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

Added some extra error checks in the VCF codec. Now that we've moved this back into the GATK, changed some of the standard exceptions to be USerErrors (instead of TribbleExceptions).

This commit is contained in:
Eric Banks 2011-07-14 11:56:10 -04:00
parent 0c54c796ed
commit 66c652d687
2 changed files with 23 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
public/java/src/org/broadinstitute/sting/utils/codecs/vcf/AbstractVCFCodec.java
View File

@ -7,6 +7,8 @@ import org.broad.tribble.NameAwareCodec;
import org.broad.tribble.TribbleException;
import org.broad.tribble.readers.LineReader;
import org.broad.tribble.util.ParsingUtils;
import org.broadinstitute.sting.utils.exceptions.ReviewedStingException;
import org.broadinstitute.sting.utils.exceptions.UserException;
import org.broadinstitute.sting.utils.variantcontext.Allele;
import org.broadinstitute.sting.utils.variantcontext.Genotype;
import org.broadinstitute.sting.utils.variantcontext.VariantContext;
@ -96,6 +98,9 @@ public abstract class AbstractVCFCodec implements FeatureCodec, NameAwareCodec,
for ( String str : headerStrings ) {
if ( !str.startsWith(VCFHeader.METADATA_INDICATOR) ) {
String[] strings = str.substring(1).split(VCFConstants.FIELD_SEPARATOR);
if ( strings.length < VCFHeader.HEADER_FIELDS.values().length )
throw new TribbleException.InvalidHeader("there are not enough columns present in the header line: " + str);
int arrayIndex = 0;
for (VCFHeader.HEADER_FIELDS field : VCFHeader.HEADER_FIELDS.values()) {
try {
@ -159,12 +164,11 @@ public abstract class AbstractVCFCodec implements FeatureCodec, NameAwareCodec,
}
private Feature reallyDecode(String line) {
try {
// the same line reader is not used for parsing the header and parsing lines, if we see a #, we've seen a header line
if (line.startsWith(VCFHeader.HEADER_INDICATOR)) return null;
// our header cannot be null, we need the genotype sample names and counts
if (header == null) throw new IllegalStateException("VCF Header cannot be null when decoding a record");
if (header == null) throw new ReviewedStingException("VCF Header cannot be null when decoding a record");
if (parts == null)
parts = new String[Math.min(header.getColumnCount(), NUM_STANDARD_FIELDS+1)];
@ -174,17 +178,18 @@ public abstract class AbstractVCFCodec implements FeatureCodec, NameAwareCodec,
// if we have don't have a header, or we have a header with no genotyping data check that we have eight columns. Otherwise check that we have nine (normal colummns + genotyping data)
if (( (header == null || (header != null && !header.hasGenotypingData())) && nParts != NUM_STANDARD_FIELDS) ||
(header != null && header.hasGenotypingData() && nParts != (NUM_STANDARD_FIELDS + 1)) )
throw new IllegalArgumentException("There aren't enough columns for line " + line + " (we expected " + (header == null ? NUM_STANDARD_FIELDS : NUM_STANDARD_FIELDS + 1) +
" tokens, and saw " + nParts + " )");
throw new UserException.MalformedVCF("there aren't enough columns for line " + line + " (we expected " + (header == null ? NUM_STANDARD_FIELDS : NUM_STANDARD_FIELDS + 1) +
" tokens, and saw " + nParts + " )", lineNo);
return parseVCFLine(parts);
} catch (TribbleException e) {
throw new TribbleException.InvalidDecodeLine(e.getMessage(), line);
}
}
protected void generateException(String message) {
throw new TribbleException.InvalidDecodeLine(message, lineNo);
throw new UserException.MalformedVCF(message, lineNo);
}
private static void generateException(String message, int lineNo) {
throw new UserException.MalformedVCF(message, lineNo);
}
/**
@ -472,10 +477,6 @@ public abstract class AbstractVCFCodec implements FeatureCodec, NameAwareCodec,
return true;
}
private static void generateException(String message, int lineNo) {
throw new TribbleException.InvalidDecodeLine(message, lineNo);
}
private static int computeForwardClipping(List<Allele> unclippedAlleles, String ref) {
boolean clipping = true;
// Note that the computation of forward clipping here is meant only to see whether there is a common

10
public/java/src/org/broadinstitute/sting/utils/exceptions/UserException.java
View File

@ -154,6 +154,16 @@ public class UserException extends ReviewedStingException {
}
}
public static class MalformedVCF extends UserException {
public MalformedVCF(String message, String line) {
super(String.format("The provided VCF file is malformed at line %s: %s", line, message));
}
public MalformedVCF(String message, int lineNo) {
super(String.format("The provided VCF file is malformed at line nmber %d: %s", lineNo, message));
}
}
public static class ReadMissingReadGroup extends MalformedBAM {
public ReadMissingReadGroup(SAMRecord read) {
super(read, String.format("Read %s is either missing the read group or its read group is not defined in the BAM header, both of which are required by the GATK. Please use http://www.broadinstitute.org/gsa/wiki/index.php/ReplaceReadGroups to fix this problem", read.getReadName()));