALLTHROUGH */ 314 case (MAN_LP): 315 if (0 == n->nchild) 316 break; 317 return(man_nmsg(m, n, MANDOCERR_ARGSLOST)); 318 default: 319 if (n->nchild) 320 break; 321 return(man_nmsg(m, n, MANDOCERR_NOARGS)); 322 } 323 324 return(1); 325 } 326 327 328 static int 329 check_bline(CHKARGS) 330 { 331 332 assert( ! (MAN_ELINE & m->flags)); 333 if (MAN_BLINE & m->flags) { 334 man_nmsg(m, n, MANDOCERR_SYNTLINESCOPE); 335 return(0); 336 } 337 338 return(1); 339 } 340 341