From 26f700519ce0b9cf7b6bc2fcde1f0e8c4420a7d7 Mon Sep 17 00:00:00 2001
From: Andrewss <Andrewss@BI2043.babraham.ac.uk>
Date: Tue, 18 Aug 2020 17:35:15 +0100
Subject: [PATCH] Add CRAM reference to the options in the wrapper.

Relates to #54
---
 fastqc                                      | 19 +++++++++++++++++--
 uk/ac/babraham/FastQC/FastQCConfig.java     | 11 ++++++++++-
 uk/ac/babraham/FastQC/Sequence/BAMFile.java |  8 +++++++-
 3 files changed, 34 insertions(+), 4 deletions(-)

diff --git a/fastqc b/fastqc
index eb52c8d..c221ee1 100755
--- a/fastqc
+++ b/fastqc
@@ -88,6 +88,7 @@ if ($^O =~/darwin/) {
 my $version;
 my $help;
 my $outdir;
+my $reference;
 my $unzip;
 my $format;
 my $contaminant;
@@ -110,6 +111,7 @@ my $result = GetOptions('version' => \$version,
 						'nogroup' => \$nogroup,
 						'expgroup' => \$expgroup,
 						'outdir=s' => \$outdir,
+						'reference=s' -> \$reference,
 						'extract!' => \$unzip,
 						'format=s' => \$format,
 						'threads=i' => \$threads,
@@ -146,6 +148,14 @@ if ($outdir) {
 	push @java_args ,"-Dfastqc.output_dir=$outdir";
 }
 
+if ($reference) {
+	unless(-e $reference and -f $reference) {
+		die "Specified reference file '$reference' does not exist\n";
+	}
+	
+	push @java_args ,"-Dfastqc.reference=$reference";
+}
+
 if ($contaminant)  {
 	unless (-e $contaminant and -r $contaminant) {
 		die "Contaminant file '$contaminant' did not exist, or could not be read\n";
@@ -387,7 +397,12 @@ DESCRIPTION
                     
     -f --format     Bypasses the normal sequence file format detection and
                     forces the program to use the specified format.  Valid
-                    formats are bam,sam,bam_mapped,sam_mapped and fastq
+                    formats are bam,sam,cram,bam_mapped,sam_mapped,cram_mapped
+                    and fastq
+                    
+    -r --reference  Only relevant when parsing CRAM files.  Specifies a fasta format
+                    file of reference sequences which can be used to decode the 
+                    CRAM entries.
                     
     -t --threads    Specifies the number of files which can be processed
                     simultaneously.  Each thread will be allocated 250MB of
@@ -427,6 +442,6 @@ DESCRIPTION
 BUGS
 
     Any bugs in fastqc should be reported either to simon.andrews@babraham.ac.uk
-    or in www.bioinformatics.babraham.ac.uk/bugzilla/
+    or in https://github.com/s-andrews/fastqc/issues
                    
     
diff --git a/uk/ac/babraham/FastQC/FastQCConfig.java b/uk/ac/babraham/FastQC/FastQCConfig.java
index 649a7a8..4bb18d5 100644
--- a/uk/ac/babraham/FastQC/FastQCConfig.java
+++ b/uk/ac/babraham/FastQC/FastQCConfig.java
@@ -32,6 +32,7 @@ public class FastQCConfig {
 	public Integer threads = null;
 	public boolean showUpdates = true;
 	public File output_dir = null;
+	public File reference = null;
 	public boolean casava = false;
 	public boolean nano = false;
 	public boolean nofilter = false;
@@ -52,7 +53,15 @@ private FastQCConfig () {
 				throw new IllegalArgumentException("Output dir "+output_dir+" doesn't exist or isn't writeable");
 			}
 		}
-		
+
+		// CRAM Reference
+		if (System.getProperty("fastqc.reference") != null) {
+			reference = new File(System.getProperty("fastqc.reference"));
+			if (!(reference.exists() && reference.canRead())) {
+				throw new IllegalArgumentException("Reference file "+reference+" doesn't exist or can't be read");
+			}
+		}
+
 		// Contaminant file
 		if (System.getProperty("fastqc.contaminant_file") != null) {
 			contaminant_file = new File(System.getProperty("fastqc.contaminant_file"));
diff --git a/uk/ac/babraham/FastQC/Sequence/BAMFile.java b/uk/ac/babraham/FastQC/Sequence/BAMFile.java
index c26eb21..ab267ab 100644
--- a/uk/ac/babraham/FastQC/Sequence/BAMFile.java
+++ b/uk/ac/babraham/FastQC/Sequence/BAMFile.java
@@ -33,6 +33,7 @@
 import htsjdk.samtools.SamReader;
 import htsjdk.samtools.SamReaderFactory;
 import htsjdk.samtools.ValidationStringency;
+import uk.ac.babraham.FastQC.FastQCConfig;
 
 
 public class BAMFile implements SequenceFile {
@@ -63,7 +64,12 @@ protected BAMFile (File file, boolean onlyMapped) throws SequenceFormatException
 		
 		SamInputResource sir = SamInputResource.of(fis);
 		
-		br = SamReaderFactory.makeDefault().validationStringency(ValidationStringency.SILENT).referenceSequence(new File("c:/Users/andrewss/Desktop/phix-illumina.fa")).open(sir);
+		if (FastQCConfig.getInstance().reference != null) {
+			br = SamReaderFactory.makeDefault().validationStringency(ValidationStringency.SILENT).referenceSequence(FastQCConfig.getInstance().reference).open(sir);
+		}
+		else {
+			br = SamReaderFactory.makeDefault().validationStringency(ValidationStringency.SILENT).open(sir);
+		}
 		
 		it = br.iterator();
 		readNext();