From fa8963522bc31ff35998f2c6c417f6d13db72212 Mon Sep 17 00:00:00 2001
From: asivache <asivache@348d0f76-0448-11de-a6fe-93d51630548a>
Date: Fri, 28 Jan 2011 21:44:33 +0000
Subject: [PATCH] Ignore header line if it happens to be passed to the codec
 again, instead of crashing on it

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5116 348d0f76-0448-11de-a6fe-93d51630548a
---
 .../playground/gatk/features/maf/MafCodec.java    | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/java/src/org/broadinstitute/sting/playground/gatk/features/maf/MafCodec.java b/java/src/org/broadinstitute/sting/playground/gatk/features/maf/MafCodec.java
index 8b45c1e97..b01baef58 100644
--- a/java/src/org/broadinstitute/sting/playground/gatk/features/maf/MafCodec.java
+++ b/java/src/org/broadinstitute/sting/playground/gatk/features/maf/MafCodec.java
@@ -102,20 +102,21 @@ public class MafCodec implements FeatureCodec {
         if (line.startsWith("#")) return null;
 
         // split the line
-        String[] tokens = line.split("\\t");
+        String[] tokens = line.split("\\t",-1);
 
         if ( expectedTokenCount == -1 ) { // do this only when we receive the first line and do not know the number of columns yet
              // we have not seen a single line yet, let's initialize the number of fields from the first line:
              expectedTokenCount = tokens.length;
+             log.info("MAF: line has "+expectedTokenCount+" fields (columns)");
              if ( expectedTokenCount == 9 ) {
                 mafType = MAF_TYPE.LITE;
                 log.info("MAF file appears to be MAF Lite");
              } else {
-                 if ( expectedTokenCount == 63 ) {
+                 if ( expectedTokenCount >= 63 ) {
                      mafType = MAF_TYPE.ANNOTATED;
                      log.info("MAF file appears to be MAF-Annotated");
                  } else {
-                     log.info("MAF file has "+expectedTokenCount +" columns, unknown file type");
+                     log.info("MAF file has "+expectedTokenCount +" columns in first line, unknown file type");
                  }
              }
              if ( line.contains("Chromosome") && line.contains("Start") && line.contains("Build")) {
@@ -135,11 +136,15 @@ public class MafCodec implements FeatureCodec {
         }
 
 
-        if (tokens.length != expectedTokenCount) {
-            log.error("MAF line contains wrong number of columns");   
+        if (tokens.length < expectedTokenCount) {
+            log.error("MAF line contains too few columns ("+tokens.length+")");
             return null;
         }
+        if (tokens.length > expectedTokenCount) {
+            log.warn("MAF line contains more columns than expected ("+tokens.length+"); extra columns discarded");
+        }
 
+        if ( tokens[CHR_COL].equals("Chromosome") ) return null; // if someone uses this codec manually and feeds it the header line multiple times...
         // create a new feature from the line:
 
         int start = Integer.valueOf(tokens[START_COL]);