zzh
/
gatk-3.8
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix for RefSeqCodec bug and better error messages 

-- RefSeqCodec bug: getFeatureClass() returned RefSeqCodec.class, not RefSeqFeature.class.  Really should change this in Tribble to require Class<T extends Feature> to get compile time type checking
-- Better error messages that actually list the available tribble types, when there's a type error
This commit is contained in:
Mark DePristo 2011-08-17 16:22:07 -04:00
parent 53006da9a5
commit d59e6ed274
3 changed files with 119 additions and 119 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
public/java/src/org/broadinstitute/sting/commandline/ArgumentTypeDescriptor.java
View File

@ -373,16 +373,16 @@ class RodBindingArgumentTypeDescriptor extends ArgumentTypeDescriptor {
if ( featureDescriptor != null ) { if ( featureDescriptor != null ) {
tribbleType = featureDescriptor.getName(); tribbleType = featureDescriptor.getName();
logger.warn("Dynamically determined type of " + file + " to be " + tribbleType); logger.warn("Dynamically determined type of " + file + " to be " + tribbleType);
} else {
throw new UserException.CommandLineException(
String.format("No tribble type was provided on the command line and the type of the file could not be determined dynamically. " +
"Please add an explicit type tag :TYPE listing the correct type from among the supported types: %s",
manager.userFriendlyListOfAvailableFeatures()));
} }
} }
} }
} }
if ( tribbleType == null ) // error handling
throw new UserException.CommandLineException(
String.format("Could not parse argument %s with value %s",
defaultDefinition.fullName, value));
Constructor ctor = (makeRawTypeIfNecessary(type)).getConstructor(Class.class, String.class, String.class, String.class, Tags.class); Constructor ctor = (makeRawTypeIfNecessary(type)).getConstructor(Class.class, String.class, String.class, String.class, Tags.class);
Class parameterType = getParameterizedTypeClass(type); Class parameterType = getParameterizedTypeClass(type);
RodBinding result = (RodBinding)ctor.newInstance(parameterType, name, value, tribbleType, tags); RodBinding result = (RodBinding)ctor.newInstance(parameterType, name, value, tribbleType, tags);
@ -395,8 +395,8 @@ class RodBindingArgumentTypeDescriptor extends ArgumentTypeDescriptor {
value, source.field.getName())); value, source.field.getName()));
} catch (Exception e) { } catch (Exception e) {
throw new UserException.CommandLineException( throw new UserException.CommandLineException(
String.format("Failed to parse value %s for argument %s.", String.format("Failed to parse value %s for argument %s. Message: %s",
value, source.field.getName())); value, source.field.getName(), e.getMessage()));
} }
} }

220
public/java/src/org/broadinstitute/sting/gatk/refdata/features/refseq/RefSeqCodec.java
View File

@ -1,110 +1,110 @@
package org.broadinstitute.sting.gatk.refdata.features.refseq; package org.broadinstitute.sting.gatk.refdata.features.refseq;
import org.apache.commons.io.filefilter.FalseFileFilter; import org.apache.commons.io.filefilter.FalseFileFilter;
import org.broad.tribble.Feature; import org.broad.tribble.Feature;
import org.broad.tribble.TribbleException; import org.broad.tribble.TribbleException;
import org.broad.tribble.readers.LineReader; import org.broad.tribble.readers.LineReader;
import org.broadinstitute.sting.gatk.refdata.ReferenceDependentFeatureCodec; import org.broadinstitute.sting.gatk.refdata.ReferenceDependentFeatureCodec;
import org.broadinstitute.sting.utils.GenomeLoc; import org.broadinstitute.sting.utils.GenomeLoc;
import org.broadinstitute.sting.utils.GenomeLocParser; import org.broadinstitute.sting.utils.GenomeLocParser;
import org.broadinstitute.sting.utils.Utils; import org.broadinstitute.sting.utils.Utils;
import org.broadinstitute.sting.utils.exceptions.UserException; import org.broadinstitute.sting.utils.exceptions.UserException;
import java.util.ArrayList; import java.util.ArrayList;
/** /**
* the ref seq codec * the ref seq codec
*/ */
public class RefSeqCodec implements ReferenceDependentFeatureCodec<RefSeqFeature> { public class RefSeqCodec implements ReferenceDependentFeatureCodec<RefSeqFeature> {
/** /**
* The parser to use when resolving genome-wide locations. * The parser to use when resolving genome-wide locations.
*/ */
private GenomeLocParser genomeLocParser; private GenomeLocParser genomeLocParser;
private boolean zero_coding_length_user_warned = false; private boolean zero_coding_length_user_warned = false;
/** /**
* Set the parser to use when resolving genetic data. * Set the parser to use when resolving genetic data.
* @param genomeLocParser The supplied parser. * @param genomeLocParser The supplied parser.
*/ */
@Override @Override
public void setGenomeLocParser(GenomeLocParser genomeLocParser) { public void setGenomeLocParser(GenomeLocParser genomeLocParser) {
this.genomeLocParser = genomeLocParser; this.genomeLocParser = genomeLocParser;
} }
@Override @Override
public Feature decodeLoc(String line) { public Feature decodeLoc(String line) {
if (line.startsWith("#")) return null; if (line.startsWith("#")) return null;
String fields[] = line.split("\t"); String fields[] = line.split("\t");
if (fields.length < 3) throw new TribbleException("RefSeq (decodeLoc) : Unable to parse line -> " + line + ", we expected at least 3 columns, we saw " + fields.length); if (fields.length < 3) throw new TribbleException("RefSeq (decodeLoc) : Unable to parse line -> " + line + ", we expected at least 3 columns, we saw " + fields.length);
String contig_name = fields[2]; String contig_name = fields[2];
try { try {
return new RefSeqFeature(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5]))); return new RefSeqFeature(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5])));
} catch ( UserException.MalformedGenomeLoc e ) { } catch ( UserException.MalformedGenomeLoc e ) {
Utils.warnUser("RefSeq file is potentially incorrect, as some transcripts or exons have a negative length ("+fields[2]+")"); Utils.warnUser("RefSeq file is potentially incorrect, as some transcripts or exons have a negative length ("+fields[2]+")");
return null; return null;
} }
} }
/** Fills this object from a text line in RefSeq (UCSC) text dump file */ /** Fills this object from a text line in RefSeq (UCSC) text dump file */
@Override @Override
public RefSeqFeature decode(String line) { public RefSeqFeature decode(String line) {
if (line.startsWith("#")) return null; if (line.startsWith("#")) return null;
String fields[] = line.split("\t"); String fields[] = line.split("\t");
// we reference postion 15 in the split array below, make sure we have at least that many columns // we reference postion 15 in the split array below, make sure we have at least that many columns
if (fields.length < 16) throw new TribbleException("RefSeq (decode) : Unable to parse line -> " + line + ", we expected at least 16 columns, we saw " + fields.length); if (fields.length < 16) throw new TribbleException("RefSeq (decode) : Unable to parse line -> " + line + ", we expected at least 16 columns, we saw " + fields.length);
String contig_name = fields[2]; String contig_name = fields[2];
RefSeqFeature feature = new RefSeqFeature(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5]))); RefSeqFeature feature = new RefSeqFeature(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5])));
feature.setTranscript_id(fields[1]); feature.setTranscript_id(fields[1]);
if ( fields[3].length()==1 && fields[3].charAt(0)=='+') feature.setStrand(1); if ( fields[3].length()==1 && fields[3].charAt(0)=='+') feature.setStrand(1);
else if ( fields[3].length()==1 && fields[3].charAt(0)=='-') feature.setStrand(-1); else if ( fields[3].length()==1 && fields[3].charAt(0)=='-') feature.setStrand(-1);
else throw new UserException.MalformedFile("Expected strand symbol (+/-), found: "+fields[3] + " for line=" + line); else throw new UserException.MalformedFile("Expected strand symbol (+/-), found: "+fields[3] + " for line=" + line);
int coding_start = Integer.parseInt(fields[6])+1; int coding_start = Integer.parseInt(fields[6])+1;
int coding_stop = Integer.parseInt(fields[7]); int coding_stop = Integer.parseInt(fields[7]);
if ( coding_start > coding_stop ) { if ( coding_start > coding_stop ) {
if ( ! zero_coding_length_user_warned ) { if ( ! zero_coding_length_user_warned ) {
Utils.warnUser("RefSeq file contains transcripts with zero coding length. "+ Utils.warnUser("RefSeq file contains transcripts with zero coding length. "+
"Such transcripts will be ignored (this warning is printed only once)"); "Such transcripts will be ignored (this warning is printed only once)");
zero_coding_length_user_warned = true; zero_coding_length_user_warned = true;
} }
return null; return null;
} }
feature.setTranscript_interval(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5]))); feature.setTranscript_interval(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(fields[4])+1, Integer.parseInt(fields[5])));
feature.setTranscript_coding_interval(genomeLocParser.createGenomeLoc(contig_name, coding_start, coding_stop)); feature.setTranscript_coding_interval(genomeLocParser.createGenomeLoc(contig_name, coding_start, coding_stop));
feature.setGene_name(fields[12]); feature.setGene_name(fields[12]);
String[] exon_starts = fields[9].split(","); String[] exon_starts = fields[9].split(",");
String[] exon_stops = fields[10].split(","); String[] exon_stops = fields[10].split(",");
String[] eframes = fields[15].split(","); String[] eframes = fields[15].split(",");
if ( exon_starts.length != exon_stops.length ) if ( exon_starts.length != exon_stops.length )
throw new UserException.MalformedFile("Data format error: numbers of exon start and stop positions differ for line=" + line); throw new UserException.MalformedFile("Data format error: numbers of exon start and stop positions differ for line=" + line);
if ( exon_starts.length != eframes.length ) if ( exon_starts.length != eframes.length )
throw new UserException.MalformedFile("Data format error: numbers of exons and exon frameshifts differ for line=" + line); throw new UserException.MalformedFile("Data format error: numbers of exons and exon frameshifts differ for line=" + line);
ArrayList<GenomeLoc> exons = new ArrayList<GenomeLoc>(exon_starts.length); ArrayList<GenomeLoc> exons = new ArrayList<GenomeLoc>(exon_starts.length);
ArrayList<Integer> exon_frames = new ArrayList<Integer>(eframes.length); ArrayList<Integer> exon_frames = new ArrayList<Integer>(eframes.length);
for ( int i = 0 ; i < exon_starts.length ; i++ ) { for ( int i = 0 ; i < exon_starts.length ; i++ ) {
exons.add(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(exon_starts[i])+1, Integer.parseInt(exon_stops[i]) ) ); exons.add(genomeLocParser.createGenomeLoc(contig_name, Integer.parseInt(exon_starts[i])+1, Integer.parseInt(exon_stops[i]) ) );
exon_frames.add(Integer.decode(eframes[i])); exon_frames.add(Integer.decode(eframes[i]));
} }
feature.setExons(exons); feature.setExons(exons);
feature.setExon_frames(exon_frames); feature.setExon_frames(exon_frames);
return feature; return feature;
} }
@Override @Override
public Object readHeader(LineReader reader) { public Object readHeader(LineReader reader) {
return null; return null;
} }
@Override @Override
public Class getFeatureType() { public Class getFeatureType() {
return RefSeqCodec.class; return RefSeqFeature.class;
} }
} }

4
public/java/src/org/broadinstitute/sting/utils/text/ListFileUtils.java
View File

@ -160,8 +160,8 @@ public class ListFileUtils {
rodBinding.getName(), rodBinding.getTribbleType(), builderForValidation.userFriendlyListOfAvailableFeatures())); rodBinding.getName(), rodBinding.getTribbleType(), builderForValidation.userFriendlyListOfAvailableFeatures()));
if ( ! rodBinding.getType().isAssignableFrom(descriptor.getFeatureClass()) ) if ( ! rodBinding.getType().isAssignableFrom(descriptor.getFeatureClass()) )
throw new UserException.BadArgumentValue(rodBinding.getName(), throw new UserException.BadArgumentValue(rodBinding.getName(),
String.format("Field %s expected type %s, but the type of the input file provided on the command line was %s. Please make sure that you have provided the correct file type and/or that you are not binding your rod to a name matching one of the available types.", String.format("Field %s expected type %s, but the type of the input file provided on the command line was %s producing %s. Please make sure that you have provided the correct file type and/or that you are not binding your rod to a name matching one of the available types.",
rodBinding.getName(), rodBinding.getType(), descriptor.getName())); rodBinding.getName(), rodBinding.getType(), descriptor.getName(), descriptor.getFeatureClass()));
rodBindings.add(triplet); rodBindings.add(triplet);