usbloadergx/source/utils/U8Archive.cpp

/****************************************************************************
 * Copyright (C) 2012 Dimok and giantpune
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 ****************************************************************************/
#include <malloc.h>
#include <stdlib.h>
#include <string.h>
#include <stdio.h>

#include "ash.h"
#include "lz77.h"
#include "gecko.h"
#include "U8Archive.h"
#include "uncompress.h"

#define RU(N, S) ((((N) + (S) - 1) / (S)) * (S))

U8Archive::U8Archive(const u8 *stuff, u32 len )
	: fst( NULL ),
	  name_table( NULL ),
	  data( NULL )
{
	if( stuff )
	{
		SetData( stuff, len );
	}
}

void U8Archive::SetData( const u8 *stuff, u32 len )
{
	fst = NULL;
	name_table = NULL;
	data = NULL;
	if( !stuff || len < 0x40 )
	{
		gprintf( "SetData(): !stuff || len < 0x40  %p %08x\n", stuff, len );
		return;
	}
	stuff = FindU8Tag( stuff, len );
	if( !stuff )
	{
		gprintf( "U8 tag not found\n" );
		return;
	}
	const U8Header * binHdr = (const U8Header *)stuff;

	const u8* fst_buffer = stuff + binHdr->rootNodeOffset;
	fst = (FstEntry *)fst_buffer;
	u32 name_table_offset = fst->filelen * 0xC;
	name_table = (char *)( fst_buffer + name_table_offset );
	data = (u8*)stuff;
}

const u8 *U8Archive::FindU8Tag( const u8* stuff, u32 len )
{
	if( !stuff || len < 0x20 )
	{
		return NULL;
	}
	if( *(u32*)( stuff ) == 0x55AA382D )
	{
		return stuff;
	}
	if( len > 0x620 && *(u32*)( stuff + 0x600 ) == 0x55AA382D )
	{
		return stuff + 0x600;
	}
	if( len > 0x660 && *(u32*)( stuff + 0x640 ) == 0x55AA382D )
	{
		return stuff + 0x640;
	}
	return NULL;
}

u8 *U8Archive::GetFile( const std::string &path, u32 *size ) const
{
	return GetFile( path.c_str(), size );
}

u8 *U8Archive::GetFile( const char *path, u32 *size ) const
{
	if( !path || !fst || !name_table )
	{
		return NULL;
	}

	s32 entryNo = EntryFromPath( path );
	if( entryNo < 1 )
	{
		//gprintf( "U8: entry wasn\'t found in the archive  \"%s\"\n", path );
		return NULL;
	}
	if( fst[ entryNo ].filetype )
	{
		gprintf( "U8: \"%s\" is a folder\n", path );
		return NULL;
	}
	if( size )
	{
		*size = fst[ entryNo ].filelen;
	}
	return data + fst[ entryNo ].fileoffset;
}

u8 *U8Archive::GetFile( u32 fstIdx, u32 *size ) const
{
	if( !fst || !name_table )
	{
		return NULL;
	}

	if( fstIdx >= fst[0].filelen || fst[ fstIdx ].filetype )
	{
		gprintf( "%i is a folder\n", fstIdx );
		return NULL;
	}

	if( size )
	{
		*size = fst[ fstIdx ].filelen;
	}
	return data + fst[ fstIdx ].fileoffset;
}

u8* U8Archive::GetFileAllocated( const std::string &path, u32 *size ) const
{
	return GetFileAllocated( path.c_str(), size );
}

u8 *U8Archive::DecompressCopy( const u8 * stuff, u32 len, u32 *size ) const
{
	// check for IMD5 header and skip it
	if( len > 0x40 && *(u32*)stuff == 0x494d4435 )// IMD5
	{
		stuff += 0x20;
		len -= 0x20;
	}

	u8* ret = NULL;
	// determine if it needs to be decompressed
	if( IsAshCompressed( stuff, len ) )
	{
		//u32 len2 = len;
		// ASH0
		ret = DecompressAsh( stuff, len );
		if( !ret )
		{
			gprintf( "out of memory\n" );
			return NULL;
		}
	}
	else if( (len > 8) && *(u32*)( stuff ) == 0x59617a30 )// Yaz0
	{
		// Yaz0 with a magic word
		Yaz0_Header *header = (Yaz0_Header *) stuff;
		// set decompress length
		len = header->decompressed_size;
		// allocate memory
		ret = (u8*) memalign(32, ALIGN32(len));
		if(!ret)
		{
			gprintf("out of memory\n");
			return NULL;
		}
		// function can not fail at this point
		uncompressYaz0(stuff, ret, len);
	}
	else if( isLZ77compressed( stuff ) )
	{
		// LZ77 with no magic word
		if( decompressLZ77content( stuff, len, &ret, &len ) )
		{
			return NULL;
		}
	}
	else if( *(u32*)( stuff ) == 0x4C5A3737 )// LZ77
	{
		// LZ77 with a magic word
		if( decompressLZ77content( stuff + 4, len - 4, &ret, &len ) )
		{
			return NULL;
		}
	}
	else
	{
		// just copy the data out of the archive
		ret = (u8*)memalign( 32, ALIGN32( len ) );
		if( !ret )
		{
			gprintf( "out of memory\n" );
			return NULL;
		}
		memcpy( ret, stuff, len );
	}
	if( size )
	{
		*size = len;
	}

	// flush the cache so if there are any textures in this data, it will be ready for the GX
	DCFlushRange( ret, len );
	return ret;
}

u8* U8Archive::GetFileAllocated( const char *path, u32 *size ) const
{
	u32 len;
	const u8 *stuff = GetFile( path, &len );
	if( !stuff )
	{
		return NULL;
	}

	return DecompressCopy( stuff, len, size );
}

u8* U8Archive::GetFileAllocated( u32 fstIdx, u32 *size ) const
{
	u32 len;
	const u8 *stuff = GetFile( fstIdx, &len );
	if( !stuff )
	{
		return NULL;
	}

	return DecompressCopy( stuff, len, size );
}


u32 U8Archive::FileDescriptor( const char *path ) const
{
	int ret = EntryFromPath( path );
	if( ret < 1 )
	{
		return 0;
	}
	return ret;
}

u8* U8Archive::GetFileFromFd( u32 fd, u32 *size )const
{
	if( !fst || !name_table || fd >= fst[ 0 ].filelen )
	{
		return NULL;
	}
	if( fst[ fd ].filetype )
	{
		gprintf( "U8: \"%s\" is a folder\n", FstName( &fst[ fd ] ) );
		return NULL;
	}
	if( size )
	{
		*size = fst[ fd ].filelen;
	}
	return data + fst[ fd ].fileoffset;
}

char *U8Archive::FstName( const FstEntry *entry ) const
{
	if( entry == &fst[ 0 ] )
	{
		return NULL;
	}

	return (char*)name_table + ( *((u32 *)entry) & 0x00ffffff );
}

int U8Archive::strcasecmp_slash( const char *s1, const char *s2 )
{
	while( toupper( *s1 ) == toupper( *s2++ ) )
	{
		if( *s1++ == 0 || *s1 == '/' )
		{
			return 0;
		}
	}

	return( *(const unsigned char*)s1 - *(const unsigned char *)( s2 - 1 ) );
}

int U8Archive::strlen_slash( const char *s )
{
	int ret = 0;
	while( *s && *s++ != '/' )
	{
		ret++;
	}

	return ret;
}

u32 U8Archive::NextEntryInFolder( u32 current, u32 directory ) const
{
	u32 next = ( fst[ current ].filetype ? fst[ current ].filelen : current + 1 );
	if( next < fst[ directory ].filelen )
		return next;

	return 0;
}

s32 U8Archive::EntryFromPath( const char *path, int d ) const
{
	//gprintf( "EntryFromPath( \"%s\", %d )\n", path, d );
	while( *path == '/' )
	{
		path++;
	}

	if( !fst[ d ].filetype )
	{
		gprintf("ERROR!!  %s is not a directory\n", FstName( &fst[ d ] ) );
		return -1;
	}

	u32 next = d + 1;

	FstEntry *entry = &fst[ next ];

	while( next )
	{
		//does this entry match.
		//strlen_slash is used because if looking for "dvd:/gameboy/" it would return a false positive if it hit "dvd:/gameboy advance/" first
		if( !strcasecmp_slash( path, FstName( entry ) ) && ( strlen( FstName( entry ) ) == (u32)strlen_slash( path ) ) )
		{
			char *slash = strchr( path, '/' );
			if( slash && *( slash + 1 ) )
			{
				//we are looking for a file somewhere in this folder
				return EntryFromPath( slash + 1, next );
			}
			//this is the actual entry we were looking for
			return next;
		}

		//find the next entry in this folder
		next = NextEntryInFolder( next, d );
		entry = &fst[ next ];
	}

	//no entry with the given path was found
	return -1;
}

U8NandArchive::U8NandArchive( const char* nandPath )
	: U8Archive( NULL, 0 ),
	  fd( -1 ),
	  dataOffset( 0 )
{
	if( nandPath )
	{
		SetFile( nandPath );
	}
}

U8NandArchive::~U8NandArchive()
{
	if( fd >= 0 )
	{
		ISFS_Close( fd );
	}
	free( fst );
}

bool U8NandArchive::SetFile( const char* nandPath )
{
	if(fst)
		free(fst);
	if(name_table)
		free(name_table);
	CloseFile();

	// open file
	if( (fd = ISFS_Open( nandPath, ISFS_OPEN_READ ) ) < 0 )
	{
		gprintf( "U8NandArchive:  ISFS_Open( \"%s\" ) failed\n", nandPath );
		return false;
	}

	// get file size
	fstats stats __attribute__(( aligned( 32 ) ));
	int ret = ISFS_GetFileStats( fd, &stats );
	if( ret < 0 )
	{
		CloseFile();
		gprintf( "U8NandArchive:  ISFS_GetFileStats( \"%s\" ) failed\n", nandPath );
		return false;
	}

	// buffer for reading the header and stuff
	u8* buffer = (u8*)memalign( 32, 0x800 );
	if( !buffer )
	{
		CloseFile();
		gprintf( "U8NandArchive: enomem\n" );
		return false;
	}

	// read a chunk big enough that it should contain the U8 header if there is going to be one
	if( (ret = ISFS_Read( fd, buffer, 0x800 )) != 0x800 )
	{
		free( buffer );
		CloseFile();
		gprintf( "U8NandArchive: ISFS_Read( 0x800 ) = %i\n", ret );
		return false;
	}

	// find the start of the U8 data
	U8Header* tagStart = (U8Header*)FindU8Tag( buffer, ret );
	if( !tagStart )
	{
		free( buffer );
		CloseFile();
		gprintf( "U8NandArchive: didn't see a U8 tag\n" );
		return false;
	}

	// remember where in the file the U8 starts
	dataOffset = ( (u8*)tagStart - buffer );

	// allocate memory and read the fst
	if( !(fst = (FstEntry *)memalign( 32, RU( tagStart->dataOffset - dataOffset, 32 ) ) )
			|| ( ISFS_Seek( fd, dataOffset + tagStart->rootNodeOffset, SEEK_SET ) != (s32)( dataOffset + tagStart->rootNodeOffset ) )
			|| ( ISFS_Read( fd, fst, tagStart->dataOffset - dataOffset ) != (s32)( tagStart->dataOffset - dataOffset ) )
			|| ( fst->filelen * 0xC > tagStart->dataOffset ) )
	{
		dataOffset = 0;
		free( buffer );
		if( fst )
			free( fst );
		CloseFile();
		gprintf( "U8NandArchive: error reading fst\n" );
	}

	// set name table pointer
	u32 name_table_offset = fst->filelen * 0xC;
	name_table = ((char *)fst) + name_table_offset;

	free( buffer );
	return true;
}

u8* U8NandArchive::GetFileAllocated( const char *path, u32 *size ) const
{
	//gprintf( "U8NandArchive::GetFileAllocated( %s )\n" );
	if( !path || !fst )
	{
		return NULL;
	}

	// find file
	int f = EntryFromPath( path, 0 );
	if( f < 1 || f >= (int)fst[ 0 ].filelen )
	{
		gprintf( "U8: \"%s\" wasn't found in the archive.\n", path );
		return NULL;
	}
	if( fst[ f ].filetype )
	{
		gprintf( "U8: \"%s\" is a folder\n", path );
		return NULL;
	}

	// create a buffer
	u8* ret = (u8*)memalign( 32, RU( fst[ f ].filelen, 32 ) );
	if( !ret )
	{
		gprintf( "U8: out of memory\n" );
		return NULL;
	}

	// seek and read
	if( ISFS_Seek( fd, dataOffset + fst[ f ].fileoffset, SEEK_SET ) != (s32)( dataOffset + fst[ f ].fileoffset )
			|| ISFS_Read( fd, ret, fst[ f ].filelen ) != (s32)fst[ f ].filelen )
	{
		free( ret );
		gprintf( "U8: error reading data from nand\n" );
		gprintf( "fd: %i  fst[ fd ].filelen: %08x\n", fd, fst[ f ].filelen );
		return NULL;
	}

	u32 len = fst[ f ].filelen;
	u8* ret2;
	// determine if it needs to be decompressed
	if( IsAshCompressed( ret, len ) )
	{
		// ASH0
		ret2 = DecompressAsh( ret, len );
		if( !ret2 )
		{
			free( ret );
			gprintf( "out of memory\n" );
			return NULL;
		}
		free( ret );
	}
	else if( isLZ77compressed( ret ) )
	{
		// LZ77 with no magic word
		if( decompressLZ77content( ret, len, &ret2, &len ) )
		{
			free( ret );
			return NULL;
		}
		free( ret );
	}
	else if( *(u32*)( ret ) == 0x4C5A3737 )// LZ77
	{
		// LZ77 with a magic word
		if( decompressLZ77content( ret + 4, len - 4, &ret2, &len ) )
		{
			free( ret );
			return NULL;
		}
		free( ret );
	}
	else
	{
		// already got what we are after
		ret2 = ret;
	}

	if( size )
	{
		*size = len;
	}

	// flush the cache so if there are any textures in this data, it will be ready for the GX
	DCFlushRange( ret2, len );
	return ret2;
}