libmaus::bitio::IndexedBitVector::unique_ptr_type libmaus::util::GenericIntervalTree::computeNonEmptyBV(::libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > const & V)
{
::libmaus::bitio::IndexedBitVector::unique_ptr_type BV(new ::libmaus::bitio::IndexedBitVector(V.size()));
for ( uint64_t i = 0; i < V.size(); ++i )
(*BV)[i] = (V[i].first != V[i].second);
BV->setupIndex();
return UNIQUE_PTR_MOVE(BV);
}
CharTermTable(uint8_t c)
: atable(257), table(atable.get()+1)
{
for ( unsigned int i = 0; i < 256; ++i )
table[i] = false;
table[-1] = true;
table[c] = true;
}
void reset()
{
lnumsyms = 0;
minlen = std::numeric_limits<uint64_t>::max();
maxlen = 0;
pc = C.begin();
p = 0;
}
SynchronousGenericOutputPosix(
std::string const & rfilename,
uint64_t const bufsize,
bool const truncate,
uint64_t const offset,
bool const rmetasync = true
)
: filename(rfilename), dirname(::libmaus::util::ArgInfo::getDirName(filename)), metasync(rmetasync),
B(bufsize), pa(B.get()), pc(pa), pe(pa+B.getN()),
fd ( -1 ),
totalwrittenbytes(0), totalwrittenwords(0)
{
while ( (fd = ::open(filename.c_str(), truncate ? (O_WRONLY|O_TRUNC|O_CREAT) : O_WRONLY , 0755 )) < 0 )
{
switch ( errno )
{
case EINTR:
{
std::cerr << "Restarting open() system call interupted by signal." << std::endl;
break;
}
default:
{
::libmaus::exception::LibMausException se;
se.getStream() << "Failed to open file "<< filename <<" in SynchronousGenericOutputPosix: " <<
strerror(errno);
se.finish();
throw se;
}
}
}
if ( lseek ( fd, offset, SEEK_SET) == static_cast<off_t>(-1) )
{
close(fd);
::libmaus::exception::LibMausException se;
se.getStream() << "Failed to seek " << filename << " in SynchronousGenericOutputPosix: " <<
strerror(errno);
se.finish();
throw se;
}
#if 0
std::cerr << "File " << filename << " opened for output in "
<< ::libmaus::util::Demangle::demangle<this_type>() << std::endl;
#endif
}
void getPattern(pattern_type & pat, uint64_t i) const
{
GetObject G(T.begin()+(*dict)[i - dict->FI.low]);
::libmaus::fastx::CompactFastQContext C;
C.nextid = i;
::libmaus::fastx::CompactFastQDecoderBase::decodePattern<GetObject>(G,*H,C,pat);
pat.patid = i;
}
void serialise(std::ostream & out) const
{
::libmaus::util::NumberSerialisation::serialiseNumber(out,slog);
::libmaus::util::NumberSerialisation::serialiseNumber(out,hashsize);
::libmaus::util::NumberSerialisation::serialiseNumber(out,hashmask);
::libmaus::util::NumberSerialisation::serialiseNumber(out,fill);
H.serialize(out);
}
unique_ptr_type extend() const
{
unique_ptr_type O(new this_type(slog+1));
for ( uint64_t i = 0; i < H.size(); ++i )
if ( H[i].first != base_type::unused() )
O->insert ( H[i].first, H[i].second );
return UNIQUE_PTR_MOVE(O);
}
void cleanup()
{
for ( uint64_t i = 0; i < alloclist.size(); ++i )
delete alloclist[i];
alloclist = libmaus::autoarray::AutoArray<element_type *>(0);
freelist = libmaus::autoarray::AutoArray<element_type *>(0);
freelistfill = 0;
}
uint64_t createFinalStream(stream_type & out)
{
flush();
uint64_t p = 0;
p += ::libmaus::util::NumberSerialisation::serialiseNumber(out,symbols); // n
p += root->serialize(out); // huffman code tree
p += ::libmaus::util::NumberSerialisation::serialiseNumber(out,contexts.size()); // number of bit vectors
std::vector<uint64_t> nodeposvec;
for ( uint64_t i = 0; i < contexts.size(); ++i )
{
nodeposvec.push_back(p);
uint64_t const blockswritten = contexts[i]->blockswritten;
uint64_t const datawordswritten = 6*blockswritten;
uint64_t const allwordswritten = 8*blockswritten;
contexts[i].reset();
tmpcnt.closeOutputTempFile(i);
// bits written
p += ::libmaus::serialize::Serialize<uint64_t>::serialize(out,64*datawordswritten);
// auto array header (words written)
p += ::libmaus::serialize::Serialize<uint64_t>::serialize(out,allwordswritten);
//std::string const filename = outputfilenames[i];
//::libmaus::aio::CheckedInputStream istr(filename);
std::istream & istr = tmpcnt.openInputTempFile(i);
// std::ifstream istr(filename.c_str(),std::ios::binary);
// std::cerr << "Copying " << allwordswritten << " from stream " << filename << std::endl;
::libmaus::util::GetFileSize::copy (istr, out, allwordswritten, sizeof(uint64_t));
p += allwordswritten * sizeof(uint64_t);
tmpcnt.closeInputTempFile(i);
// remove(filename.c_str());
}
uint64_t const indexpos = p;
p += ::libmaus::util::NumberSerialisation::serialiseNumberVector(out,nodeposvec);
p += ::libmaus::util::NumberSerialisation::serialiseNumber(out,indexpos);
out.flush();
return p;
}
libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > libmaus::util::GenericIntervalTree::computeNonEmpty(::libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > const & V)
{
uint64_t nonempty = 0;
for ( uint64_t i = 0; i < V.size(); ++i )
if ( V[i].first != V[i].second )
nonempty++;
if ( nonempty == 0 )
std::cerr << "all of the " << V.size() << " intervals are empty." << std::endl;
::libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > R(nonempty);
nonempty = 0;
for ( uint64_t i = 0; i < V.size(); ++i )
if ( V[i].first != V[i].second )
R [ nonempty++ ] = V[i];
return R;
}
BamSeqEncodeTable()
: A(256)
{
char const * s = "=ACMGRSVTWYHKDBN";
for ( uint64_t i = 0; i < A.size(); ++i )
A[i] = strlen(s);
for ( uint64_t i = 0; i < strlen(s); ++i )
A [ s[i] ] = i;
}
void flush()
{
for ( uint64_t i = 0; i < contexts.size(); ++i )
{
contexts[i]->writeBit(0);
contexts[i]->flush();
// std::cerr << "Flushed context " << i << std::endl;
}
}
BgzfDeflateParallel(std::ostream & rdeflateout, uint64_t const rnumthreads, uint64_t const rnumbuffers, int const level, std::ostream * rdeflateindexostr = 0)
: deflategloblist(), deflatecontext(deflategloblist,rdeflateout,rnumbuffers,level,BgzfDeflateParallelContext::getDefaultDeflateGetCur(),rdeflateindexostr), T(rnumthreads)
{
for ( uint64_t i = 0; i < T.size(); ++i )
{
BgzfDeflateParallelThread::unique_ptr_type tTi(new BgzfDeflateParallelThread(deflatecontext));
T[i] = UNIQUE_PTR_MOVE(tTi);
T[i]->start();
}
}
element_type * get()
{
if ( ! freelistfill )
{
// allocate more alignment objects
libmaus::autoarray::AutoArray<element_type *> nalloclist(
std::max(
static_cast<uint64_t>(1),
static_cast<uint64_t>(2*alloclist.size())
)
,false
);
std::copy(alloclist.begin(),alloclist.end(),nalloclist.begin());
element_type * nullp = 0;
std::fill(nalloclist.begin()+alloclist.size(),nalloclist.end(),nullp);
for ( element_type ** p = nalloclist.begin()+alloclist.size();
p != nalloclist.end(); ++p )
*p = new element_type;
libmaus::autoarray::AutoArray<element_type *> nfreelist(
std::max(
static_cast<uint64_t>(1),
static_cast<uint64_t>(2*freelist.size())
)
,false
);
std::copy(freelist.begin(),freelist.end(),nfreelist.begin());
std::fill(nfreelist.begin()+freelist.size(),nfreelist.end(),nullp);
freelist = nfreelist;
for ( element_type ** p = nalloclist.begin()+alloclist.size();
p != nalloclist.end(); ++p )
freelist[freelistfill++] = *p;
alloclist = nalloclist;
}
return freelist[--freelistfill];
}
SynchronousOutputBuffer8(std::string const & rfilename, uint64_t const bufsize, bool truncate = true)
: filename(rfilename), B(bufsize), pa(B.get()), pc(pa), pe(pa+B.getN())
{
if ( truncate )
{
std::ofstream ostr(filename.c_str(), std::ios::binary);
ostr.flush();
}
}
::libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > computeSymAccu()
{
uint64_t numint = 0;
for ( uint64_t i = 0; i < index.size(); ++i )
numint += index[i].size();
::libmaus::autoarray::AutoArray<uint64_t> preaccu(numint+1);
uint64_t k = 0;
for ( uint64_t i = 0; i < index.size(); ++i )
for ( uint64_t j = 0; j < index[i].size(); ++j )
preaccu[k++] = index[i][j].vcnt;
preaccu.prefixSums();
::libmaus::autoarray::AutoArray< std::pair<uint64_t,uint64_t> > accu(numint);
for ( uint64_t i = 1; i < preaccu.size(); ++i )
accu[i-1] = std::pair<uint64_t,uint64_t>(
std::pair<uint64_t,uint64_t>(preaccu[i-1],preaccu[i])
);
return accu;
}
SocketFastReaderBase(::libmaus::network::SocketBase * rsocket, uint64_t const bufsize)
:
socket(rsocket),
B(bufsize),
pa(B.get()),
pc(pa),
pe(pc),
c(0)
{
}
~FileBunchLRU()
{
for ( uint64_t i = 0; i < files.size(); ++i )
if ( files[i].get() )
{
files[i] -> flush();
files[i] -> close();
files[i].reset();
}
}
static void writeArray(::libmaus::autoarray::AutoArray<data_type> const & A,
std::string const & outputfilename)
{
this_type out(outputfilename,64*1024);
for ( uint64_t i = 0; i < A.getN(); ++i )
out.put(A[i]);
out.flush();
}
AsynchronousBufferReader(
std::string const & filename,
uint64_t const rnumbufs,
uint64_t const rbufsize,
uint64_t const offset
)
: std::ifstream(filename.c_str()), bufsize(rnumbufs * rbufsize),
abuffer(bufsize), buffer(abuffer.get()), av(true)
{
std::ifstream::seekg(offset,std::ios::beg);
}
bool getNextTriple(TripleEdge & triple)
{
if ( ! curbufleft )
{
istr.read ( reinterpret_cast<char *>(B.get()), B.getN() * sizeof(TripleEdge) );
if ( istr.gcount() == 0 )
return false;
assert ( istr.gcount() % sizeof(TripleEdge) == 0 );
curbufleft = istr.gcount() / sizeof(TripleEdge);
assert ( curbufleft );
curtrip = B.get();
}
triple = *(curtrip++);
curbufleft -= 1;
return true;
}
char const * prevStart(char const * e) const
{
if ( e == C.begin() )
return 0;
assert ( e[-1] == '\n' );
// step over last/quality line's newline
--e;
// search for plus line's newline
while ( *--e != '\n' ) {}
// search for sequence line's newline
while ( *--e != '\n' ) {}
// search for id line's newline
while ( *--e != '\n' ) {}
// search for start of line
while ( e != C.begin() && e[-1] != '\n' )
--e;
return e;
}
uint64_t lookupValue(uint64_t const v) const
{
if ( !index.size() )
return 0;
typedef IndexEntryValueGetAdapter<IndexEntry const *> adapter_type;
adapter_type IEKGA(index.get());
::libmaus::util::ConstIterator<adapter_type,uint64_t> const ita(&IEKGA);
::libmaus::util::ConstIterator<adapter_type,uint64_t> ite(ita);
ite += index.size();
::libmaus::util::ConstIterator<adapter_type,uint64_t> R =
::std::lower_bound(ita,ite,v);
if ( R == ite )
return index.size()-1;
if ( v == *R )
return R-ita;
else
return (R-ita)-1;
}
void extendInternal()
{
unique_ptr_type O(new this_type(slog+1));
for ( uint64_t i = 0; i < H.size(); ++i )
if ( H[i].first != base_type::unused() )
O->insert ( H[i].first, H[i].second );
slog = O->slog;
hashsize = O->hashsize;
hashmask = O->hashmask;
fill = O->fill;
H = O->H;
}
int_type underflow()
{
// if there is still data, then return it
if ( gptr() < egptr() )
return static_cast<int_type>(*uptr());
assert ( gptr() == egptr() );
// number of bytes for putback buffer
uint64_t const putbackcopy = std::min(
static_cast<uint64_t>(gptr() - eback()),
putbackspace
);
// copy bytes
std::copy(
gptr()-putbackcopy,
gptr(),
buffer.begin() + putbackspace - putbackcopy
);
// load data
uint64_t const uncompressedsize = stream.read(
buffer.begin()+putbackspace,
buffer.size()-putbackspace
);
// set buffer pointers
setg(
buffer.begin()+putbackspace-putbackcopy,
buffer.begin()+putbackspace,
buffer.begin()+putbackspace+uncompressedsize);
symsread += uncompressedsize;
if ( uncompressedsize )
return static_cast<int_type>(*uptr());
else
return traits_type::eof();
}
void checkSpace(uint64_t const outlen)
{
// buffer overflow?
if ( freeSpace() < outlen )
{
flush();
assert ( opc == opa );
if ( outlen > outbuf.size() )
{
::libmaus::autoarray::AutoArray<uint8_t> newbuf(outlen);
std::copy( outbuf.begin(), outbuf.end(), newbuf.begin() );
outbuf = newbuf;
opa = outbuf.begin();
opc = opa;
ope = outbuf.end();
}
}
assert ( freeSpace() >= outlen );
}
/* decode next block */
bool decodeBlock()
{
/* open new file if necessary */
bool changedfile = false;
while ( fileptr < idda.data.size() && blockptr == idda.data[fileptr].numentries )
{
fileptr++;
blockptr = 0;
changedfile = true;
}
if ( fileptr == idda.data.size() )
return false;
if ( changedfile )
openNewFile();
/* align to word boundary */
GD->flush();
/* read block size */
uint64_t const blocksize = GD->decodeWord(32);
/* increase size of memory buffer if necessary */
if ( blocksize > decodebuf.size() )
decodebuf.resize(blocksize);
/* set buffer pointers */
pa = decodebuf.begin();
pc = pa;
pe = pa + blocksize;
/* decode block */
for ( uint64_t i = 0; i < blocksize; ++i )
decodebuf[i] = GD->decode();
/* increment block pointer */
blockptr++;
return true;
}
static void merge(std::vector<std::string> const & inputfilenames, std::string const & outputfilename)
{
CheckOverlapResultMergeInput in(inputfilenames);
std::ofstream ostr(outputfilename.c_str(),std::ios::binary);
CheckOverlapResult::shared_ptr_type ptr;
while ( (ptr=in.get()) )
ptr->serialise(ostr);
ostr.flush();
assert ( ostr );
ostr.close();
}
void internalFlush()
{
if ( pathigh != patlow )
{
#if defined(LIBMAUS_FASTX_FASTQBGZFWRITER_PARALLEL)
uint64_t const bcnt = bgzfenc->writeSyncedCount(C.begin(),pc-C.begin());
libmaus::util::UTF8::encodeUTF8(bcnt,*bgzfidxcntoutstr);
libmaus::fastx::FastInterval const FI(patlow,pathigh,0,0,lnumsyms,minlen,maxlen);
#else
std::pair<uint64_t,uint64_t> bcntccnt = bgzfenc->writeSyncedCount(C.begin(),pc-C.begin());
libmaus::fastx::FastInterval const FI(patlow,pathigh,cacc,cacc+bcntccnt.second,lnumsyms,minlen,maxlen);
cacc += bcntccnt.second;
#endif
(*fioutstr) << FI.serialise();
blockcnt += 1;
std::cerr << FI << std::endl;
reset();
patlow = pathigh;
}
}
SynchronousGenericInput(std::istream & ristr, uint64_t const rbufsize,
uint64_t const rtotalwords = std::numeric_limits<uint64_t>::max(),
bool const rcheckmod = true
)
: bufsize(rbufsize), buffer(bufsize,false),
pa(buffer.get()), pc(pa), pe(pa),
Pistr(),
istr(ristr),
totalwords ( rtotalwords ),
totalwordsread(0),
checkmod(rcheckmod)
{
}
