/**
* @method HandleCharset [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Reads and stores information contained in the command
* CHARSET within an ASSUMPTIONS block.
*/
void AssumptionsBlock::HandleCharset( NexusToken& token )
{
bool asterisked = false;
// Next token should be either an asterisk or the name of a charset
token.GetNextToken();
if( token.Equals("*") ) {
asterisked = true;
token.GetNextToken();
}
// Token now stored should be the name of a charset
nxsstring charset_name = token.GetToken();
// Now grab the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' in CHARSET definition but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
assert( charBlockPtr != NULL );
CharactersBlock& charBlock = *charBlockPtr;
IntSet s;
int totalChars = charBlock.GetNCharTotal();
SetReader( token, totalChars, s, charBlock, SetReader::charset ).Run();
charsets[charset_name] = s;
if( asterisked )
def_charset = charset_name;
}
/**
* @method HandleTaxset [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* ?
*/
void AssumptionsBlock::HandleTaxset( NexusToken& token )
{
bool asterisked = false;
// Next token should be either an asterisk or the name of a taxset
token.GetNextToken();
if( token.Equals("*") ) {
asterisked = true;
token.GetNextToken();
}
// Token now stored should be the name of a taxset
nxsstring taxset_name = token.GetToken();
// Now grab the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' in TAXSET definition but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
IntSet s;
int totalTaxa = taxa.GetNumTaxonLabels();
SetReader( token, totalTaxa, s, *this, SetReader::taxset ).Run();
taxsets[taxset_name] = s;
if( asterisked )
def_taxset = taxset_name;
}
/**
* @method HandleExecute [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XBull
*
* Handles everything after the EXecute keyword and the terminating
* semicolon. Flushes all blocks before executing file specified,
* and no warning is given of this
*/
void BullShell::HandleExecute( NexusToken& token ) {
token.GetNextToken();
std::string fn = token.GetToken();
token.GetNextToken();
RequireSemicolon(token, "EXECUTE");
cout << "\nExecuting " << fn << "..." << endl;
readFile(fn, false);
cout << "Done with file "<<fn<<endl;
}
EncodingType BullShell::InterpretOutputType(NexusToken &token) const {
if (token.Abbreviation("PROtein") || token.Abbreviation("AMino"))
return EncodingType(AminoAcid);
if (token.Abbreviation("Dna"))
return EncodingType(DNANoGap);
errormsg.clear();
errormsg << "The output type " << token.GetToken() << " is invalid";
throw XBull(errormsg, token);
}
void BullShell::ThrowNoSemicolon(NexusToken & token, const char *cmd) const {
errormsg = "Expecting ';'";
if (cmd != NULL) {
errormsg += " to terminate the ";
errormsg.append(cmd);
errormsg += "command,";
}
errormsg += " but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XBull(errormsg, token);
}
void MyCharactersBlock::handleContMatrix (NexusToken& token)
//: read in a matrix of continuous data and store it
{
std::vector<double> theRow;
for(;;)
{
// grab the first token in the row
string theTokenVal;
token.GetNextToken();
theTokenVal = (token.GetToken()).c_str();
// if reached the end of the matrix, break loop
if (token.Equals (";"))
break;
// otherwise store the name, read and store the data
mTaxaNames.push_back (theTokenVal);
theRow.clear();
for (int i = 0; i < nchar; i++)
{
token.GetNextToken();
theTokenVal = (token.GetToken()).c_str();
theRow.push_back (sbl::toDouble (theTokenVal));
}
// store the row in the matrix
mContData.appendRow (theRow);
}
assert (mTaxaNames.size() == mContData.size());
for (sbl::ulong i = 1; i < mContData.size(); i++)
assert (mContData[1].size() == mContData[i].size());
// store names in continuous data matrix
for (unsigned int i = 0; i < mTaxaNames.size(); i++)
mContData.setRowName (i, mTaxaNames[i].c_str());
}
void ReadNexusVector(std::vector<std::string> *v, NexusToken &token) { // TAH changed from above
if (token.Equals("(")) {
token.GetNextToken();
while (!token.Equals(")")) {
v->push_back(token.GetToken());
token.GetNextToken();
}
}
else
v->push_back(token.GetToken());
}
/**
* @method HandleEndblock [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Called when the END or ENDBLOCK command needs to be parsed
* from within the ASSUMPTIONS block. Basically just checks to make
* sure the next token in the data file is a semicolon.
*/
void AssumptionsBlock::HandleEndblock( NexusToken& token )
{
// get the semicolon following END or ENDBLOCK token
//
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate the END or ENDBLOCK command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
/**
* @method HandleNextPass [int:protected]
* @param token [NexusToken&] the token we are using for reading the data file
* @param offset [int&] the offset (see below)
*
* Called from within HandleMatrix, this function is used to deal with interleaved
* matrices. It is called once for each pass through the taxa.
*
* <P>The local variable jmax records the number of columns read in the current
* interleaved page and is used to determine the offset used for j in subsequent
* pages.
*/
int DistancesBlock::HandleNextPass( NexusToken& token, int& offset )
{
int i, j, k, jmax = 0, done = 0;
int i_first = 0;
if( triangle == lower )
i_first = offset;
int i_last = ntax;
for( i = i_first; i < i_last; i++ )
{
// Deal with taxon label if provided
//
if( labels && ( !newtaxa || offset>0 ) )
{
do {
token.SetLabileFlagBit( NexusToken::newlineIsToken );
token.GetNextToken();
} while( token.AtEOL() );
try {
k = taxa.FindTaxon( token.GetToken() );
if( taxonPos[i]==-1 )
taxonPos[i] = k;
else if( taxonPos[i] != k ) {
errormsg = "Taxon labeled ";
errormsg += token.GetToken();
errormsg += " is out of order compared to previous interleave pages";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
catch (std::out_of_range) {
errormsg = "Could not find ";
errormsg += token.GetToken();
errormsg += " among taxa previously defined";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
else if( labels && newtaxa ) {
do {
token.SetLabileFlagBit( NexusToken::newlineIsToken );
token.GetNextToken();
} while( token.AtEOL() );
taxa.AddTaxonLabel( token.GetToken() );
taxonPos[i] = i;
}
// Now deal with the row of distance values
//
int true_j = 0;
for( j = 0; j < ntax; j++ )
{
if( i == ntax-1 && j == ntax-1 ) {
done = 1;
}
if( i == ntax-1 && true_j == ntax-1 ) {
done = 1;
break;
}
if( i == ntax-1 && !diagonal && triangle==upper ) {
done = 1;
break;
}
if( !diagonal && triangle==lower && j == ntax-offset-1) {
done = 1;
break;
}
token.SetLabileFlagBit( NexusToken::newlineIsToken );
token.GetNextToken();
if( token.AtEOL() ) {
if( j > jmax ) {
jmax = j;
if( !diagonal && triangle == upper && i >= offset )
jmax++;
if( interleave && triangle == upper )
i_last = jmax + offset;
}
break;
}
true_j = j+offset;
if( triangle==upper && i > offset )
true_j += ( i - offset );
if( !diagonal && triangle==upper && i >= offset )
true_j++;
if( true_j == ntax ) {
errormsg = "Too many distances specified in row just read in";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
std::string t = token.GetToken();
if( token.GetTokenLength() == 1 && t[0] == missing )
SetMissing( i, true_j );
else
SetDistance( i, true_j, atof( t.c_str() ) );
}
}
offset += jmax;
if( done )
return 1;
else
return 0;
}
SimulateOpts BullShell::ParseSimulateCommand(NexusToken& token ) const {
token.GetNextToken();
int nOutputs = 0;
int nSimChars = 0;
SimulateOpts simOpts;
const unsigned nTrees = kernel.getNumTrees();
for (; token.GetToken() != ";"; token.GetNextToken()) {
if (token.Abbreviation("NReps")) {
DemandEquals(token, "after NReps subcommand of Simulate command");
simOpts.nReps = DemandPositiveInt(token, "after NReps subcommand of Simulate command");
}
if (token.Abbreviation("Seed")) {
DemandEquals(token, "after Seed subcommand of Simulate command");
simOpts.seed = DemandPositiveInt(token, "after Seed subcommand of Simulate command");
}
else if (token.Abbreviation("COLLapsefile")) {
DemandEquals(token, "after CollapseFile subcommand of Simulate command");
token.GetNextToken();
simOpts.collapsedTreeFilename = token.GetTokenReference();
}
else if (token.Abbreviation("NChars")) {
DemandEquals(token, "after NChars subcommand of Simulate command");
simOpts.nSimChars = DemandPositiveInt(token, "after NChars subcommand of Simulate command");
}
else if (token.Abbreviation("Concat")) {
DemandEquals(token, "after Concat subcommand of GetTrees command");
simOpts.concatenations = DemandPositiveInt(token, "after Concat subcommand of GetTrees command");
}
else if (token.Abbreviation("Verbose")) {
simOpts.verbose = true;
}
else if (token.Abbreviation("NOUTput")) {
DemandEquals(token, "after NOutput subcommand of Simulate command");
nOutputs = DemandPositiveInt(token, "after NOutput subcommand of Simulate command");
if (nOutputs < 1) {
errormsg="nOutputs must be > 0";
throw XBull( errormsg, token);
}
}
else if (token.Abbreviation("Paupblockfile")) {
DemandEquals(token, "after Paupblockfile subcommand of Simulate");
token.GetNextToken();
ReadNexusVector(&(simOpts.paupBlockFilename), token);
}
else if (token.Abbreviation("OVERWrite"))
simOpts.overwrite = true;
else if (token.Abbreviation("File")) {
DemandEquals(token, "after File subcommand of Simulate");
token.GetNextToken();
ReadNexusVector(&(simOpts.outputFilenames), token);
}
else if (token.Abbreviation("TAg")) {
DemandEquals(token, "after Tag subcommand of Simulate command");
token.GetNextToken();
simOpts.tagname = token.GetTokenReference();
}
else if (token.Abbreviation("AUTOmatic"))
simOpts.automatic = true;
else if (token.Abbreviation("OUTputtypes")) {
DemandEquals(token, "after OutputTypes subcommand of Simulate");
token.GetNextToken();
EncodingType et;
if (token.Equals("(")) {
token.GetNextToken();
while (!token.Equals(")")) {
et = InterpretOutputType(token);
simOpts.outTypes.push_back(et);
token.GetNextToken();
}
}
else {
et = InterpretOutputType(token);
simOpts.outTypes.push_back(et);
}
}
else {
Tree * temptree;
long tn;
if (NxsString::to_long(token.GetTokenAsCStr(), &tn)) {
if ((unsigned) tn > nTrees || tn < 1) {
errormsg = "Tree ";
errormsg += token.GetToken();
errormsg += " unknown. simulation aborted.";
throw XBull( errormsg, token);
}
temptree = kernel.getTree(tn - 1);
}
else {
try {
temptree = kernel.FindTreeFromName(token.GetToken());
}
catch (NoSuchTree) {
errormsg << "Tree " << token.GetToken() << " unknown.";
throw XBull( errormsg, token);
}
}
simOpts.treeNames.push_back(token.GetToken()); //keeps track of what the user called the tree
simOpts.treeAlias.push_back(temptree);
}
}
if (simOpts.outTypes.size() == 0){
simOpts.outTypes.push_back(EncodingType(DNANoGap));
}
if ((nOutputs > 0) && simOpts.outTypes.size() != (unsigned) nOutputs) {
errormsg = "The number of OutputTypes in a Simulate command must match the number of specified in the NOutput option";
throw XBull( errormsg, token);
}
if (simOpts.automatic) {
if (simOpts.tagname.empty()) {
errormsg = "The Tag option of the Simulate command must be used if the Automatic naming mode is requested.";
throw XBull( errormsg, token);
}
if (!simOpts.outputFilenames.empty()) {
errormsg = "The FILE option of the Simulate command cannot be used if the Automatic naming mode is requested (use the Tag instead of file).";
throw XBull( errormsg, token);
}
simOpts.outputFilenames.clear();
std::vector<EncodingType>::const_iterator ot = simOpts.outTypes.begin();
for (; ot != simOpts.outTypes.end(); ot++) {
string ofname = EncodingTypeToString(*ot);
ofname.append(simOpts.tagname);
simOpts.outputFilenames.push_back(ofname);
}
}
else if (simOpts.outTypes.size() != simOpts.outputFilenames.size()) {
errormsg = "The number of OutputTypes in a Simulate command must match the number of Files specified";
throw XBull( errormsg, token);
}
if ((!simOpts.paupBlockFilename.empty()) && simOpts.paupBlockFilename.size() != simOpts.outTypes.size()) {
errormsg = "The number of PaupBlockFile in a Simulate command must match the number of OutputTypes specified";
throw XBull( errormsg, token);
}
//TEMPORARY lots of the interaction with SSettings assumes that you are using SSRFCodonSubModel
if (!kernel.hasModel()) {
errormsg="You can't simulate a codon model without first declaring the models using codLikeStartVal";
throw XBull( errormsg, token);
}
if ( simOpts.concatenations == 0) {
if (nSimChars == 0) {
errormsg="Either the number of characters or the number of concatenations must be specified";
throw XBull( errormsg, token);
}
const unsigned naa = kernel.getModelConstRef().getNumAASites();
const unsigned ndna = 3*naa;
if (nSimChars % ndna) {
errormsg="The number of characters must be a multiple of the number of amino acids";
throw XBull( errormsg, token);
}
simOpts.concatenations = ndna/nSimChars;
}
return simOpts;
}
/**
* @method HandleLog [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XBull
*
* Called when the LOG command needs to be parsed
* from within the BullShell block.
*/
void BullShell::HandleLog( NexusToken& token )
{
bool starting = false;
bool stopping = false;
bool appending = false;
bool replacing = false;
std::string logfname;
errormsg.clear();
// Retrieve all tokens for this command, stopping only in the event
// of a semicolon or an unrecognized keyword
//
for (token.GetNextToken(); !token.Equals(";"); token.GetNextToken()) {
if ( token.Abbreviation("STOp") )
stopping = true;
else if ( token.Abbreviation("STArt") )
starting = true;
else if ( token.Abbreviation("Replace") )
replacing = true;
else if ( token.Abbreviation("Append") )
appending = true;
else if ( token.Abbreviation("File") ) {
DemandEquals(token, "after file subcommand of Log command");
token.GetNextToken();
logfname = token.GetTokenReference();
}
else {
errormsg << "Unexpected keyword (" << token.GetToken() << ") encountered reading LOG command";
throw XBull( errormsg, token);
}
}
// Check for incompatible combinations of keywords
if ( stopping && ( starting || appending || replacing || !logfname.empty() ) ) {
errormsg = "Cannot specify STOP with any of the following START, APPEND, REPLACE, FILE";
throw XBull( errormsg, token);
}
if ( appending && replacing ) {
errormsg = "Cannot specify APPEND and REPLACE at the same time";
throw XBull( errormsg, token);
}
if ( stopping ) {
ioObject.stopLogging();
kernel.logEachStep=false;
return;
}
if (logfname.empty() ) {
errormsg << "Must provide a file name when opening a log file\ne.g., log file=doofus.txt start replace;";
throw XBull(errormsg, token);
}
if (!appending && !replacing && FileExists(logfname.c_str())) {
errormsg << "File " << logfname << " exists.\n APPEND or REPLACE must be specified in the LOG command.";
throw XBull( errormsg, token);
}
ioObject.startLogging(logfname, appending);
}
/**
* @method HandleMatrixCommand [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Called when MATRIX command needs to be parsed from within the
* DISTANCES block. Deals with everything after the token MATRIX
* up to and including the semicolon that terminates the MATRIX
* command.
*/
void DistancesBlock::HandleMatrixCommand( NexusToken& token )
{
int i;
if( ntax == 0 )
ntax = taxa.GetNumTaxonLabels();
if( ntax == 0 ) {
errormsg = "MATRIX command cannot be read if NTAX is zero";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
if( triangle == both && !diagonal ) {
errormsg = "Cannot specify NODIAGONAL and TRIANGLE=BOTH at the same time";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
if( newtaxa )
taxa.Reset();
// allocate memory to hold the taxonPos array
//
if( taxonPos != NULL )
delete [] taxonPos;
taxonPos = new int[ntax];
for( i = 0; i < ntax; i++ )
taxonPos[i] = -1;
// allocate memory to hold the matrix
//
if( matrix != NULL )
delete matrix;
matrix = new DistanceDatum*[ntax];
for( i = 0; i < ntax; i++ )
matrix[i] = new DistanceDatum[ntax];
int offset = 0;
int done = 0;
while( !done ) {
done = HandleNextPass( token, offset );
}
// token should be equal to the terminating semicolon
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate MATRIX command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
/**
* @method HandleTaxlabelsCommand [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Called when TAXLABELS command needs to be parsed from within the
* DISTANCES block. Deals with everything after the token TAXLABELS
* up to and including the semicolon that terminates the TAXLABELS
* command.
*/
void DistancesBlock::HandleTaxlabelsCommand( NexusToken& token )
{
if( !newtaxa ) {
errormsg = "NEWTAXA must have been specified in DIMENSIONS command to use the TAXLABELS command in a ";
errormsg += id;
errormsg += " block";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
if( ntax == 0 ) {
errormsg = "NTAX must be specified before TAXLABELS command";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
for( int i = 0; i < ntax; i++ ) {
token.GetNextToken();
taxa.AddTaxonLabel( token.GetToken() );
}
token.GetNextToken(); // this should be terminating semicolon
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate TAXLABELS command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// OPEN ISSUE: Some may object to setting newtaxa to 0 here, because then the
// fact that new taxa were specified in this DISTANCES block rather than in
// a preceding TAXA block is lost. This will only be important if we wish to
// recreate the original data file, which I don't anticipate anyone doing with
// this code (too difficult to remember all comments, the order of blocks in
// the file, etc.)
newtaxa = 0;
}
GetTreesOpts BullShell::ParseGetTrees(NexusToken& token ) const {
errormsg.clear();
if (kernel.getNumTaxa() == 0) {
errormsg="You can't get trees without an active set of taxa";
throw XBull( errormsg, token);
}
GetTreesOpts gto;
bool readingMostRecent = false;
gto.fromTree = -1;
gto.toTree = -1;
gto.mode = 7;
std::string filePref;
token.GetNextToken();
while (token.GetToken()!=";") {
if (token.Abbreviation("STOREBrlens")) {
token.GetNextToken();
gto.storeBrLensFromFile=true;
if (token.GetToken() == "=") {
token.GetNextToken();
if (token.Abbreviation("No"))
gto.storeBrLensFromFile=false;
else if (!token.Abbreviation("Yes")) {
errormsg="Expecting YES or NO after CurrentBranchLengths = option to LSCORE command";
throw XBull(errormsg, token);
}
}
}
else if (token.Abbreviation("FRom")) {
DemandEquals(token, "after From subcommand of GetTrees command");
gto.fromTree = DemandPositiveInt(token, "after From subcommand of GetTrees command");
if (gto.fromTree < 0) {
errormsg="Expecting positive integer after FROM option of GETTREES";
throw XBull(errormsg, token);
}
}
else if (token.Abbreviation("To")) {
DemandEquals(token, "after To subcommand of GetTrees command");
gto.toTree = DemandPositiveInt(token, "after To subcommand of GetTrees command");
if (gto.toTree < 0) {
errormsg="Expecting positive integer after TO option of GETTREES";
throw XBull(errormsg, token);
}
}
else if (token.Abbreviation("REplace"))
gto.mode=3;
else if (token.Abbreviation("Mode")) {
DemandEquals(token, "after Mode subcommand of GetTrees command");
gto.mode = DemandPositiveInt(token, "after Mode subcommand of GetTrees command");
if (gto.mode!=3 && gto.mode!=7) {
errormsg="Right now bull only get trees with mode 3 or mode 7";
throw XBull(errormsg, token);
}
}
else if (token.Abbreviation("FIle")) {
DemandEquals(token, "after File subcommand of GetTrees command");
token.GetNextToken();
gto.filename = token.GetTokenReference();
}
else if (token.Abbreviation("PREfix")) {
token.GetNextToken();
if (token.GetToken() == "=")
token.GetNextToken();
filePref = token.GetToken();
}
else if (token.Abbreviation("MOStrecent"))
readingMostRecent=true;
else {
errormsg = "Unrecognized subcommand (";
errormsg << token.GetTokenReference() << ") in GetTrees command.";
throw XBull(errormsg, token);
}
token.GetNextToken();
}
if (readingMostRecent) {
int highNum = GetHighestFileNum(filePref);
if (highNum < 0) {
errormsg << "Couldn't open file " << filePref << "0";
throw XBull(errormsg, token);
}
gto.filename = filePref;
gto.filename += highNum;
}
if (gto.fromTree > -1 && gto.toTree > gto.fromTree) {
errormsg << "The FromTree setting (" << gto.fromTree << ") cannot be less than the ToTree setting (" << gto.toTree << ")";
throw XBull(errormsg, token);
}
gto.fromTree = (gto.fromTree < 1 ? -1 : gto.fromTree - 1);
gto.toTree = (gto.toTree < 1 ? -1 : gto.toTree - 1);
return gto;
}
/**
* @method Read [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* This function provides the ability to read everything following the block name
* (which is read by the Nexus object) to the end or endblock statement.
* Characters are read from the input stream in. Overrides the
* abstract virtual function in the base class.
*/
void DistancesBlock::Read( NexusToken& token )
{
isEmpty = false;
token.GetNextToken(); // this should be the semicolon after the block name
if( !token.Equals(";") ) {
errormsg = "Expecting ';' after ";
errormsg += id;
errormsg += " block name, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
for(;;)
{
token.GetNextToken();
if( token.Equals("DIMENSIONS") ) {
HandleDimensionsCommand( token );
}
else if( token.Equals("FORMAT") ) {
HandleFormatCommand( token );
}
else if( token.Equals("TAXLABELS") ) {
HandleTaxlabelsCommand( token );
}
else if( token.Equals("MATRIX") ) {
HandleMatrixCommand( token );
}
else if( token.Equals("END") ) {
// get the semicolon following END
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate the END command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
break;
}
else if( token.Equals("ENDBLOCK") ) {
// get the semicolon following ENDBLOCK
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate the ENDBLOCK command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
break;
}
else {
SkippingCommand( token.GetToken() );
do {
token.GetNextToken();
} while( !token.AtEOF() && !token.Equals(";") );
if( token.AtEOF() ) {
errormsg = "Unexpected end of file encountered";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
}
}
/**
* @method HandleEndblock [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XBull
*
* Called when the END or ENDBLOCK command needs to be parsed
* from within the BullShell block. Basically just checks to make
* sure the next token in the data file is a semicolon.
*/
void BullShell::HandleEndblock( NexusToken& token ) {
token.GetNextToken();
RequireSemicolon(token, "END");
}
/**
* @method HandleFormatCommand [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Called when FORMAT command needs to be parsed from within the
* DISTANCES block. Deals with everything after the token FORMAT
* up to and including the semicolon that terminates the FORMAT
* command.
*/
void DistancesBlock::HandleFormatCommand( NexusToken& token )
{
for(;;)
{
token.GetNextToken();
// token should either be ';' or the name of a subcommand
//
if( token.Equals(";") ) {
break;
}
else if( token.Equals("TRIANGLE") ) {
// this should be the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// this should be LOWER, UPPER, or BOTH
token.GetNextToken();
if( token.Equals("LOWER") )
triangle = lower;
else if( token.Equals("UPPER") )
triangle = upper;
else if( token.Equals("BOTH") )
triangle = both;
else {
errormsg = "Expecting UPPER, LOWER, or BOTH but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
else if( token.Equals("DIAGONAL") ) {
diagonal = 1;
}
else if( token.Equals("NODIAGONAL") ) {
diagonal = 0;
}
else if( token.Equals("LABELS") ) {
labels = 1;
}
else if( token.Equals("NOLABELS") ) {
labels = 0;
}
else if( token.Equals("INTERLEAVE") ) {
interleave = 1;
}
else if( token.Equals("NOINTERLEAVE") ) {
interleave = 0;
}
else if( token.Equals("MISSING") ) {
// this should be the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// this should be the missing data symbol
token.GetNextToken();
if( token.GetTokenLength() != 1 ) {
errormsg = "Missing data symbol specified (";
errormsg += token.GetToken();
errormsg += ") is invalid (must be a single character)";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
missing = token.GetToken()[0];
}
else {
errormsg = "Token specified (";
errormsg += token.GetToken();
errormsg += ") is an invalid subcommand for the FORMAT command";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
}
/// Throws an XBull excpetion if token is not ";"
void BullShell::RequireSemicolon(NexusToken & token, const char *cmd) const {
if ( !token.Equals(";"))
ThrowNoSemicolon(token, cmd);
}
void SkipEqualsIfNext(NexusToken & token)
{
token.GetNextToken();
if (token.Equals("="))
token.GetNextToken();
}
void MyCharactersBlock::HandleMatrix (NexusToken& token)
{
int i, j;
if (ntax == 0)
{
errormsg = "Must precede ";
errormsg += id;
errormsg += " block with a TAXA block or specify NEWTAXA and NTAX in the DIMENSIONS command";
throw XNexus (errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn());
}
if (ntaxTotal == 0)
ntaxTotal = taxa.GetNumTaxonLabels();
// We use >= rather than just > below because someone might have
// ELIMINATEd all characters, and we should allow that (even though it
// is absurd)
assert (nchar >= 0);
if (matrix != NULL)
delete matrix;
mContData.clear();
if (GetDataType() != continuous)
{
matrix = new DiscreteMatrix (ntax, nchar);
}
// Allocate memory for (and initialize) the arrays activeTaxon and
// activeChar.
// All characters and all taxa are initially active.
activeTaxon = new bool[ntax];
for (i = 0; i < ntax; i++)
activeTaxon[i] = true;
activeChar = new bool[nchar];
for (j = 0; j < nchar; j++)
activeChar[j] = true;
// The value of ncharTotal is normally identical to the value of nchar
// specified in the CHARACTERS block DIMENSIONS command. If an
// ELIMINATE command is processed, however, nchar < ncharTotal. Note
// that the ELIMINATE command will have already been read by now, and
// the ELIMINATEd character numbers will be stored in the IntSet
// eliminated.
//
// Note that if an ELIMINATE command has been read, charPos will have
// already been created; thus, we only need to allocate and initialize
// charPos if user did not specify an ELIMINATE command
if (charPos == NULL)
BuildCharPosArray();
// The value of ntaxTotal equals the number of taxa specified in the
// TAXA block, whereas ntax equals the number of taxa specified in
// the DIMENSIONS command of the CHARACTERS block. These two numbers
// will be identical unless some taxa were left out of the MATRIX
// command of the CHARACTERS block, in which case ntax < ntaxTotal.
if (taxonPos != NULL)
delete [] taxonPos;
taxonPos = new int[ntaxTotal];
for (i = 0; i < ntaxTotal; i++)
taxonPos[i] = -1;
if (GetDataType() == continuous)
handleContMatrix (token);
else if (transposing)
HandleTransposedMatrix (token);
else
HandleStdMatrix (token);
// If we've gotten this far, presumably it is safe to
// tell the ASSUMPTIONS block that were ready to take on
// the responsibility of being the current character-containing
// block (to be consulted if characters are excluded or included
// or if taxa are deleted or restored)
assumptionsBlock.SetCallback(this);
// this should be the terminating semicolon at the end of the matrix command
if (GetDataType() != continuous)
token.GetNextToken();
if (!token.Equals(";")) {
errormsg = "Expecting ';' at the end of the MATRIX command; found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus (errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn());
}
}
/**
* @method Execute [void:public]
* @param token [NexusToken&] the token object used to grab Nexus tokens
* @param notifyStartStop [bool] if true, ExecuteStarting and ExecuteStopping will be called (true by default)
*
* Reads the Nexus data file from the input stream provided by token. This function
* is responsible for reading through the name of a each block. Once
* it has read a block name, it searches blockList for a block object
* to handle reading the remainder of the block's contents. The block
* object is responsible for reading the end or endblock command as well
* as the trailing semicolon. This function also handles reading
* comments that are outside of blocks, as well as the initial #NEXUS
* keyword. The notifyStartStop argument is provided in case you do not
* wish the ExecuteStart and ExecuteStop functions to be called. These functions
* are primarily used for creating and destroying a dialog box to show progress,
* and nested Execute calls can thus cause problems (e.g., a dialog box is
* destroyed when the inner Execute calls ExecuteStop and the outer Execute
* still expects the dialog box to be available). Specifying notifyStartStop
* false for all the nested Execute calls thus allows the outermost Execute call
* to control creation and destruction of the dialog box.
*/
void Nexus::Execute( NexusToken& token, bool notifyStartStop /* = true */ )
{
char id_str[256];
bool disabledBlock = false;
nxsstring errormsg;
try {
token.GetNextToken();
}
catch( XNexus x ) {
NexusError( token.errormsg, 0, 0, 0 );
return;
}
if( !token.Equals("#NEXUS") ) {
errormsg = "Expecting #NEXUS to be the first token in the file, but found ";
errormsg += token.GetToken();
errormsg += " instead";
NexusError( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
return;
}
if( notifyStartStop )
ExecuteStarting();
for(;;)
{
token.SetLabileFlagBit( NexusToken::saveCommandComments );
token.GetNextToken();
if( token.AtEOF() )
break;
if( token.Equals("BEGIN") )
{
disabledBlock = false;
token.GetNextToken();
NexusBlock* curr;
for( curr = blockList; curr != NULL; curr = curr->next )
{
if( !token.Equals( curr->GetID() ) )
continue;
if( !curr->IsEnabled() ) {
disabledBlock = true;
SkippingDisabledBlock( token.GetToken() );
continue;
}
strcpy( id_str, curr->GetID().c_str() );
EnteringBlock( id_str /*curr->GetID()*/ );
curr->Reset();
try {
curr->Read( token );
}
catch( XNexus x ) {
NexusError( curr->errormsg, x.pos, x.line, x.col );
curr->Reset();
return;
}
ExitingBlock( id_str /*curr->GetID()*/ );
break;
}
if( curr == NULL )
{
token.BlanksToUnderscores();
nxsstring currBlock = token.GetToken();
if( !disabledBlock )
SkippingBlock( currBlock );
for(;;)
{
token.GetNextToken();
if( token.Equals("END") || token.Equals("ENDBLOCK") ) {
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' after END or ENDBLOCK command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
NexusError( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
return;
}
break;
}
if( token.AtEOF() ) {
errormsg = "Encountered end of file before END or ENDBLOCK in block ";
errormsg += currBlock;
NexusError( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
return;
}
}
} // if token not found amongst known block IDs
} // if token equals BEGIN
else if( token.Equals("&SHOWALL") ) {
NexusBlock* curr;
for( curr = blockList; curr != NULL; curr = curr->next )
{
DebugReportBlock(*curr);
}
}
else if( token.Equals("&LEAVE") ) {
break;
}
} // for(;;)
if( notifyStartStop )
ExecuteStopping();
}
void BullShell::ParseCodLikeStartValCommand(CodLikeStartOpts &opts, NexusToken& token) const
{
opts.aaFreqs.clear();
DblVector blankRow(21, 1.0/20.0);
blankRow[20] = 0.0; //zero out the freq of the stop codon.
opts.genetic_code = GenCode(MITO);
token.GetNextToken();
opts.gtrParams.assign(N_MUT_PARAMS, .25);
while (token.GetToken()!=";") {
if (token.Abbreviation("CURREntbranchlengths") || token.Abbreviation("CURRBranchlengths")) {
token.GetNextToken();
opts.currbrlen = true;
if (token.Equals("=")) {
token.GetNextToken();
if (token.Abbreviation("No"))
opts.currbrlen = false;
else if (!token.Abbreviation("Yes"))
throw XBull("Expecting YES or NO after CurrentBranchLengths = option to CodLikeStartVal command");
token.GetNextToken();
}
}
else if (token.Abbreviation("GEneticCode")) {
SkipEqualsIfNext(token);
if (token.Abbreviation("Mitochondrial"))
opts.genetic_code = GenCode(MITO);
else if (token.Abbreviation("Nuclear"))
opts.genetic_code = GenCode(NUCLEAR);
else
throw XBull("Expecting either Mito or Nuclear code");
}
else if (token.Abbreviation("BASEfreq")) {
SkipEqualsIfNext(token);
if (!token.Equals("("))
throw XBull("Expecting ( after basefreq option to CodLikeStartVal command");
for (unsigned i = 0; i < 3; i++) {
token.GetNextToken();
opts.gtrParams[i] = atof(token.GetToken().c_str());
if (opts.gtrParams[i] <= 0.0)
throw XBull("basefreqs must be > 0.0 CodLikeStartVal command");
}
if (opts.gtrParams[1] + opts.gtrParams[2] + opts.gtrParams[0] >= 1.0)
throw XBull("Sum of A, C, and G must be <1.0 CodLikeStartVal command");
token.GetNextToken();
if (!token.Equals(")"))
throw XBull("Expecting ) after basefreq option to CodLikeStartVal command");
}
else if (token.Abbreviation("RMATrix")) {
SkipEqualsIfNext(token);
if (!token.Equals("("))
throw XBull("Expecting ( after rmatrix option to CodLikeStartVal command");
for (int i = 0; i < 6; i++) {
token.GetNextToken();
opts.gtrParams[3 + i] = atof(token.GetToken().c_str());
if (opts.gtrParams[3 + i] <= 0.0)
throw XBull("rmatrix must be >0.0 CodLikeStartVal command");
}
token.GetNextToken();
if (!token.Equals(")"))
throw XBull("Expecting ) after rmatrix option to CodLikeStartVal command");
}
# ifdef ALLOWMULTIHITS
else if (token.Abbreviation("DOublehit")) { //not a great name, but allows multiplier abbreviation to work
SkipEqualsIfNext(token);
opts.gtrParams[MULTI_HIT_PARAM_INDEX] = atof(token.GetToken().c_str());
if (opts.gtrParams[MULTI_HIT_PARAM_INDEX] < 0.0)
throw XBull("DOublehit (MultipleHitProb) must be >= 0.0 CodLikeStartVal command");
}
# endif
else if (token.Abbreviation("AAFreq")) {
SkipEqualsIfNext(token);
if (!token.Equals("("))
throw XBull("Expecting ( after aafreq option to CodLikeStartVal command");
token.GetNextToken();
long tmp;
if (!NxsString::to_long(token.GetTokenAsCStr(), &tmp))
throw XBull("Expecting number of codons after aafreq=( option to CodLikeStartVal command");
if (tmp < 1)
throw XBull("number of codons must be >0 in CodLikeStartVal command");
unsigned naa = (unsigned) tmp;
opts.aaFreqs.assign(naa, blankRow);
for (unsigned ii = 0; ii < naa; ii++) {
token.GetNextToken();
if (!token.Equals("("))
throw XBull("Expecting ( for next site in aafreq option to CodLikeStartVal command", token);
DblVector & row = opts.aaFreqs[ii];
for (int j=0; j < 20; j++) {
token.SetLabileFlagBit(NxsToken::hyphenNotPunctuation);
token.GetNextToken();
row[j] = atof(token.GetToken().c_str());
if (row[j]<0.0)
throw XBull("each amino acid freq must be >0.0 CodLikeStartVal command", token);
}
token.GetNextToken();
if (!token.Equals(")"))
throw XBull("Expecting ) for next site in aafreq option to CodLikeStartVal command", token);
}
token.GetNextToken();
if (!token.Equals(")"))
throw XBull("Expecting ) after aafreq option to CodLikeStartVal command");
}
else if (token.Abbreviation("MUlt")) {
SkipEqualsIfNext(token);
if (!token.Equals("("))
throw XBull("Expecting ( after mult option to CodLikeStartVal command");
if (opts.aaFreqs.size() < 1)
throw XBull("number of codons must be specified (in the AAFreq subcommand) before the mult option is used in CodLikeStartVal command");
opts.multipliers.assign(opts.aaFreqs.size(), 1.0);
for (unsigned ii = 0; ii < opts.aaFreqs.size(); ii++) {
token.GetNextToken();
opts.multipliers[ii] = atof(token.GetToken().c_str());
if (opts.multipliers[ii] < 0.0)
throw XBull("multiplier must be > 0.0 CodLikeStartVal command");
}
token.GetNextToken();
if (!token.Equals(")"))
throw XBull("Expecting ) after mult option to CodLikeStartVal command");
}
else if (token.Abbreviation("TReescale")) {
SkipEqualsIfNext(token);
opts.treeScale = atof(token.GetToken().c_str());
if (opts.treeScale < 0.0)
throw XBull("The Tree scaling must be >0.0 CodLikeStartVal command");
}
token.GetNextToken();
}
}
/**
* @method Read [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* This function provides the ability to read everything following
* the block name (which is read by the Nexus object) to the end or
* endblock statement. Characters are read from the input stream
* in. Overrides the pure virtual function in the base class.
*/
void AssumptionsBlock::Read( NexusToken& token )
{
isEmpty = false;
// this should be the semicolon after the block name
//
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' after ";
errormsg += id;
errormsg += " block name, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
for(;;)
{
token.GetNextToken();
if( token.Equals("EXSET") ) {
HandleExset( token );
}
else if( token.Equals("TAXSET") ) {
HandleTaxset( token );
}
else if( token.Equals("CHARSET") ) {
HandleCharset( token );
}
else if( token.Equals("END") ) {
HandleEndblock( token );
break;
}
else if( token.Equals("ENDBLOCK") ) {
HandleEndblock( token );
break;
}
else
{
SkippingCommand( token.GetToken() );
do {
token.GetNextToken();
} while( !token.AtEOF() && !token.Equals(";") );
if( token.AtEOF() ) {
errormsg = "Unexpected end of file encountered";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
}
}
/**
* @method Read [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* This function provides the ability to read everything following the block name
* (which is read by the Nexus object) to the end or endblock statement.
* Characters are read from the input stream in. Overrides the
* abstract virtual function in the base class.
*/
void TaxaBlock::Read( NexusToken& token )
{
isEmpty = false;
token.GetNextToken(); // this should be the semicolon after the block name
if( !token.Equals(";") ) {
errormsg = "Expecting ';' after TAXA block name, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
for(;;)
{
token.GetNextToken();
if( token.Equals("DIMENSIONS") ) {
token.GetNextToken(); // this should be the NTAX keyword
if( !token.Equals("NTAX") ) {
errormsg = "Expecting NTAX keyword, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
token.GetNextToken(); // this should be the equals sign
if( !token.Equals("=") ) {
errormsg = "Expecting '=', but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
token.GetNextToken(); // this should be the number of taxa
ntax = atoi( token.GetToken().c_str() );
if( ntax <= 0 ) {
errormsg = "NTAX should be greater than zero (";
errormsg += token.GetToken();
errormsg += " was specified)";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
token.GetNextToken(); // this should be the terminating semicolon
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate DIMENSIONS command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
else if( token.Equals("TAXLABELS") ) {
if( ntax <= 0 ) {
errormsg = "NTAX must be specified before TAXLABELS command";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
for( int i = 0; i < ntax; i++ ) {
token.GetNextToken();
taxonLabels.push_back( token.GetToken() );
}
token.GetNextToken(); // this should be terminating semicolon
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate TAXLABELS command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
else if( token.Equals("END") ) {
// get the semicolon following END
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate the END command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
break;
}
else if( token.Equals("ENDBLOCK") ) {
// get the semicolon following ENDBLOCK
token.GetNextToken();
if( !token.Equals(";") ) {
errormsg = "Expecting ';' to terminate the ENDBLOCK command, but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
break;
}
else {
SkippingCommand( token.GetToken() );
do {
token.GetNextToken();
} while( !token.AtEOF() && !token.Equals(";") );
if( token.AtEOF() ) {
errormsg = "Unexpected end of file encountered";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
}
}
}
/**
* @method HandleDimensionsCommand [void:protected]
* @param token [NexusToken&] the token used to read from in
* @throws XNexus
*
* Called when DIMENSIONS command needs to be parsed from within the
* DISTANCES block. Deals with everything after the token DIMENSIONS
* up to and including the semicolon that terminates the DIMENSIONS
* command.
*/
void DistancesBlock::HandleDimensionsCommand( NexusToken& token )
{
for(;;)
{
token.GetNextToken();
// token should either be ';' or the name of a subcommand
//
if( token.Equals(";") ) {
break;
}
else if( token.Equals("NEWTAXA") ) {
ntax = 0;
newtaxa = 1;
}
else if( token.Equals("NTAX") ) {
if( !newtaxa ) {
errormsg = "Must specify NEWTAXA before NTAX if new taxa are being defined";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// this should be the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// this should be the number of taxa
token.GetNextToken();
ntax = atoi( token.GetToken().c_str() );
}
else if( token.Equals("NCHAR") ) {
// this should be the equals sign
token.GetNextToken();
if( !token.Equals("=") ) {
errormsg = "Expecting '=' but found ";
errormsg += token.GetToken();
errormsg += " instead";
throw XNexus( errormsg, token.GetFilePosition(), token.GetFileLine(), token.GetFileColumn() );
}
// this should be the number of characters
token.GetNextToken();
nchar = atoi( token.GetToken().c_str() );
}
}
if( ntax == 0 )
ntax = taxa.GetNumTaxonLabels();
}
