added changes for the readsTraversal to accomidate design changes

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@553 348d0f76-0448-11de-a6fe-93d51630548a

java/src/org/broadinstitute/sting/gatk/dataSources/shards/ReadShard.java

@@ -31,22 +31,24 @@ import org.broadinstitute.sting.utils.GenomeLoc;
 public class ReadShard implements Shard {
 
     // the count of the reads we want to copy off
-    int size = 0;
+    private int size = 0;
 
     // this is going to get gross
     private final ReadShardStrategy str;
 
     /**
      * create a read shard, given a read size
+     *
      * @param size
      */
     public ReadShard(int size) {
         this.str = null;
-        this.size = size;     
+        this.size = size;
     }
 
     /**
      * create a read shard, given a read size
+     *
      * @param size
      */
     ReadShard(ReadShardStrategy caller, int size) {
@@ -56,9 +58,15 @@ public class ReadShard implements Shard {
 
     /** @return the genome location represented by this shard */
     public GenomeLoc getGenomeLoc() {
-        return null;  //To change body of implemented methods use File | Settings | File Templates.
+        throw new UnsupportedOperationException("Reads based sharding isn't genome loc aware");
     }
 
+    /** @return the genome location represented by this shard */
+    public int getSize() {
+        return size;
+    }
+
+
     /**
      * what kind of shard do we return
      *

java/src/org/broadinstitute/sting/gatk/dataSources/shards/ReadShardStrategy.java

@@ -36,7 +36,7 @@ public class ReadShardStrategy implements ShardStrategy {
     private boolean unMappedReads = true;
 
     // our read bucket size, default
-    public long readCount = 100000;
+    protected long readCount = 100000L;
 
     // our sequence dictionary
     final private SAMSequenceDictionary dic;
@@ -60,7 +60,7 @@ public class ReadShardStrategy implements ShardStrategy {
     }
 
     public Shard next() {
-        return null;  //To change body of implemented methods use File | Settings | File Templates.
+        return new ReadShard((int)readCount);  //To change body of implemented methods use File | Settings | File Templates.
     }
 
     public void remove() {

java/src/org/broadinstitute/sting/gatk/dataSources/shards/ShardStrategyFactory.java

@@ -56,6 +56,8 @@ public class ShardStrategyFactory {
                 return new LinearLocusShardStrategy(dic, startingSize);
             case EXPONENTIAL:
                 return new ExpGrowthLocusShardStrategy(dic, startingSize);
+            case READS:
+                return new ReadShardStrategy(dic, startingSize);
             default:
                 throw new RuntimeException("Strategy: " + strat + " isn't implemented");
         }
@@ -93,9 +95,12 @@ public class ShardStrategyFactory {
     static public ShardStrategy shatter(SHATTER_STRATEGY strat, SAMSequenceDictionary dic, long startingSize, List<GenomeLoc> lst) {
         switch (strat) {
             case LINEAR:
-                return new LinearLocusShardStrategy(dic, startingSize , lst);
+                return new LinearLocusShardStrategy(dic, startingSize, lst);
             case EXPONENTIAL:
-                return new ExpGrowthLocusShardStrategy(dic, startingSize , lst);
+                return new ExpGrowthLocusShardStrategy(dic, startingSize, lst);
+            case READS:
+                // return new ReadShardStrategy(dic, startingSize);
+                throw new RuntimeException("Strategy: " + strat + " isn't implemented for intervals");
             default:
                 throw new RuntimeException("Strategy: " + strat + " isn't implemented");
         }
@@ -109,7 +114,7 @@ public class ShardStrategyFactory {
      * @return
      */
     static public ShardStrategy shatterByReadCount(SAMSequenceDictionary dic, long readCount) {
-        return null;
+        return new ReadShardStrategy(dic, readCount);
     }
 
 }

java/src/org/broadinstitute/sting/gatk/traversals/TraverseReads.java

@@ -4,9 +4,9 @@ import net.sf.samtools.SAMRecord;
 import org.apache.log4j.Logger;
 import org.broadinstitute.sting.gatk.LocusContext;
 import org.broadinstitute.sting.gatk.dataSources.providers.LocusContextProvider;
+import org.broadinstitute.sting.gatk.dataSources.shards.ReadShard;
 import org.broadinstitute.sting.gatk.dataSources.shards.Shard;
 import org.broadinstitute.sting.gatk.iterators.BoundedReadIterator;
-import org.broadinstitute.sting.gatk.refdata.RefMetaDataTracker;
 import org.broadinstitute.sting.gatk.refdata.ReferenceOrderedData;
 import org.broadinstitute.sting.gatk.refdata.ReferenceOrderedDatum;
 import org.broadinstitute.sting.gatk.walkers.ReadWalker;
@@ -15,6 +15,7 @@ import org.broadinstitute.sting.utils.GenomeLoc;
 
 import java.io.File;
 import java.util.List;
+import java.util.Arrays;
 
 /**
  *
@@ -77,13 +78,14 @@ public class TraverseReads extends TraversalEngine {
                              BoundedReadIterator iter,
                              T sum) {
 
-        logger.debug(String.format("TraverseReads.traverse Genomic interval is %s", shard.getGenomeLoc()));
+        logger.debug(String.format("TraverseReads.traverse Genomic interval is %s", ((ReadShard)shard).getSize()));
 
         if (!(walker instanceof ReadWalker))
             throw new IllegalArgumentException("Walker isn't a read walker!");
 
         ReadWalker<M, T> readWalker = (ReadWalker<M, T>) walker;
-        GenomeLoc loc = shard.getGenomeLoc();
+
+        int readCNT = 0;
 
         // while we still have more reads
         for (SAMRecord read: iter) {
@@ -91,19 +93,15 @@ public class TraverseReads extends TraversalEngine {
             // get the genome loc from the read
             GenomeLoc site = new GenomeLoc(read);
 
+            // Jump forward in the reference to this locus location
+            LocusContext locus = new LocusContext(site, Arrays.asList(read), Arrays.asList(0));
+
             // update the number of reads we've seen
             TraversalStatistics.nRecords++;
 
-            // Iterate forward to get all reference ordered data covering this locus
+            // we still have to fix the locus context provider to take care of this problem with > 1 length contexts
-            final RefMetaDataTracker tracker = getReferenceOrderedDataAtLocus(site);
+            // LocusContext locus = locusProvider.getLocusContext(site);
-            //ReferenceIterator refSite = referenceProvider.getReferenceSequence(site);
+            
-                             
-            LocusContext locus = locusProvider.getLocusContext(site);
-            //locus.setReferenceContig(refSite.getCurrentContig());
-
-            if (DOWNSAMPLE_BY_COVERAGE)
-                locus.downsampleToCoverage(downsamplingCoverage);
-
             final boolean keepMeP = readWalker.filter(locus, read);
             if (keepMeP) {
                 M x = readWalker.map(locus, read);