transmission/libtransmission/metainfo.c

695 lines
20 KiB
C

/******************************************************************************
* $Id$
*
* Copyright (c) 2005-2008 Transmission authors and contributors
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
*****************************************************************************/
#include <assert.h>
#include <ctype.h> /* isspace */
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <unistd.h> /* unlink, stat */
#include "transmission.h"
#include "bencode.h"
#include "crypto.h" /* tr_sha1 */
#include "metainfo.h"
#include "platform.h"
#include "trcompat.h" /* strlcpy */
#include "utils.h"
/***********************************************************************
* Local prototypes
**********************************************************************/
static int getannounce( tr_info * inf, tr_benc * meta );
static char * announceToScrape( const char * announce );
static int parseFiles( tr_info * inf, tr_benc * name,
tr_benc * files, tr_benc * length );
/***
****
***/
#define WANTBYTES( want, got ) \
if( (want) > (got) ) { return; } else { (got) -= (want); }
static void
strlcat_utf8( void * dest, const void * src, size_t len, char skip )
{
char * s = dest;
const char * append = src;
const char * p;
/* don't overwrite the nul at the end */
len--;
/* Go to the end of the destination string */
while( s[0] )
{
s++;
len--;
}
/* Now start appending, converting on the fly if necessary */
for( p = append; p[0]; )
{
/* skip over the requested character */
if( skip == p[0] )
{
p++;
continue;
}
if( !( p[0] & 0x80 ) )
{
/* ASCII character */
WANTBYTES( 1, len );
*(s++) = *(p++);
continue;
}
if( ( p[0] & 0xE0 ) == 0xC0 && ( p[1] & 0xC0 ) == 0x80 )
{
/* 2-bytes UTF-8 character */
WANTBYTES( 2, len );
*(s++) = *(p++); *(s++) = *(p++);
continue;
}
if( ( p[0] & 0xF0 ) == 0xE0 && ( p[1] & 0xC0 ) == 0x80 &&
( p[2] & 0xC0 ) == 0x80 )
{
/* 3-bytes UTF-8 character */
WANTBYTES( 3, len );
*(s++) = *(p++); *(s++) = *(p++);
*(s++) = *(p++);
continue;
}
if( ( p[0] & 0xF8 ) == 0xF0 && ( p[1] & 0xC0 ) == 0x80 &&
( p[2] & 0xC0 ) == 0x80 && ( p[3] & 0xC0 ) == 0x80 )
{
/* 4-bytes UTF-8 character */
WANTBYTES( 4, len );
*(s++) = *(p++); *(s++) = *(p++);
*(s++) = *(p++); *(s++) = *(p++);
continue;
}
/* ISO 8859-1 -> UTF-8 conversion */
WANTBYTES( 2, len );
*(s++) = 0xC0 | ( ( *p & 0xFF ) >> 6 );
*(s++) = 0x80 | ( *(p++) & 0x3F );
}
}
static void
savedname( const tr_handle * handle,
char * name,
size_t len,
const char * hash )
{
const char * torDir = tr_getTorrentDir( handle );
if( !handle->tag )
{
tr_buildPath( name, len, torDir, hash, NULL );
}
else
{
char base[1024];
snprintf( base, sizeof(base), "%s-%s", hash, handle->tag );
tr_buildPath( name, len, torDir, base, NULL );
}
}
int
tr_metainfoParse( const tr_handle * handle,
tr_info * inf,
const tr_benc * meta_in )
{
tr_piece_index_t i;
tr_benc * beInfo, * val, * val2;
tr_benc * meta = (tr_benc *) meta_in;
char buf[4096];
/* info_hash: urlencoded 20-byte SHA1 hash of the value of the info key
* from the Metainfo file. Note that the value will be a bencoded
* dictionary, given the definition of the info key above. */
if(( beInfo = tr_bencDictFindType( meta, "info", TYPE_DICT )))
{
int len;
char * str = tr_bencSave( beInfo, &len );
tr_sha1( inf->hash, str, len, NULL );
tr_free( str );
}
else
{
tr_err( _( "Missing metadata entry \"%s\"" ), "info" );
return TR_EINVALID;
}
tr_sha1_to_hex( inf->hashString, inf->hash );
savedname( handle, buf, sizeof( buf ), inf->hashString );
tr_free( inf->torrent );
inf->torrent = tr_strdup( buf );
/* comment */
memset( buf, '\0', sizeof( buf ) );
val = tr_bencDictFindFirst( meta, "comment.utf-8", "comment", NULL );
if( tr_bencIsString( val ) )
strlcat_utf8( buf, val->val.s.s, sizeof( buf ), 0 );
tr_free( inf->comment );
inf->comment = tr_strdup( buf );
/* creator */
memset( buf, '\0', sizeof( buf ) );
val = tr_bencDictFindFirst( meta, "created by.utf-8", "created by", NULL );
if( tr_bencIsString( val ) )
strlcat_utf8( buf, val->val.s.s, sizeof( buf ), 0 );
tr_free( inf->creator );
inf->creator = tr_strdup( buf );
/* Date created */
inf->dateCreated = 0;
val = tr_bencDictFind( meta, "creation date" );
if( tr_bencIsInt( val ) )
inf->dateCreated = val->val.i;
/* Private torrent */
val = tr_bencDictFind( beInfo, "private" );
val2 = tr_bencDictFind( meta, "private" );
if( ( tr_bencIsInt(val) && val->val.i ) ||
( tr_bencIsInt(val2) && val2->val.i ) )
{
inf->isPrivate = 1;
}
/* Piece length */
val = tr_bencDictFind( beInfo, "piece length" );
if( !tr_bencIsInt( val ) )
{
if( val )
tr_err( _( "Invalid metadata entry \"%s\"" ), "piece length" );
else
tr_err( _( "Missing metadata entry \"%s\"" ), "piece length" );
goto fail;
}
inf->pieceSize = val->val.i;
/* Hashes */
val = tr_bencDictFind( beInfo, "pieces" );
if( !tr_bencIsString( val ) )
{
if( val )
tr_err( _( "Invalid metadata entry \"%s\"" ), "pieces" );
else
tr_err( _( "Missing metadata entry \"%s\"" ), "pieces" );
goto fail;
}
if( val->val.s.i % SHA_DIGEST_LENGTH )
{
tr_err( _( "Invalid metadata entry \"%s\"" ), "pieces" );
goto fail;
}
inf->pieceCount = val->val.s.i / SHA_DIGEST_LENGTH;
inf->pieces = calloc ( inf->pieceCount, sizeof(tr_piece) );
for ( i=0; i<inf->pieceCount; ++i )
{
memcpy (inf->pieces[i].hash, &val->val.s.s[i*SHA_DIGEST_LENGTH], SHA_DIGEST_LENGTH);
}
/* get file or top directory name */
val = tr_bencDictFindFirst( beInfo, "name.utf-8", "name", NULL );
if( parseFiles( inf, tr_bencDictFindFirst( beInfo,
"name.utf-8", "name", NULL ),
tr_bencDictFind( beInfo, "files" ),
tr_bencDictFind( beInfo, "length" ) ) )
{
goto fail;
}
if( !inf->fileCount || !inf->totalSize )
{
tr_err( _( "Torrent is corrupt" ) ); /* the content is missing! */
goto fail;
}
/* TODO add more tests so we don't crash on weird files */
if( (uint64_t) inf->pieceCount !=
( inf->totalSize + inf->pieceSize - 1 ) / inf->pieceSize )
{
tr_err( _( "Torrent is corrupt" ) ); /* size of hashes and files don't match */
goto fail;
}
/* get announce or announce-list */
if( getannounce( inf, meta ) )
{
goto fail;
}
return TR_OK;
fail:
tr_metainfoFree( inf );
return TR_EINVALID;
}
void tr_metainfoFree( tr_info * inf )
{
tr_file_index_t ff;
int i, j;
for( ff=0; ff<inf->fileCount; ++ff )
tr_free( inf->files[ff].name );
tr_free( inf->pieces );
tr_free( inf->files );
tr_free( inf->comment );
tr_free( inf->creator );
tr_free( inf->torrent );
tr_free( inf->name );
tr_free( inf->primaryAddress );
for( i=0; i<inf->trackerTiers; ++i ) {
for( j=0; j<inf->trackerList[i].count; ++j )
tr_trackerInfoClear( &inf->trackerList[i].list[j] );
tr_free( inf->trackerList[i].list );
}
tr_free( inf->trackerList );
memset( inf, '\0', sizeof(tr_info) );
}
static int
getfile( char ** setme, const char * prefix, tr_benc * name )
{
const char ** list;
int ii, jj;
char buf[4096];
if( !tr_bencIsList( name ) )
return TR_EINVALID;
list = calloc( name->val.l.count, sizeof( list[0] ) );
if( !list )
return TR_EINVALID;
for( ii = jj = 0; name->val.l.count > ii; ii++ )
{
tr_benc * dir = &name->val.l.vals[ii];
if( !tr_bencIsString( dir ) )
continue;
if( 0 == strcmp( "..", dir->val.s.s ) )
{
if( 0 < jj )
{
jj--;
}
}
else if( 0 != strcmp( ".", dir->val.s.s ) )
{
list[jj] = dir->val.s.s;
jj++;
}
}
if( 0 == jj )
{
free( list );
return TR_EINVALID;
}
memset( buf, 0, sizeof( buf ) );
strlcat_utf8( buf, prefix, sizeof(buf), 0 );
for( ii = 0; jj > ii; ii++ )
{
strlcat_utf8( buf, TR_PATH_DELIMITER_STR, sizeof(buf), 0 );
strlcat_utf8( buf, list[ii], sizeof(buf), TR_PATH_DELIMITER );
}
free( list );
tr_free( *setme );
*setme = tr_strdup( buf );
return TR_OK;
}
static int getannounce( tr_info * inf, tr_benc * meta )
{
tr_benc * val, * urlval;
char * address, * announce;
int ii, jj, port, random;
tr_tracker_info * sublist;
void * swapping;
/* Announce-list */
val = tr_bencDictFind( meta, "announce-list" );
if( tr_bencIsList(val) && 0 < val->val.l.count )
{
inf->trackerTiers = 0;
inf->trackerList = calloc( val->val.l.count,
sizeof( inf->trackerList[0] ) );
/* iterate through the announce-list's tiers */
for( ii = 0; ii < val->val.l.count; ii++ )
{
int subcount = 0;
tr_benc * subval = &val->val.l.vals[ii];
if( !tr_bencIsList(subval) || 0 >= subval->val.l.count )
continue;
sublist = calloc( subval->val.l.count, sizeof( sublist[0] ) );
/* iterate through the tier's items */
for( jj = 0; jj < subval->val.l.count; jj++ )
{
tr_tracker_info tmp;
urlval = &subval->val.l.vals[jj];
if( TYPE_STR != urlval->type ||
tr_trackerInfoInit( &tmp, urlval->val.s.s, urlval->val.s.i ) )
{
continue;
}
if( !inf->primaryAddress ) {
char buf[1024];
snprintf( buf, sizeof(buf), "%s:%d", tmp.address, tmp.port );
inf->primaryAddress = tr_strdup( buf );
}
/* place the item info in a random location in the sublist */
random = tr_rand( subcount + 1 );
if( random != subcount )
sublist[subcount] = sublist[random];
sublist[random] = tmp;
subcount++;
}
/* just use sublist as-is if it's full */
if( subcount == subval->val.l.count )
{
inf->trackerList[inf->trackerTiers].list = sublist;
inf->trackerList[inf->trackerTiers].count = subcount;
inf->trackerTiers++;
}
/* if we skipped some of the tier's items then trim the sublist */
else if( 0 < subcount )
{
inf->trackerList[inf->trackerTiers].list = calloc( subcount, sizeof( sublist[0] ) );
memcpy( inf->trackerList[inf->trackerTiers].list, sublist,
sizeof( sublist[0] ) * subcount );
inf->trackerList[inf->trackerTiers].count = subcount;
inf->trackerTiers++;
free( sublist );
}
/* drop the whole sublist if we didn't use any items at all */
else
{
free( sublist );
}
}
/* did we use any of the tiers? */
if( 0 == inf->trackerTiers )
{
tr_inf( _( "Invalid metadata entry \"%s\"" ), "announce-list" );
free( inf->trackerList );
inf->trackerList = NULL;
}
/* trim unused sublist pointers */
else if( inf->trackerTiers < val->val.l.count )
{
swapping = inf->trackerList;
inf->trackerList = calloc( inf->trackerTiers,
sizeof( inf->trackerList[0] ) );
memcpy( inf->trackerList, swapping,
sizeof( inf->trackerList[0] ) * inf->trackerTiers );
free( swapping );
}
}
/* Regular announce value */
val = tr_bencDictFind( meta, "announce" );
if( !tr_bencIsString( val ) )
{
tr_err( _( "Missing metadata entry \"%s\"" ), "announce" );
return TR_EINVALID;
}
if( !inf->trackerTiers )
{
char buf[4096], *pch;
strlcpy( buf, val->val.s.s, sizeof( buf ) );
pch = buf;
while( isspace( *pch ) )
++pch;
if( tr_httpParseURL( pch, -1, &address, &port, &announce ) )
{
tr_err( _( "Invalid announce URL \"%s\"" ), val->val.s.s );
return TR_EINVALID;
}
sublist = calloc( 1, sizeof( sublist[0] ) );
sublist[0].address = address;
sublist[0].port = port;
sublist[0].announce = announce;
sublist[0].scrape = announceToScrape( announce );
inf->trackerList = calloc( 1, sizeof( inf->trackerList[0] ) );
inf->trackerList[0].list = sublist;
inf->trackerList[0].count = 1;
inf->trackerTiers = 1;
if( !inf->primaryAddress ) {
char buf[1024];
snprintf( buf, sizeof(buf), "%s:%d", sublist[0].address, sublist[0].port );
inf->primaryAddress = tr_strdup( buf );
}
}
return TR_OK;
}
static char * announceToScrape( const char * announce )
{
char old[] = "announce";
int oldlen = 8;
char new[] = "scrape";
int newlen = 6;
char * slash, * scrape;
size_t scrapelen, used;
slash = strrchr( announce, '/' );
if( NULL == slash )
{
return NULL;
}
slash++;
if( 0 != strncmp( slash, old, oldlen ) )
{
return NULL;
}
scrapelen = strlen( announce ) - oldlen + newlen;
scrape = calloc( scrapelen + 1, 1 );
if( NULL == scrape )
{
return NULL;
}
assert( ( size_t )( slash - announce ) < scrapelen );
memcpy( scrape, announce, slash - announce );
used = slash - announce;
strncat( scrape, new, scrapelen - used );
used += newlen;
assert( strlen( scrape ) == used );
if( used < scrapelen )
{
assert( strlen( slash + oldlen ) == scrapelen - used );
strncat( scrape, slash + oldlen, scrapelen - used );
}
return scrape;
}
int
tr_trackerInfoInit( tr_tracker_info * info,
const char * address,
int address_len )
{
int ret = tr_httpParseURL( address, address_len,
&info->address,
&info->port,
&info->announce );
if( !ret )
info->scrape = announceToScrape( info->announce );
return ret;
}
void
tr_trackerInfoClear( tr_tracker_info * info )
{
tr_free( info->address );
tr_free( info->announce );
tr_free( info->scrape );
memset( info, '\0', sizeof(tr_tracker_info) );
}
void
tr_metainfoRemoveSaved( const tr_handle * handle,
const char * hashString )
{
char file[MAX_PATH_LENGTH];
savedname( handle, file, sizeof file, hashString );
unlink( file );
}
/* Save a copy of the torrent file in the saved torrent directory */
int
tr_metainfoSave( const tr_handle * handle,
const char * hash,
const uint8_t * buf,
size_t buflen )
{
char path[MAX_PATH_LENGTH];
FILE * file;
savedname( handle, path, sizeof path, hash );
file = fopen( path, "wb+" );
if( !file )
{
tr_err( _( "Couldn't open \"%1$s\": %2$s" ), path, tr_strerror( errno ) );
return TR_EINVALID;
}
fseek( file, 0, SEEK_SET );
if( fwrite( buf, 1, buflen, file ) != buflen )
{
tr_err( _( "Couldn't save file \"%1$s\": %2$s" ), path, tr_strerror( errno ) );
fclose( file );
return TR_EINVALID;
}
fclose( file );
return TR_OK;
}
static int
parseFiles( tr_info * inf, tr_benc * name,
tr_benc * files, tr_benc * length )
{
tr_benc * item, * path;
int ii;
if( !tr_bencIsString( name ) )
{
if( name )
tr_err( _( "Invalid metadata entry \"%s\"" ), "name" );
else
tr_err( _( "Missing metadata entry \"%s\"" ), "name" );
return TR_EINVALID;
}
tr_free( inf->name );
inf->name = tr_strdup( name->val.s.s );
if( !inf->name || !*inf->name )
{
tr_err( _( "Invalid metadata entry \"%s\"" ), "name" );
return TR_EINVALID;
}
inf->totalSize = 0;
if( tr_bencIsList( files ) )
{
/* Multi-file mode */
inf->isMultifile = 1;
inf->fileCount = files->val.l.count;
inf->files = calloc( inf->fileCount, sizeof( inf->files[0] ) );
if( !inf->files )
return TR_EINVALID;
for( ii = 0; files->val.l.count > ii; ii++ )
{
item = &files->val.l.vals[ii];
path = tr_bencDictFindFirst( item, "path.utf-8", "path", NULL );
if( getfile( &inf->files[ii].name, inf->name, path ) )
{
if( path )
tr_err( _( "Invalid metadata entry \"%s\"" ), "path" );
else
tr_err( _( "Missing metadata entry \"%s\"" ), "path" );
return TR_EINVALID;
}
length = tr_bencDictFind( item, "length" );
if( !tr_bencIsInt( length ) )
{
if( length )
tr_err( _( "Invalid metadata entry \"%s\"" ), "length" );
else
tr_err( _( "Missing metadata entry \"%s\"" ), "length" );
return TR_EINVALID;
}
inf->files[ii].length = length->val.i;
inf->totalSize += length->val.i;
}
}
else if( tr_bencIsInt( length ) )
{
char buf[4096];
/* Single-file mode */
inf->isMultifile = 0;
inf->fileCount = 1;
inf->files = calloc( 1, sizeof( inf->files[0] ) );
if( !inf->files )
return TR_EINVALID;
memset( buf, 0, sizeof( buf ) );
strlcat_utf8( buf, name->val.s.s, sizeof(buf), TR_PATH_DELIMITER );
tr_free( inf->files[0].name );
inf->files[0].name = tr_strdup( buf );
inf->files[0].length = length->val.i;
inf->totalSize += length->val.i;
}
else
{
tr_err( _( "Invalid or missing metadata entries \"length\" and \"files\"" ) );
}
return TR_OK;
}