1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/src/parse.c	Fri Sep 26 17:05:23 2008 +0200
     1.3 @@ -0,0 +1,450 @@
     1.4 +/*  MasqMail
     1.5 +    Copyright (C) 1999-2001 Oliver Kurth
     1.6 +
     1.7 +    This program is free software; you can redistribute it and/or modify
     1.8 +    it under the terms of the GNU General Public License as published by
     1.9 +    the Free Software Foundation; either version 2 of the License, or
    1.10 +    (at your option) any later version.
    1.11 +
    1.12 +    This program is distributed in the hope that it will be useful,
    1.13 +    but WITHOUT ANY WARRANTY; without even the implied warranty of
    1.14 +    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    1.15 +    GNU General Public License for more details.
    1.16 +
    1.17 +    You should have received a copy of the GNU General Public License
    1.18 +    along with this program; if not, write to the Free Software
    1.19 +    Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
    1.20 +*/
    1.21 +
    1.22 +#ifndef PARSE_TEST
    1.23 +#include "masqmail.h"
    1.24 +#endif
    1.25 +
    1.26 +/* This is really dangerous. I hope that I was careful enough,
    1.27 +   but maybe there is some malformed address possible that causes
    1.28 +   this to segfault or be caught in endless loops.
    1.29 +
    1.30 +   If you find something like that, PLEASE mail the string to me
    1.31 +   (no matter how idiotic it is), so that I can debug that.
    1.32 +   Those things really should not happen.
    1.33 +*/
    1.34 +
    1.35 +static gchar *specials = "()<>@,;:\\\".[]`";
    1.36 +
    1.37 +char *parse_error = NULL;
    1.38 +
    1.39 +static
    1.40 +gchar *skip_comment(gchar *p)
    1.41 +{
    1.42 +
    1.43 +#ifdef PARSE_TEST
    1.44 +  g_print("skip_comment: %s\n", p);
    1.45 +#endif
    1.46 +
    1.47 +  p++;
    1.48 +  while(*p && *p != ')'){
    1.49 +    p++;
    1.50 +    if(*p == '(')
    1.51 +      p = skip_comment(p);
    1.52 +  }
    1.53 +  p++;
    1.54 +
    1.55 +  return p;
    1.56 +}
    1.57 +
    1.58 +static
    1.59 +gboolean read_word(gchar *p, gchar **b, gchar **e)
    1.60 +{
    1.61 +#ifdef PARSE_TEST
    1.62 +  g_print("read_word: %s\n", p);
    1.63 +#endif
    1.64 +  /* eat leading spaces */
    1.65 +  while(*p && isspace(*p)) p++;
    1.66 +  
    1.67 +  *b = p;
    1.68 +  /*  b = &p;*/
    1.69 +  if(*p == '\"'){
    1.70 +    /* quoted-string */
    1.71 +    p++;
    1.72 +    while(*p && (*p != '\"')) p++;
    1.73 +    p++;
    1.74 +  }else{
    1.75 +    /* atom */
    1.76 +    while(*p && !strchr(specials, *p) && !iscntrl(*p) && !isspace(*p))
    1.77 +      p++;
    1.78 +  }
    1.79 +  *e = p;
    1.80 +  return TRUE;
    1.81 +}
    1.82 +
    1.83 +static
    1.84 +gboolean read_word_with_dots(gchar *p, gchar **b, gchar **e)
    1.85 +{
    1.86 +  gchar *b0 = p;
    1.87 +
    1.88 +#ifdef PARSE_TEST
    1.89 +  g_print("read_word_with_dots: %s\n", p);
    1.90 +#endif
    1.91 +  while(TRUE){
    1.92 +    if(!read_word(p, b, e))
    1.93 +      return FALSE;
    1.94 +    p = *e;
    1.95 +    if(*p != '.') break;
    1.96 +    p++;
    1.97 +  }
    1.98 +  *b = b0;
    1.99 +  *e = p;
   1.100 +  return TRUE;
   1.101 +}
   1.102 +
   1.103 +static
   1.104 +gboolean read_domain(gchar *p, gchar **b, gchar **e)
   1.105 +{
   1.106 +#ifdef PARSE_TEST
   1.107 +  g_print("read_domain: %s\n", p);
   1.108 +#endif
   1.109 +  *b = p;
   1.110 +  if(*p != '['){
   1.111 +    while(isalnum(*p) || (*p == '-') || (*p == '.'))
   1.112 +      p++;
   1.113 +  }else{
   1.114 +    p++;
   1.115 +    while(isalpha(*p) || (*p == '.'))
   1.116 +      p++;
   1.117 +    if(*p != ']'){
   1.118 +      parse_error =
   1.119 +	g_strdup_printf("']' expected at end of literal address %s", *b);
   1.120 +      return FALSE;
   1.121 +    }
   1.122 +    p++;
   1.123 +  }
   1.124 +  *e = p;
   1.125 +  return TRUE;
   1.126 +}
   1.127 +
   1.128 +gboolean parse_address_rfc822(gchar *string,
   1.129 +		       gchar **local_begin, gchar **local_end,
   1.130 +		       gchar **domain_begin, gchar **domain_end,
   1.131 +		       gchar **address_end)
   1.132 +{
   1.133 +  gint angle_brackets = 0;
   1.134 +
   1.135 +  gchar *p = string;
   1.136 +  gchar *b, *e;
   1.137 +
   1.138 +  *local_begin = *local_end = NULL;
   1.139 +  *domain_begin = *domain_end = NULL;
   1.140 +
   1.141 +  /* might be some memory left from previous call: */
   1.142 +  if(parse_error != NULL){
   1.143 +    g_free(parse_error);
   1.144 +    parse_error = NULL;
   1.145 +  }
   1.146 +
   1.147 +  /* leading spaces and angle brackets */
   1.148 +  while(*p && (isspace(*p) || (*p == '<'))){
   1.149 +    if(*p == '<')
   1.150 +      angle_brackets++;
   1.151 +    p++;
   1.152 +  }
   1.153 + 
   1.154 +  if(*p){
   1.155 +    while(TRUE){
   1.156 +      if(read_word_with_dots(p, &b, &e)){
   1.157 +	p = e;
   1.158 +#ifdef PARSE_TEST
   1.159 +	g_print("after read_word_with_dots: %s\n", p);
   1.160 +#endif
   1.161 +	/* eat white spaces and comments */
   1.162 +	while((*p && (isspace(*p))) || (*p == '(')){
   1.163 +	  if(*p == '('){
   1.164 +	    if(!(p = skip_comment(p))){
   1.165 +	      parse_error =
   1.166 +		g_strdup("missing right bracket ')'");
   1.167 +	      return FALSE;
   1.168 +	    }
   1.169 +	  }else
   1.170 +	    p++;
   1.171 +	}
   1.172 +	/* we now have a non-space char that is not
   1.173 +	   the beginning of a comment */
   1.174 +
   1.175 +	if(*p == '@'){
   1.176 +	  /* the last word was the local_part
   1.177 +	     of an addr-spec */
   1.178 +	  *local_begin = b;
   1.179 +	  *local_end = e;
   1.180 +#ifdef PARSE_TEST
   1.181 +	  g_print("found local part: %s\n", *local_begin);
   1.182 +#endif
   1.183 +	  if(*p == '@'){
   1.184 +	    p++; /* skip @ */
   1.185 +	    /* now the domain */
   1.186 +	    if(read_domain(p, &b, &e)){
   1.187 +	      p = e;
   1.188 +	      *domain_begin = b;
   1.189 +	      *domain_end = e;
   1.190 +	    }
   1.191 +	    else
   1.192 +	      return FALSE;
   1.193 +	  }else{
   1.194 +	    /* unqualified? */
   1.195 +	    *domain_begin = *domain_end = NULL;
   1.196 +	  }
   1.197 +	  break;
   1.198 +	}else if(*p == '<'){
   1.199 +	  /* addr-spec follows */
   1.200 +	  while(isspace(*p) || (*p == '<')){
   1.201 +	    if(*p == '<')
   1.202 +	      angle_brackets++;
   1.203 +	    p++;
   1.204 +	  }
   1.205 +	  if(read_word_with_dots(p, &b, &e)){
   1.206 +	    p = e;
   1.207 +	    *local_begin = b;
   1.208 +	    *local_end = e;
   1.209 +#ifdef PARSE_TEST
   1.210 +	  g_print("found local part: %s\n", *local_begin);
   1.211 +#endif
   1.212 +	  }else
   1.213 +	    return FALSE;
   1.214 +	  if(*p == '@'){
   1.215 +	    p++;
   1.216 +	    if(read_domain(p, &b, &e)){
   1.217 +	      p = e;
   1.218 +	      *domain_begin = b;
   1.219 +	      *domain_end = e;
   1.220 +	    }else
   1.221 +	      return FALSE;
   1.222 +	  }else{
   1.223 +	    /* may be unqualified address */
   1.224 +	    *domain_begin = *domain_end = NULL;
   1.225 +	  }
   1.226 +	  break;
   1.227 +	}else if(!*p || *p == '>'){
   1.228 +	  *local_begin = b;
   1.229 +	  *local_end = e;
   1.230 +#ifdef PARSE_TEST
   1.231 +	  g_print("found local part: %s\n", *local_begin);
   1.232 +#endif
   1.233 +	  *domain_begin = *domain_end = NULL;
   1.234 +	  break;
   1.235 +	}else if(strchr(specials, *p) || iscntrl(*p) || isspace(*p)){
   1.236 +	  parse_error = g_strdup_printf("unexpected character: %c", *p);
   1.237 +	  return FALSE;
   1.238 +	}
   1.239 +      }else
   1.240 +	return FALSE;
   1.241 +    }
   1.242 +   /* trailing spaces and angle brackets */
   1.243 +#ifdef PARSE_TEST
   1.244 +    g_print("down counting trailing '>'\n");
   1.245 +#endif
   1.246 +    while(*p && (isspace(*p) || (*p == '>'))){
   1.247 +      if(*p == '>')
   1.248 +	angle_brackets--;
   1.249 +      p++;
   1.250 +    }
   1.251 +
   1.252 +    *address_end = p;
   1.253 +
   1.254 +    if(angle_brackets != 0){
   1.255 +      if(angle_brackets > 0)
   1.256 +	parse_error = g_strdup("missing '>' at end of string");
   1.257 +      else
   1.258 +	parse_error = g_strdup("superfluous '>' at end of string");
   1.259 +      return FALSE;
   1.260 +    }else{
   1.261 +      /* we successfully parsed the address */
   1.262 +      return TRUE;
   1.263 +    }
   1.264 +    /* we never get here */
   1.265 +  }
   1.266 +  return FALSE;
   1.267 +}
   1.268 +
   1.269 +gboolean parse_address_rfc821(gchar *string,
   1.270 +			      gchar **local_begin, gchar **local_end,
   1.271 +			      gchar **domain_begin, gchar **domain_end,
   1.272 +			      gchar **address_end)
   1.273 +{
   1.274 +  gint angle_brackets = 0;
   1.275 +
   1.276 +  gchar *p = string;
   1.277 +  gchar *b, *e;
   1.278 +
   1.279 +  *local_begin = *local_end = NULL;
   1.280 +  *domain_begin = *domain_end = NULL;
   1.281 +
   1.282 +  /* might be some memory left from previous call: */
   1.283 +  if(parse_error != NULL){
   1.284 +    g_free(parse_error);
   1.285 +    parse_error = NULL;
   1.286 +  }
   1.287 +
   1.288 +  /* leading spaces and angle brackets */
   1.289 +  while(*p && (isspace(*p) || (*p == '<'))){
   1.290 +    if(*p == '<')
   1.291 +      angle_brackets++;
   1.292 +    p++;
   1.293 +  }
   1.294 + 
   1.295 +  if(*p){
   1.296 +    while(TRUE){
   1.297 +      if(read_word_with_dots(p, &b, &e)){
   1.298 +	p = e;
   1.299 +#ifdef PARSE_TEST
   1.300 +	g_print("after read_word_with_dots: %s\n", p);
   1.301 +#endif
   1.302 +	*local_begin = b;
   1.303 +	*local_end = e;
   1.304 +#ifdef PARSE_TEST
   1.305 +	  g_print("found local part: %s\n", *local_begin);
   1.306 +	  g_print("local_end = %s\n", *local_end);
   1.307 +#endif
   1.308 +	if(!(*p) || isspace(*p) || (*p == '>')){
   1.309 +	  /* unqualified ?*/
   1.310 +	  domain_begin = domain_end = NULL;
   1.311 +	  break;
   1.312 +	}else if(*p == '@'){
   1.313 +	  p++;
   1.314 +	  if(read_domain(p, &b, &e)){
   1.315 +	    p = e;
   1.316 +	    *domain_begin = b;
   1.317 +	    *domain_end = e;
   1.318 +	  }
   1.319 +	  break;
   1.320 +	}else{
   1.321 +	  parse_error =
   1.322 +	      g_strdup_printf("unexpected character after local part '%c'",*p);
   1.323 +	  return FALSE;
   1.324 +	}
   1.325 +      } else
   1.326 +         return FALSE;
   1.327 +    }
   1.328 +
   1.329 +    /* trailing spaces and angle brackets */
   1.330 +#ifdef PARSE_TEST
   1.331 +    g_print("down counting trailing '>'\n");
   1.332 +#endif
   1.333 +    while(*p && (isspace(*p) || (*p == '>'))){
   1.334 +      if(*p == '>')
   1.335 +	angle_brackets--;
   1.336 +      p++;
   1.337 +    }
   1.338 +    *address_end = p;
   1.339 +
   1.340 +    if(angle_brackets != 0){
   1.341 +      if(angle_brackets > 0)
   1.342 +	parse_error = g_strdup("missing '>' at end of string");
   1.343 +      else
   1.344 +	parse_error = g_strdup("superfluous '>' at end of string");
   1.345 +      return FALSE;
   1.346 +    }else{
   1.347 +      /* we successfully parsed the address */
   1.348 +      return TRUE;
   1.349 +    }
   1.350 +    /* we never get here */
   1.351 +  }
   1.352 +  return FALSE;
   1.353 +}
   1.354 +
   1.355 +/*
   1.356 +  allocate address, reading from string.
   1.357 +  On failure, returns NULL.
   1.358 +  after call, end contatins a pointer to the end of the parsed string
   1.359 +  end may be NULL, if we are not interested.
   1.360 +
   1.361 +  parses both rfc 821 and rfc 822 addresses, depending on flag is_rfc821
   1.362 +*/
   1.363 +
   1.364 +address *_create_address(gchar *string, gchar **end, gboolean is_rfc821)
   1.365 +{
   1.366 +  gchar *loc_beg, *loc_end;
   1.367 +  gchar *dom_beg, *dom_end;
   1.368 +  gchar *addr_end;
   1.369 +
   1.370 +  if (string && (string[0] == 0)) {
   1.371 +    address *addr = g_malloc(sizeof(address));
   1.372 +    addr->address = g_strdup("");
   1.373 +    addr->local_part = g_strdup("");
   1.374 +    addr->domain = g_strdup(""); /* 'NULL' address (failure notice),
   1.375 +        "" makes sure it will not be qualified with a hostname */
   1.376 +    return addr;
   1.377 +  }
   1.378 +
   1.379 +  if(is_rfc821 ?
   1.380 +     parse_address_rfc821(string,
   1.381 +			  &loc_beg, &loc_end, &dom_beg, &dom_end, &addr_end) :
   1.382 +     parse_address_rfc822(string,
   1.383 +			  &loc_beg, &loc_end, &dom_beg, &dom_end, &addr_end)){
   1.384 +    address *addr = g_malloc(sizeof(address));
   1.385 +    gchar *p = addr_end;
   1.386 +    
   1.387 +
   1.388 +    memset(addr, 0, sizeof(address));
   1.389 +
   1.390 +    if(loc_beg[0] == '|'){
   1.391 +      parse_error = g_strdup("no pipe allowed for RFC 822/821 address");
   1.392 +      return NULL;
   1.393 +    }
   1.394 +
   1.395 +    while(*p && (*p != ',')) p++;
   1.396 +    addr->address = g_strndup(string, p - string);
   1.397 +
   1.398 +    addr->local_part = g_strndup(loc_beg, loc_end - loc_beg);
   1.399 +
   1.400 +#ifdef PARSE_TEST
   1.401 +    g_print("addr->local_part = %s\n", addr->local_part);
   1.402 +#endif
   1.403 +
   1.404 +    if(dom_beg != NULL){
   1.405 +      addr->domain = g_strndup(dom_beg, dom_end - dom_beg);
   1.406 +    }else{
   1.407 +      if(addr->local_part[0] == 0)
   1.408 +	addr->domain = g_strdup(""); /* 'NULL' address (failure notice),
   1.409 +		      "" makes sure it will not be qualified with a hostname */
   1.410 +      else
   1.411 +	addr->domain = NULL;
   1.412 +    }
   1.413 +
   1.414 +    if(end != NULL)
   1.415 +      *end = p;
   1.416 +
   1.417 +#ifndef PARSE_TEST
   1.418 +    addr_unmark_delivered(addr);
   1.419 +#endif
   1.420 +
   1.421 +    return addr;
   1.422 +  }
   1.423 +  return NULL;
   1.424 +}
   1.425 +
   1.426 +address *create_address_rfc822(gchar *string, gchar **end){
   1.427 +  return _create_address(string, end, FALSE);
   1.428 +}
   1.429 +
   1.430 +address *create_address_rfc821(gchar *string, gchar **end){
   1.431 +  return _create_address(string, end, TRUE);
   1.432 +}
   1.433 +
   1.434 +GList *addr_list_append_rfc822(GList *addr_list, gchar *string, gchar *domain)
   1.435 +{
   1.436 +  gchar *p = string;
   1.437 +  gchar *end;
   1.438 +
   1.439 +  while(*p){
   1.440 +    address *addr = _create_address(p, &end, FALSE);
   1.441 +    if(addr){
   1.442 +      if(domain)
   1.443 +	if(addr->domain == NULL)
   1.444 +	  addr->domain = g_strdup(domain);
   1.445 +
   1.446 +      addr_list = g_list_append(addr_list, addr);
   1.447 +      p = end;
   1.448 +    }else
   1.449 +      break;
   1.450 +    while(*p == ',' || isspace(*p)) p++;
   1.451 +  }
   1.452 +  return addr_list;
   1.453 +}