jpayne@68: package prok;
jpayne@68: 
jpayne@68: import java.io.PrintStream;
jpayne@68: import java.util.ArrayList;
jpayne@68: import java.util.Arrays;
jpayne@68: 
jpayne@68: import aligner.SingleStateAlignerFlat2;
jpayne@68: import fileIO.ByteFile;
jpayne@68: import fileIO.FileFormat;
jpayne@68: import fileIO.ReadWrite;
jpayne@68: import shared.Parse;
jpayne@68: import shared.Parser;
jpayne@68: import shared.PreParser;
jpayne@68: import shared.ReadStats;
jpayne@68: import shared.Shared;
jpayne@68: import shared.Timer;
jpayne@68: import shared.Tools;
jpayne@68: import stream.ConcurrentReadInputStream;
jpayne@68: import stream.ConcurrentReadOutputStream;
jpayne@68: import stream.FastaReadInputStream;
jpayne@68: import stream.Read;
jpayne@68: import structures.ListNum;
jpayne@68: import template.Accumulator;
jpayne@68: import template.ThreadWaiter;
jpayne@68: 
jpayne@68: /**
jpayne@68:  * Splits a mix of ribosomal sequences (such as Silva) into different files per type (16S, 18S, etc).
jpayne@68:  * 
jpayne@68:  * @author Brian Bushnell
jpayne@68:  * @date November 19, 2015
jpayne@68:  *
jpayne@68:  */
jpayne@68: public class SplitRibo implements Accumulator<SplitRibo.ProcessThread> {
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------        Initialization        ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/**
jpayne@68: 	 * Code entrance from the command line.
jpayne@68: 	 * @param args Command line arguments
jpayne@68: 	 */
jpayne@68: 	public static void main(String[] args){
jpayne@68: 		//Start a timer immediately upon code entrance.
jpayne@68: 		Timer t=new Timer();
jpayne@68: 		
jpayne@68: 		//Create an instance of this class
jpayne@68: 		SplitRibo x=new SplitRibo(args);
jpayne@68: 		
jpayne@68: 		//Run the object
jpayne@68: 		x.process(t);
jpayne@68: 		
jpayne@68: 		//Close the print stream if it was redirected
jpayne@68: 		Shared.closeStream(x.outstream);
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/**
jpayne@68: 	 * Constructor.
jpayne@68: 	 * @param args Command line arguments
jpayne@68: 	 */
jpayne@68: 	public SplitRibo(String[] args){
jpayne@68: 		
jpayne@68: 		{//Preparse block for help, config files, and outstream
jpayne@68: 			PreParser pp=new PreParser(args, getClass(), false);
jpayne@68: 			args=pp.args;
jpayne@68: 			outstream=pp.outstream;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		//Set shared static variables prior to parsing
jpayne@68: 		ReadWrite.USE_PIGZ=ReadWrite.USE_UNPIGZ=true;
jpayne@68: 		ReadWrite.MAX_ZIP_THREADS=Shared.threads();
jpayne@68: 		Shared.capBufferLen(50);
jpayne@68: 		ReadWrite.ZIPLEVEL=9;
jpayne@68: 		
jpayne@68: 		{//Parse the arguments
jpayne@68: 			final Parser parser=parse(args);
jpayne@68: 			Parser.processQuality();
jpayne@68: 			
jpayne@68: 			maxReads=parser.maxReads;
jpayne@68: 			overwrite=ReadStats.overwrite=parser.overwrite;
jpayne@68: 			append=ReadStats.append=parser.append;
jpayne@68: 			
jpayne@68: 			in1=parser.in1;
jpayne@68: 			qfin1=parser.qfin1;
jpayne@68: 			extin=parser.extin;
jpayne@68: 
jpayne@68: 			outPattern=parser.out1;
jpayne@68: 			extout=parser.extout;
jpayne@68: 		}
jpayne@68: 
jpayne@68: 		validateParams();
jpayne@68: 		fixExtensions(); //Add or remove .gz or .bz2 as needed
jpayne@68: 		checkFileExistence(); //Ensure files can be read and written
jpayne@68: 		checkStatics(); //Adjust file-related static fields as needed for this program 
jpayne@68: 
jpayne@68: 		//Create input FileFormat objects
jpayne@68: 		ffin1=FileFormat.testInput(in1, FileFormat.FASTQ, extin, true, true);
jpayne@68: 		
jpayne@68: 		numTypes=sequenceTypes.length;
jpayne@68: 		readsOut=new long[numTypes];
jpayne@68: 		basesOut=new long[numTypes];
jpayne@68: 		consensusSequences=loadConsensusSequenceFromFile();
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------    Initialization Helpers    ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/** Parse arguments from the command line */
jpayne@68: 	private Parser parse(String[] args){
jpayne@68: 		
jpayne@68: 		//Create a parser object
jpayne@68: 		Parser parser=new Parser();
jpayne@68: 		
jpayne@68: 		//Set any necessary Parser defaults here
jpayne@68: 		//parser.foo=bar;
jpayne@68: 		
jpayne@68: 		//Parse each argument
jpayne@68: 		for(int i=0; i<args.length; i++){
jpayne@68: 			String arg=args[i];
jpayne@68: 			
jpayne@68: 			//Break arguments into their constituent parts, in the form of "a=b"
jpayne@68: 			String[] split=arg.split("=");
jpayne@68: 			String a=split[0].toLowerCase();
jpayne@68: 			String b=split.length>1 ? split[1] : null;
jpayne@68: 			if(b!=null && b.equalsIgnoreCase("null")){b=null;}
jpayne@68: 			
jpayne@68: 			if(a.equals("verbose")){
jpayne@68: 				verbose=Parse.parseBoolean(b);
jpayne@68: 			}else if(a.equals("ordered")){
jpayne@68: 				ordered=Parse.parseBoolean(b);
jpayne@68: 			}else if(a.equalsIgnoreCase("minid")){
jpayne@68: 				minID=Float.parseFloat(b);
jpayne@68: 			}else if(a.equalsIgnoreCase("minid2") || a.equalsIgnoreCase("refineid")){
jpayne@68: 				refineID=Float.parseFloat(b);
jpayne@68: 			}else if(a.equals("out") || a.equals("pattern") || a.equals("outpattern")){
jpayne@68: 				parser.out1=b;
jpayne@68: 			}else if(a.equals("type") || a.equals("types")){
jpayne@68: 				parseTypes(b);
jpayne@68: 			}else if(a.equals("parse_flag_goes_here")){
jpayne@68: 				long fake_variable=Parse.parseKMG(b);
jpayne@68: 				//Set a variable here
jpayne@68: 			}else if(parser.parse(arg, a, b)){//Parse standard flags in the parser
jpayne@68: 				//do nothing
jpayne@68: 			}else{
jpayne@68: 				outstream.println("Unknown parameter "+args[i]);
jpayne@68: 				assert(false) : "Unknown parameter "+args[i];
jpayne@68: 			}
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		return parser;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	private void parseTypes(String b){
jpayne@68: 		sequenceTypes=null;
jpayne@68: 		if(b==null){
jpayne@68: 			assert(false) : "'types' flag requires a list of types, such as 'types=16S,18S'";
jpayne@68: 			sequenceTypes=new String[] {"Other"};
jpayne@68: 		}else{
jpayne@68: 			String[] split=b.split(",");
jpayne@68: 			sequenceTypes=new String[split.length+1];
jpayne@68: 			sequenceTypes[0]="Other";
jpayne@68: 			for(int i=0; i<split.length; i++){
jpayne@68: 				String s=split[i].replace('s', 'S');
jpayne@68: 				if(s.startsWith("its")){s=s.replaceFirst("its", "ITS");}
jpayne@68: 				sequenceTypes[i+1]=s;
jpayne@68: 			}
jpayne@68: 		}
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/** Add or remove .gz or .bz2 as needed */
jpayne@68: 	private void fixExtensions(){
jpayne@68: 		in1=Tools.fixExtension(in1);
jpayne@68: 		qfin1=Tools.fixExtension(qfin1);
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/** Ensure files can be read and written */
jpayne@68: 	private void checkFileExistence(){
jpayne@68: 		
jpayne@68: 		//Ensure input files can be read
jpayne@68: 		if(!Tools.testInputFiles(false, true, in1)){
jpayne@68: 			throw new RuntimeException("\nCan't read some input files.\n");  
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		if(outPattern==null){return;}
jpayne@68: 		
jpayne@68: 		if(!outPattern.contains("#")){
jpayne@68: 			throw new RuntimeException("OutPattern must contain '#' symbol: "+outPattern);
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		for(String type : sequenceTypes) {
jpayne@68: 			String out=outPattern.replaceFirst("#", type);
jpayne@68: 			
jpayne@68: 			//Ensure output files can be written
jpayne@68: 			if(!Tools.testOutputFiles(overwrite, append, false, out)){
jpayne@68: 				outstream.println((outPattern==null)+", "+(out==null)+", "+outPattern+", "+out);
jpayne@68: 				throw new RuntimeException("\n\noverwrite="+overwrite+"; Can't write to output file "+out+"\n");
jpayne@68: 			}
jpayne@68: 
jpayne@68: 			//Ensure that no file was specified multiple times
jpayne@68: 			if(!Tools.testForDuplicateFiles(true, in1, out)){
jpayne@68: 				throw new RuntimeException("\nSome file names were specified multiple times.\n");
jpayne@68: 			}
jpayne@68: 		}
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/** Adjust file-related static fields as needed for this program */
jpayne@68: 	private static void checkStatics(){
jpayne@68: 		//Adjust the number of threads for input file reading
jpayne@68: 		if(!ByteFile.FORCE_MODE_BF1 && !ByteFile.FORCE_MODE_BF2 && Shared.threads()>2){
jpayne@68: 			ByteFile.FORCE_MODE_BF2=true;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		assert(FastaReadInputStream.settingsOK());
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/** Ensure parameter ranges are within bounds and required parameters are set */
jpayne@68: 	private boolean validateParams(){
jpayne@68: //		assert(minfoo>0 && minfoo<=maxfoo) : minfoo+", "+maxfoo;
jpayne@68: 		if(in1==null){throw new RuntimeException("Error - at least one input file is required.");}
jpayne@68: 		return true;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	private final Read[][] loadConsensusSequenceFromFile(){
jpayne@68: 		Read[][] seqs=new Read[numTypes][];
jpayne@68: 		m16S_index=Tools.find("m16S", sequenceTypes);
jpayne@68: 		m18S_index=Tools.find("m18S", sequenceTypes);
jpayne@68: 		p16S_index=Tools.find("p16S", sequenceTypes);
jpayne@68: 		boolean stripM16S=(m16S_index>=0);
jpayne@68: 		boolean stripM18S=(m18S_index>=0);
jpayne@68: 		boolean stripP16S=(p16S_index>=0);
jpayne@68: 		for(int st=1; st<numTypes; st++){
jpayne@68: 			String name=sequenceTypes[st];
jpayne@68: 			boolean is16S=name.equalsIgnoreCase("16S");
jpayne@68: 			boolean is18S=name.equalsIgnoreCase("18S");
jpayne@68: 			seqs[st]=ProkObject.loadConsensusSequenceType(name, ((is16S && stripM16S) || (is18S && stripM18S)), (is16S && stripP16S));
jpayne@68: 		}
jpayne@68: 		return seqs;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------         Outer Methods        ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 
jpayne@68: 	/** Create read streams and process all data */
jpayne@68: 	void process(Timer t){
jpayne@68: 		
jpayne@68: 		//Turn off read validation in the input threads to increase speed
jpayne@68: 		final boolean vic=Read.VALIDATE_IN_CONSTRUCTOR;
jpayne@68: 		Read.VALIDATE_IN_CONSTRUCTOR=Shared.threads()<4;
jpayne@68: 		
jpayne@68: 		//Create a read input stream
jpayne@68: 		final ConcurrentReadInputStream cris=makeCris();
jpayne@68: 		
jpayne@68: 		//Optionally create a read output stream
jpayne@68: 		final ConcurrentReadOutputStream[] rosa=makeCrosArray();
jpayne@68: 		
jpayne@68: 		//Reset counters
jpayne@68: 		readsProcessed=0;
jpayne@68: 		basesProcessed=0;
jpayne@68: 		Arrays.fill(readsOut, 0);
jpayne@68: 		Arrays.fill(basesOut, 0);
jpayne@68: 		
jpayne@68: 		//Process the reads in separate threads
jpayne@68: 		spawnThreads(cris, rosa);
jpayne@68: 		
jpayne@68: 		if(verbose){outstream.println("Finished; closing streams.");}
jpayne@68: 		
jpayne@68: 		//Write anything that was accumulated by ReadStats
jpayne@68: 		errorState|=ReadStats.writeAll();
jpayne@68: 		//assert(!errorState);
jpayne@68: 		//Close the read streams
jpayne@68: 		errorState|=ReadWrite.closeStreams(cris, rosa);
jpayne@68: 		//assert(!errorState);
jpayne@68: 		
jpayne@68: 		//Reset read validation
jpayne@68: 		Read.VALIDATE_IN_CONSTRUCTOR=vic;
jpayne@68: 
jpayne@68: 		long readsOut2=Tools.sum(readsOut)-readsOut[0];
jpayne@68: 		long basesOut2=Tools.sum(basesOut)-basesOut[0];
jpayne@68: 		
jpayne@68: 		//Report timing and results
jpayne@68: 		t.stop();
jpayne@68: 		outstream.println(Tools.timeReadsBasesProcessed(t, readsProcessed, basesProcessed, 8));
jpayne@68: 		outstream.println(Tools.readsBasesOut(readsProcessed, basesProcessed, readsOut2, basesOut2, 8, true));
jpayne@68: 
jpayne@68: 		outstream.println();
jpayne@68: 		outstream.println(Tools.string("Type", "Count", 8));
jpayne@68: 		for(int type=0; type<numTypes; type++){
jpayne@68: 			outstream.println(Tools.number(sequenceTypes[type], readsOut[type], 8));
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		//Throw an exception of there was an error in a thread
jpayne@68: 		if(errorState){
jpayne@68: 			throw new RuntimeException(getClass().getName()+" terminated in an error state; the output may be corrupt.");
jpayne@68: 		}
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	private ConcurrentReadInputStream makeCris(){
jpayne@68: 		ConcurrentReadInputStream cris=ConcurrentReadInputStream.getReadInputStream(maxReads, true, ffin1, null, qfin1, null);
jpayne@68: 		cris.start(); //Start the stream
jpayne@68: 		if(verbose){outstream.println("Started cris");}
jpayne@68: 		return cris;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	private ConcurrentReadOutputStream[] makeCrosArray(){
jpayne@68: 		ConcurrentReadOutputStream[] rosa=new ConcurrentReadOutputStream[numTypes];
jpayne@68: 		for(int i=0; i<numTypes; i++){
jpayne@68: 			String type=sequenceTypes[i];
jpayne@68: 			final ConcurrentReadOutputStream ros=makeCros(type);
jpayne@68: 			rosa[i]=ros;
jpayne@68: 		}
jpayne@68: 		return rosa;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	private ConcurrentReadOutputStream makeCros(String type){
jpayne@68: 		if(outPattern==null){return null;}
jpayne@68: 
jpayne@68: 		//Select output buffer size based on whether it needs to be ordered
jpayne@68: 		final int buff=(ordered ? Tools.mid(2, 16, (Shared.threads()*2)/3) : 4);
jpayne@68: 		final String fname=outPattern.replaceFirst("#", type);
jpayne@68: 		FileFormat ff=FileFormat.testOutput(fname, FileFormat.FASTA, extout, true, overwrite, append, ordered);
jpayne@68: 
jpayne@68: 		final ConcurrentReadOutputStream ros=ConcurrentReadOutputStream.getStream(ff, null, buff, null, false);
jpayne@68: 		ros.start(); //Start the stream
jpayne@68: 		return ros;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------       Thread Management      ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/** Spawn process threads */
jpayne@68: 	private void spawnThreads(final ConcurrentReadInputStream cris, final ConcurrentReadOutputStream[] rosa){
jpayne@68: 		
jpayne@68: 		//Do anything necessary prior to processing
jpayne@68: 		
jpayne@68: 		//Determine how many threads may be used
jpayne@68: 		final int threads=Shared.threads();
jpayne@68: 		
jpayne@68: 		//Fill a list with ProcessThreads
jpayne@68: 		ArrayList<ProcessThread> alpt=new ArrayList<ProcessThread>(threads);
jpayne@68: 		for(int i=0; i<threads; i++){
jpayne@68: 			alpt.add(new ProcessThread(cris, rosa, i));
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		//Start the threads and wait for them to finish
jpayne@68: 		boolean success=ThreadWaiter.startAndWait(alpt, this);
jpayne@68: 		errorState&=!success;
jpayne@68: 		//assert(!errorState);
jpayne@68: 		
jpayne@68: 		//Do anything necessary after processing
jpayne@68: 		
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	@Override
jpayne@68: 	public final void accumulate(ProcessThread pt){
jpayne@68: 		readsProcessed+=pt.readsProcessedT;
jpayne@68: 		basesProcessed+=pt.basesProcessedT;
jpayne@68: 		Tools.add(readsOut, pt.readsOutT);
jpayne@68: 		Tools.add(basesOut, pt.basesOutT);
jpayne@68: 		errorState|=(!pt.success);
jpayne@68: 		//assert(!errorState);
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	@Override
jpayne@68: 	public final boolean success(){return !errorState;}
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------         Inner Methods        ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------         Inner Classes        ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/** This class is static to prevent accidental writing to shared variables.
jpayne@68: 	 * It is safe to remove the static modifier. */
jpayne@68: 	class ProcessThread extends Thread {
jpayne@68: 		
jpayne@68: 		//Constructor
jpayne@68: 		ProcessThread(final ConcurrentReadInputStream cris_, final ConcurrentReadOutputStream[] rosa_, final int tid_){
jpayne@68: 			cris=cris_;
jpayne@68: 			rosa=rosa_;
jpayne@68: 			tid=tid_;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		//Called by start()
jpayne@68: 		@Override
jpayne@68: 		public void run(){
jpayne@68: 			//Do anything necessary prior to processing
jpayne@68: 			
jpayne@68: 			//Process the reads
jpayne@68: 			processInner();
jpayne@68: 			
jpayne@68: 			//Do anything necessary after processing
jpayne@68: 			
jpayne@68: 			//Indicate successful exit status
jpayne@68: 			success=true;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		/** Iterate through the reads */
jpayne@68: 		void processInner(){
jpayne@68: 			
jpayne@68: 			//Grab the first ListNum of reads
jpayne@68: 			ListNum<Read> ln=cris.nextList();
jpayne@68: 
jpayne@68: 			//Check to ensure pairing is as expected
jpayne@68: 			if(ln!=null && !ln.isEmpty()){
jpayne@68: 				Read r=ln.get(0);
jpayne@68: 				assert(r.mate==null);
jpayne@68: 			}
jpayne@68: 
jpayne@68: 			//As long as there is a nonempty read list...
jpayne@68: 			while(ln!=null && ln.size()>0){
jpayne@68: //				if(verbose){outstream.println("Fetched "+reads.size()+" reads.");} //Disabled due to non-static access
jpayne@68: 				
jpayne@68: 				processList(ln);
jpayne@68: 				
jpayne@68: 				//Notify the input stream that the list was used
jpayne@68: 				cris.returnList(ln);
jpayne@68: //				if(verbose){outstream.println("Returned a list.");} //Disabled due to non-static access
jpayne@68: 				
jpayne@68: 				//Fetch a new list
jpayne@68: 				ln=cris.nextList();
jpayne@68: 			}
jpayne@68: 
jpayne@68: 			//Notify the input stream that the final list was used
jpayne@68: 			if(ln!=null){
jpayne@68: 				cris.returnList(ln.id, ln.list==null || ln.list.isEmpty());
jpayne@68: 			}
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		void processList(ListNum<Read> ln){
jpayne@68: 
jpayne@68: 			//Grab the actual read list from the ListNum
jpayne@68: 			final ArrayList<Read> reads=ln.list;
jpayne@68: 			
jpayne@68: 			@SuppressWarnings("unchecked")
jpayne@68: 			final ArrayList<Read>[] out=new ArrayList[numTypes];
jpayne@68: 			for(int i=0; i<numTypes; i++){
jpayne@68: 				ArrayList<Read> list=new ArrayList<Read>(50);
jpayne@68: 				out[i]=list;
jpayne@68: 			}
jpayne@68: 			
jpayne@68: 			//Loop through each read in the list
jpayne@68: 			for(int idx=0; idx<reads.size(); idx++){
jpayne@68: 				final Read r1=reads.get(idx);
jpayne@68: 				
jpayne@68: 				//Validate reads in worker threads
jpayne@68: 				if(!r1.validated()){r1.validate(true);}
jpayne@68: 
jpayne@68: 				//Track the initial length for statistics
jpayne@68: 				final int initialLength1=r1.length();
jpayne@68: 				final int initialLength2=r1.mateLength();
jpayne@68: 
jpayne@68: 				//Increment counters
jpayne@68: 				readsProcessedT+=r1.pairCount();
jpayne@68: 				basesProcessedT+=initialLength1+initialLength2;
jpayne@68: 				
jpayne@68: 				{
jpayne@68: 					//Reads are processed in this block.
jpayne@68: 					final int type=processRead(r1);
jpayne@68: 					readsOutT[type]+=r1.pairCount();
jpayne@68: 					basesOutT[type]+=r1.pairLength();
jpayne@68: 					out[type].add(r1);
jpayne@68: 				}
jpayne@68: 			}
jpayne@68: 
jpayne@68: 			//Output reads to the output stream
jpayne@68: 			if(rosa!=null){
jpayne@68: 				for(int type=0; type<numTypes; type++){
jpayne@68: 					rosa[type].add(out[type], ln.id);
jpayne@68: 				}
jpayne@68: 			}
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		/**
jpayne@68: 		 * Process a read.
jpayne@68: 		 * @param r1 Read 1
jpayne@68: 		 * @return The best-matching type, or 0 for no matches.
jpayne@68: 		 */
jpayne@68: 		private int processRead(final Read r){
jpayne@68: 			int bestType=0;
jpayne@68: 			float bestID=-1;
jpayne@68: 			for(int type=1; type<numTypes; type++){//Align to only the overall consensus
jpayne@68: 				Read[] refs=consensusSequences[type];
jpayne@68: 				float id=align(r, refs, 0, 1);
jpayne@68: 				if(id>bestID && id>=minID){
jpayne@68: 					bestType=type;
jpayne@68: 					bestID=id;
jpayne@68: 				}
jpayne@68: 			}
jpayne@68: 			if(bestType<1 || bestID<refineID || bestType==p16S_index){//If nothing met minID, or if it matched chloro, align to clade-specific consensuses
jpayne@68: 				for(int type=1; type<numTypes; type++){
jpayne@68: 					Read[] refs=consensusSequences[type];
jpayne@68: 					float id=align(r, refs, 1, refs.length);
jpayne@68: 					if(id>bestID && id>=minID){
jpayne@68: 						bestType=type;
jpayne@68: 						bestID=id;
jpayne@68: 					}
jpayne@68: 				}
jpayne@68: 			}
jpayne@68: 			r.obj=bestID;//If desired...  in actuality, more info might be useful, like alignment length
jpayne@68: 			return bestID<minID ? 0 : bestType;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		private float align(Read r, Read[] refs, int minRef, int maxRef){
jpayne@68: 			float bestID=-1;
jpayne@68: 			if(refs!=null){
jpayne@68: 				for(int i=minRef; i<maxRef; i++){
jpayne@68: 					Read ref=refs[i];
jpayne@68: 					float id=align(r.bases, ref.bases);
jpayne@68: 					bestID=Tools.max(id,  bestID);
jpayne@68: 				}
jpayne@68: 			}
jpayne@68: 			return bestID;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		private float align(byte[] query, byte[] ref){
jpayne@68: 			int a=0, b=ref.length-1;
jpayne@68: 			int[] max=ssa.fillUnlimited(query, ref, a, b, -9999);
jpayne@68: 			if(max==null){return 0;}
jpayne@68: 			
jpayne@68: 			final int rows=max[0];
jpayne@68: 			final int maxCol=max[1];
jpayne@68: 			final int maxState=max[2];
jpayne@68: 			final float id=ssa.tracebackIdentity(query, ref, a, b, rows, maxCol, maxState, null);
jpayne@68: 			return id;
jpayne@68: 		}
jpayne@68: 		
jpayne@68: 		SingleStateAlignerFlat2 ssa=new SingleStateAlignerFlat2();
jpayne@68: 
jpayne@68: 		/** Number of reads processed by this thread */
jpayne@68: 		protected long readsProcessedT=0;
jpayne@68: 		/** Number of bases processed by this thread */
jpayne@68: 		protected long basesProcessedT=0;
jpayne@68: 		
jpayne@68: 		/** Number of reads retained by this thread */
jpayne@68: 		protected long[] readsOutT=new long[numTypes];
jpayne@68: 		/** Number of bases retained by this thread */
jpayne@68: 		protected long[] basesOutT=new long[numTypes];
jpayne@68: 		
jpayne@68: 		/** True only if this thread has completed successfully */
jpayne@68: 		boolean success=false;
jpayne@68: 		
jpayne@68: 		/** Shared input stream */
jpayne@68: 		private final ConcurrentReadInputStream cris;
jpayne@68: 		/** Shared output stream */
jpayne@68: 		private final ConcurrentReadOutputStream[] rosa;
jpayne@68: 		/** Thread ID */
jpayne@68: 		final int tid;
jpayne@68: 	}
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------            Fields            ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 
jpayne@68: 	/** Primary input file path */
jpayne@68: 	private String in1=null;
jpayne@68: 	
jpayne@68: 	private String qfin1=null;
jpayne@68: 
jpayne@68: 	/** Primary output file path */
jpayne@68: 	private String outPattern=null;
jpayne@68: 	
jpayne@68: 	/** Override input file extension */
jpayne@68: 	private String extin=null;
jpayne@68: 	/** Override output file extension */
jpayne@68: 	private String extout=null;
jpayne@68: 
jpayne@68: 	float minID=0.59f; //This could be a per-type value
jpayne@68: 	float refineID=0.70f; //Refine alignment if best is less than this
jpayne@68: 	
jpayne@68: 	private int m16S_index=-2;
jpayne@68: 	private int m18S_index=-2;
jpayne@68: 	private int p16S_index=-2;
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/** Number of reads processed */
jpayne@68: 	protected long readsProcessed=0;
jpayne@68: 	/** Number of bases processed */
jpayne@68: 	protected long basesProcessed=0;
jpayne@68: 	
jpayne@68: 	/** Quit after processing this many input reads; -1 means no limit */
jpayne@68: 	private long maxReads=-1;
jpayne@68: 	
jpayne@68: 	private String[] sequenceTypes=new String[] {"Other", "16S", "18S", "23S", "5S", "m16S", "m18S", "p16S"};
jpayne@68: 	private final int numTypes;//=sequenceTypes.length;
jpayne@68: 	final Read[][] consensusSequences;
jpayne@68: 	
jpayne@68: 	/** Number of reads retained */
jpayne@68: 	final long[] readsOut;
jpayne@68: 	/** Number of bases retained */
jpayne@68: 	final long[] basesOut;
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------         Final Fields         ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 
jpayne@68: 	/** Primary input file */
jpayne@68: 	private final FileFormat ffin1;
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------         Static Fields        ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	/*----------------        Common Fields         ----------------*/
jpayne@68: 	/*--------------------------------------------------------------*/
jpayne@68: 	
jpayne@68: 	/** Print status messages to this output stream */
jpayne@68: 	private PrintStream outstream=System.err;
jpayne@68: 	/** Print verbose messages */
jpayne@68: 	public static boolean verbose=false;
jpayne@68: 	/** True if an error was encountered */
jpayne@68: 	public boolean errorState=false;
jpayne@68: 	/** Overwrite existing output files */
jpayne@68: 	private boolean overwrite=false;
jpayne@68: 	/** Append to existing output files */
jpayne@68: 	private boolean append=false;
jpayne@68: 	/** Reads are output in input order */
jpayne@68: 	private boolean ordered=true;
jpayne@68: 	
jpayne@68: }