int main (int argc, char* argv[])
{
/** We create a command line parser. */
OptionsParser parser ("BankFilter");
parser.push_back (new OptionOneParam (STR_URI_INPUT, "bank reference", true));
parser.push_back (new OptionOneParam (STR_URI_SEQ_IDS, "file holding indexes of bank", true));
try
{
/** We parse the user options. */
IProperties* options = parser.parse (argc, argv);
/** We read the list of indexes. */
set<size_t> indexes;
FILE* file = fopen (options->getStr(STR_URI_SEQ_IDS).c_str(), "r");
if (file != 0)
{
char buffer[128];
while (fgets (buffer, sizeof(buffer), file)) { indexes.insert (atoi(buffer)); }
fclose (file);
}
cout << "found " << indexes.size() << " indexes" << endl;
/** We open the output bank. */
string outputBankUri = options->getStr(STR_URI_INPUT) + "_" + System::file().getBaseName (options->getStr(STR_URI_SEQ_IDS));
IBank* outputBank = Bank::open (outputBankUri);
LOCAL (outputBank);
/** We loop the input bank. */
IBank* inputBank = Bank::open (options->getStr(STR_URI_INPUT));
LOCAL (inputBank);
/** We use another iterator for filtering out some sequences. */
FilterIterator<Sequence,FilterFunctor> itSeq (inputBank->iterator(), FilterFunctor(indexes));
/** We loop the sequences. */
for (itSeq.first(); !itSeq.isDone(); itSeq.next())
{
outputBank->insert (itSeq.item());
}
/** We flush the output bank. */
outputBank->flush();
}
catch (OptionFailure& e)
{
return e.displayErrors (cout);
}
catch (Exception& e)
{
cerr << "EXCEPTION: " << e.getMessage() << endl;
}
}
int main (int argc, char* argv[])
{
/** We create a command line parser. */
OptionsParser parser ("BankFilter");
parser.push_back (new OptionOneParam (STR_URI_INPUT, "bank input", true));
parser.push_back (new OptionOneParam (STR_FILTER_RATIO, "skip a sequence if 'good letters number / seq.len > X'", false, "0.8"));
try
{
/** We parse the user options. */
IProperties* options = parser.parse (argc, argv);
/** Shortcuts. */
double percentThreshold = options->getDouble(STR_FILTER_RATIO);
/** We open the input bank. */
IBank* inBank = Bank::open (options->getStr(STR_URI_INPUT));
LOCAL (inBank);
/** We create the output inBank. */
IBank* outBank = new BankFasta (options->getStr(STR_URI_INPUT) + "_filtered");
LOCAL (outBank);
/** We iterate the inBank. NOTE: WE USE A LAMBDA EXPRESSION HERE. */
inBank->iterate ([&] (Sequence& s)
{
/** Shortcut. */
char* data = s.getDataBuffer();
size_t nbOK = 0;
for (size_t i=0; i<s.getDataSize(); i++)
{
if (data[i]=='A' || data[i]=='C' || data[i]=='G' || data[i]=='T') { nbOK++; }
}
if ((double)nbOK / (double)s.getDataSize() > percentThreshold) { outBank->insert (s); }
});
/** We flush the output bank. */
outBank->flush();
}
catch (OptionFailure& e)
{
return e.displayErrors (cout);
}
catch (Exception& e)
{
cerr << "EXCEPTION: " << e.getMessage() << endl;
}
}
int main (int argc, char* argv[])
{
/** We create a command line parser. */
OptionsParser parser ("BankSplitter");
parser.push_back (new OptionOneParam (STR_URI_INPUT, "bank reference", true));
parser.push_back (new OptionOneParam (STR_MAX_INPUT_SIZE, "average db size per split", true));
parser.push_back (new OptionOneParam (STR_URI_OUTPUT_DIR, "output directory", false, "."));
parser.push_back (new OptionNoParam (STR_OUTPUT_FASTQ, "fastq output", false));
parser.push_back (new OptionNoParam (STR_OUTPUT_GZ, "gzip output", false));
// We define a try/catch block in case some method fails (bad filename for instance)
try
{
/** We parse the user options. */
IProperties* options = parser.parse (argc, argv);
/** Shortcuts. */
u_int64_t maxDbSize = options->getInt(STR_MAX_INPUT_SIZE);
// We declare an input Bank
IBank* inputBank = Bank::open (options->getStr(STR_URI_INPUT));
LOCAL (inputBank);
// We get the basename of the input bank.
string inputBasename = System::file().getBaseName (options->getStr(STR_URI_INPUT));
/** We set the name of the output directory. */
stringstream ss; ss << inputBasename << "_S" << maxDbSize;
string outputDirName = ss.str();
/** We create the output directory. */
string outputDir = options->getStr(STR_URI_OUTPUT_DIR) + "/" + outputDirName;
System::file().mkdir (outputDir, S_IRWXU);
// We create the album bank.
BankAlbum album (outputDir + "/album.txt");
/** We get estimations about the bank. */
u_int64_t number, totalSize, maxSize;
inputBank->estimate (number, totalSize, maxSize);
u_int64_t estimationNbSeqToIterate = number;
// We create an iterator over the input bank
ProgressIterator<Sequence> itSeq (*inputBank, "split");
// We loop over sequences to get the exact number of sequences.
int64_t nbBanksOutput = -1;
u_int64_t nbSequences = 0;
u_int64_t dbSize = ~0;
bool isFastq = options->get(STR_OUTPUT_FASTQ) != 0;
bool isGzipped = options->get(STR_OUTPUT_GZ) != 0;
IBank* currentBank = 0;
for (itSeq.first(); !itSeq.isDone(); itSeq.next())
{
if (dbSize > maxDbSize)
{
if (currentBank != 0) { currentBank->flush(); currentBank->finalize(); }
nbBanksOutput ++;
/** We build the uri of the current bank. */
stringstream ss; ss << inputBasename << "_" << nbBanksOutput << (isFastq ? ".fastq" : ".fasta");
if (isGzipped) { ss << ".gz"; }
/** We create a new bank and put it in the album. */
currentBank = album.addBank (outputDir, ss.str(), isFastq, isGzipped);
/** We reinit the db size counter. */
dbSize = 0;
}
dbSize += itSeq->getDataSize();
/** We insert the sequence into the current output bank. */
currentBank->insert (*itSeq);
}
if (currentBank != 0) { currentBank->flush(); }
}
catch (OptionFailure& e)
{
return e.displayErrors (cout);
}
catch (Exception& e)
{
cerr << "EXCEPTION: " << e.getMessage() << endl;
}
}