stream/url.c

   1 /*
   2  * URL Helper
   3  *
   4  * Copyright (C) 2001 Bertrand Baudet <bertrand_baudet@yahoo.com>
   5  *
   6  * This file is part of MPlayer.
   7  *
   8  * MPlayer is free software; you can redistribute it and/or modify
   9  * it under the terms of the GNU General Public License as published by
  10  * the Free Software Foundation; either version 2 of the License, or
  11  * (at your option) any later version.
  12  *
  13  * MPlayer is distributed in the hope that it will be useful,
  14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16  * GNU General Public License for more details.
  17  *
  18  * You should have received a copy of the GNU General Public License along
  19  * with MPlayer; if not, write to the Free Software Foundation, Inc.,
  20  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  21  */
  22
  23 #include <string.h>
  24 #include <stdlib.h>
  25 #include <stdio.h>
  26 #include <ctype.h>
  27 #include <inttypes.h>
  28
  29 #include "url.h"
  30 #include "mp_msg.h"
  31
  32 #ifndef SIZE_MAX
  33 #define SIZE_MAX ((size_t)-1)
  34 #endif
  35
  36 URL_t *url_redirect(URL_t **url, const char *redir) {
  37   URL_t *u = *url;
  38   URL_t *res;
  39   if (!strchr(redir, '/') || *redir == '/') {
  40     char *tmp;
  41     char *newurl = malloc(strlen(u->url) + strlen(redir) + 1);
  42     strcpy(newurl, u->url);
  43     if (*redir == '/') {
  44       redir++;
  45       tmp = strstr(newurl, "://");
  46       if (tmp) tmp = strchr(tmp + 3, '/');
  47     } else
  48       tmp = strrchr(newurl, '/');
  49     if (tmp) tmp[1] = 0;
  50     strcat(newurl, redir);
  51     res = url_new(newurl);
  52     free(newurl);
  53   } else
  54     res = url_new(redir);
  55   url_free(u);
  56   *url = res;
  57   return res;
  58 }
  59
  60 URL_t*
  61 url_new(const char* url) {
  62         int pos1, pos2,v6addr = 0;
  63         URL_t* Curl = NULL;
  64         char *escfilename=NULL;
  65         char *ptr1=NULL, *ptr2=NULL, *ptr3=NULL, *ptr4=NULL;
  66         int jumpSize = 3;
  67
  68         if( url==NULL ) return NULL;
  69
  70         if (strlen(url) > (SIZE_MAX / 3 - 1)) {
  71                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
  72                 goto err_out;
  73         }
  74         escfilename=malloc(strlen(url)*3+1);
  75         if (!escfilename ) {
  76                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
  77                 goto err_out;
  78         }
  79
  80         // Create the URL container
  81         Curl = malloc(sizeof(URL_t));
  82         if( Curl==NULL ) {
  83                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
  84                 goto err_out;
  85         }
  86
  87         // Initialisation of the URL container members
  88         memset( Curl, 0, sizeof(URL_t) );
  89
  90         url_escape_string(escfilename,url);
  91
  92         // Copy the url in the URL container
  93         Curl->url = strdup(escfilename);
  94         if( Curl->url==NULL ) {
  95                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
  96                 goto err_out;
  97         }
  98         mp_msg(MSGT_OPEN,MSGL_V,"Filename for url is now %s\n",escfilename);
  99
 100         // extract the protocol
 101         ptr1 = strstr(escfilename, "://");
 102         if( ptr1==NULL ) {
 103                 // Check for a special case: "sip:" (without "//"):
 104                 if (strstr(escfilename, "sip:") == escfilename) {
 105                         ptr1 = (char *)&url[3]; // points to ':'
 106                         jumpSize = 1;
 107                 } else {
 108                         mp_msg(MSGT_NETWORK,MSGL_V,"Not an URL!\n");
 109                         goto err_out;
 110                 }
 111         }
 112         pos1 = ptr1-escfilename;
 113         Curl->protocol = malloc(pos1+1);
 114         if( Curl->protocol==NULL ) {
 115                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 116                 goto err_out;
 117         }
 118         strncpy(Curl->protocol, escfilename, pos1);
 119         Curl->protocol[pos1] = '\0';
 120
 121         // jump the "://"
 122         ptr1 += jumpSize;
 123         pos1 += jumpSize;
 124
 125         // check if a username:password is given
 126         ptr2 = strstr(ptr1, "@");
 127         ptr3 = strstr(ptr1, "/");
 128         if( ptr3!=NULL && ptr3<ptr2 ) {
 129                 // it isn't really a username but rather a part of the path
 130                 ptr2 = NULL;
 131         }
 132         if( ptr2!=NULL ) {
 133                 // We got something, at least a username...
 134                 int len = ptr2-ptr1;
 135                 Curl->username = malloc(len+1);
 136                 if( Curl->username==NULL ) {
 137                         mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 138                         goto err_out;
 139                 }
 140                 strncpy(Curl->username, ptr1, len);
 141                 Curl->username[len] = '\0';
 142
 143                 ptr3 = strstr(ptr1, ":");
 144                 if( ptr3!=NULL && ptr3<ptr2 ) {
 145                         // We also have a password
 146                         int len2 = ptr2-ptr3-1;
 147                         Curl->username[ptr3-ptr1]='\0';
 148                         Curl->password = malloc(len2+1);
 149                         if( Curl->password==NULL ) {
 150                                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 151                                 goto err_out;
 152                         }
 153                         strncpy( Curl->password, ptr3+1, len2);
 154                         Curl->password[len2]='\0';
 155                         url_unescape_string(Curl->password, Curl->password);
 156                 }
 157                 url_unescape_string(Curl->username, Curl->username);
 158                 ptr1 = ptr2+1;
 159                 pos1 = ptr1-escfilename;
 160         }
 161
 162         // before looking for a port number check if we have an IPv6 type numeric address
 163         // in IPv6 URL the numeric address should be inside square braces.
 164         ptr2 = strstr(ptr1, "[");
 165         ptr3 = strstr(ptr1, "]");
 166         ptr4 = strstr(ptr1, "/");
 167         if( ptr2!=NULL && ptr3!=NULL && ptr2 < ptr3 && (!ptr4 || ptr4 > ptr3)) {
 168                 // we have an IPv6 numeric address
 169                 ptr1++;
 170                 pos1++;
 171                 ptr2 = ptr3;
 172                 v6addr = 1;
 173         } else {
 174                 ptr2 = ptr1;
 175
 176         }
 177
 178         // look if the port is given
 179         ptr2 = strstr(ptr2, ":");
 180         // If the : is after the first / it isn't the port
 181         ptr3 = strstr(ptr1, "/");
 182         if(ptr3 && ptr3 - ptr2 < 0) ptr2 = NULL;
 183         if( ptr2==NULL ) {
 184                 // No port is given
 185                 // Look if a path is given
 186                 if( ptr3==NULL ) {
 187                         // No path/filename
 188                         // So we have an URL like http://www.hostname.com
 189                         pos2 = strlen(escfilename);
 190                 } else {
 191                         // We have an URL like http://www.hostname.com/file.txt
 192                         pos2 = ptr3-escfilename;
 193                 }
 194         } else {
 195                 // We have an URL beginning like http://www.hostname.com:1212
 196                 // Get the port number
 197                 Curl->port = atoi(ptr2+1);
 198                 pos2 = ptr2-escfilename;
 199         }
 200         if( v6addr ) pos2--;
 201         // copy the hostname in the URL container
 202         Curl->hostname = malloc(pos2-pos1+1);
 203         if( Curl->hostname==NULL ) {
 204                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 205                 goto err_out;
 206         }
 207         strncpy(Curl->hostname, ptr1, pos2-pos1);
 208         Curl->hostname[pos2-pos1] = '\0';
 209
 210         // Look if a path is given
 211         ptr2 = strstr(ptr1, "/");
 212         if( ptr2!=NULL ) {
 213                 // A path/filename is given
 214                 // check if it's not a trailing '/'
 215                 if( strlen(ptr2)>1 ) {
 216                         // copy the path/filename in the URL container
 217                         Curl->file = strdup(ptr2);
 218                         if( Curl->file==NULL ) {
 219                                 mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 220                                 goto err_out;
 221                         }
 222                 }
 223         }
 224         // Check if a filename was given or set, else set it with '/'
 225         if( Curl->file==NULL ) {
 226                 Curl->file = malloc(2);
 227                 if( Curl->file==NULL ) {
 228                         mp_tmsg(MSGT_NETWORK,MSGL_FATAL,"Memory allocation failed.\n");
 229                         goto err_out;
 230                 }
 231                 strcpy(Curl->file, "/");
 232         }
 233
 234         free(escfilename);
 235         return Curl;
 236 err_out:
 237         free(escfilename);
 238         if (Curl) url_free(Curl);
 239         return NULL;
 240 }
 241
 242 void
 243 url_free(URL_t* url) {
 244         if(!url) return;
 245         free(url->url);
 246         free(url->protocol);
 247         free(url->hostname);
 248         free(url->file);
 249         free(url->username);
 250         free(url->password);
 251         free(url);
 252 }
 253
 254
 255 /* Replace escape sequences in an URL (or a part of an URL) */
 256 /* works like strcpy(), but without return argument,
 257    except that outbuf == inbuf is allowed */
 258 void
 259 url_unescape_string(char *outbuf, const char *inbuf)
 260 {
 261         unsigned char c,c1,c2;
 262         int i,len=strlen(inbuf);
 263         for (i=0;i<len;i++){
 264                 c = inbuf[i];
 265                 if (c == '%' && i<len-2) { //must have 2 more chars
 266                         c1 = toupper(inbuf[i+1]); // we need uppercase characters
 267                         c2 = toupper(inbuf[i+2]);
 268                         if (    ((c1>='0' && c1<='9') || (c1>='A' && c1<='F')) &&
 269                                 ((c2>='0' && c2<='9') || (c2>='A' && c2<='F')) ) {
 270                                 if (c1>='0' && c1<='9') c1-='0';
 271                                 else c1-='A'-10;
 272                                 if (c2>='0' && c2<='9') c2-='0';
 273                                 else c2-='A'-10;
 274                                 c = (c1<<4) + c2;
 275                                 i=i+2; //only skip next 2 chars if valid esc
 276                         }
 277                 }
 278                 *outbuf++ = c;
 279         }
 280         *outbuf++='\0'; //add nullterm to string
 281 }
 282
 283 static void
 284 url_escape_string_part(char *outbuf, const char *inbuf) {
 285         unsigned char c,c1,c2;
 286         int i,len=strlen(inbuf);
 287
 288         for  (i=0;i<len;i++) {
 289                 c = inbuf[i];
 290                 if ((c=='%') && i<len-2 ) { //need 2 more characters
 291                     c1=toupper(inbuf[i+1]); c2=toupper(inbuf[i+2]); // need uppercase chars
 292                    } else {
 293                     c1=129; c2=129; //not escape chars
 294                    }
 295
 296                 if(     (c >= 'A' && c <= 'Z') ||
 297                         (c >= 'a' && c <= 'z') ||
 298                         (c >= '0' && c <= '9') ||
 299                         (c >= 0x7f)) {
 300                         *outbuf++ = c;
 301                 } else if ( c=='%' && ((c1 >= '0' && c1 <= '9') || (c1 >= 'A' && c1 <= 'F')) &&
 302                            ((c2 >= '0' && c2 <= '9') || (c2 >= 'A' && c2 <= 'F'))) {
 303                                                               // check if part of an escape sequence
 304                             *outbuf++=c;                      // already
 305
 306                                                               // dont escape again
 307                             mp_tmsg(MSGT_NETWORK,MSGL_ERR,"String appears to be already escaped in url_escape %c%c1%c2\n",c,c1,c2);
 308                                                               // error as this should not happen against RFC 2396
 309                                                               // to escape a string twice
 310                 } else {
 311                         /* all others will be escaped */
 312                         c1 = ((c & 0xf0) >> 4);
 313                         c2 = (c & 0x0f);
 314                         if (c1 < 10) c1+='0';
 315                         else c1+='A'-10;
 316                         if (c2 < 10) c2+='0';
 317                         else c2+='A'-10;
 318                         *outbuf++ = '%';
 319                         *outbuf++ = c1;
 320                         *outbuf++ = c2;
 321                 }
 322         }
 323         *outbuf++='\0';
 324 }
 325
 326 /* Replace specific characters in the URL string by an escape sequence */
 327 /* works like strcpy(), but without return argument */
 328 void
 329 url_escape_string(char *outbuf, const char *inbuf) {
 330         unsigned char c;
 331         int i = 0,j,len = strlen(inbuf);
 332         char* tmp,*unesc = NULL, *in;
 333
 334         // Look if we have an ip6 address, if so skip it there is
 335         // no need to escape anything in there.
 336         tmp = strstr(inbuf,"://[");
 337         if(tmp) {
 338                 tmp = strchr(tmp+4,']');
 339                 if(tmp && (tmp[1] == '/' || tmp[1] == ':' ||
 340                            tmp[1] == '\0')) {
 341                         i = tmp+1-inbuf;
 342                         strncpy(outbuf,inbuf,i);
 343                         outbuf += i;
 344                         tmp = NULL;
 345                 }
 346         }
 347
 348         tmp = NULL;
 349         while(i < len) {
 350                 // look for the next char that must be kept
 351                 for  (j=i;j<len;j++) {
 352                         c = inbuf[j];
 353                         if(c=='-' || c=='_' || c=='.' || c=='!' || c=='~' ||    /* mark characters */
 354                            c=='*' || c=='\'' || c=='(' || c==')' ||             /* do not touch escape character */
 355                            c==';' || c=='/' || c=='?' || c==':' || c=='@' ||    /* reserved characters */
 356                            c=='&' || c=='=' || c=='+' || c=='$' || c==',')      /* see RFC 2396 */
 357                                 break;
 358                 }
 359                 // we are on a reserved char, write it out
 360                 if(j == i) {
 361                         *outbuf++ = c;
 362                         i++;
 363                         continue;
 364                 }
 365                 // we found one, take that part of the string
 366                 if(j < len) {
 367                         if(!tmp) tmp = malloc(len+1);
 368                         strncpy(tmp,inbuf+i,j-i);
 369                         tmp[j-i] = '\0';
 370                         in = tmp;
 371                 } else // take the rest of the string
 372                         in = (char*)inbuf+i;
 373
 374                 if(!unesc) unesc = malloc(len+1);
 375                 // unescape first to avoid escaping escape
 376                 url_unescape_string(unesc,in);
 377                 // then escape, including mark and other reserved chars
 378                 // that can come from escape sequences
 379                 url_escape_string_part(outbuf,unesc);
 380                 outbuf += strlen(outbuf);
 381                 i += strlen(in);
 382         }
 383         *outbuf = '\0';
 384         free(tmp);
 385         free(unesc);
 386 }
 387
 388 #ifdef URL_DEBUG
 389 void
 390 url_debug(const URL_t *url) {
 391         if( url==NULL ) {
 392                 printf("URL pointer NULL\n");
 393                 return;
 394         }
 395         if( url->url!=NULL ) {
 396                 printf("url=%s\n", url->url );
 397         }
 398         if( url->protocol!=NULL ) {
 399                 printf("protocol=%s\n", url->protocol );
 400         }
 401         if( url->hostname!=NULL ) {
 402                 printf("hostname=%s\n", url->hostname );
 403         }
 404         printf("port=%d\n", url->port );
 405         if( url->file!=NULL ) {
 406                 printf("file=%s\n", url->file );
 407         }
 408         if( url->username!=NULL ) {
 409                 printf("username=%s\n", url->username );
 410         }
 411         if( url->password!=NULL ) {
 412                 printf("password=%s\n", url->password );
 413         }
 414 }
 415 #endif /* URL_DEBUG */