update metaSNP_COV

metaSNP_COV

 #!/bin/bash
 
+################################################
+#  metaSNP Step I:   `Coverage estimation`     #
+################################################
+#
+# Helper script for the coverage estimation.
+# -- This script will execute the coverage estimation for each sample (default)
+#
+# Parallele computing mode (-P) :
+# -- The script will generate a commandline job for each sample in the list.
+# -- each job has to be run in order to compute SNPs for the entire population
+#
+# This code is part of the metagenomic SNP calling pipeline (metaSNP)
+# Copyright (c) 2016 Robin Munch
+# Licenced under the GNU General Public License (see LICENSE) 
+
+
+# Abort on any errors
+set -e
+
 display_usage() { 
 	echo >&2 ""
 	echo >&2 "	Usage: $(basename $0) project_dir/ all_samples"
@@ -9,8 +28,15 @@ display_usage() {
 	echo >&2 "	project_dir/	DIR	A metaSNP initialized project directory (metaSNP_NEW)"
 	echo >&2 "	all_samples	FILE	Input list of bam files, one file per line."
 	echo >&2 ""
+        echo >&2 "  Options:"
+        echo >&2 "      -P		Print a commandline job for each sample in all_samples"
+        echo >&2 ""
 } 
 
+missing_sample(){
+        echo >&2 "ERROR: '$1' in '$2' does not exist" 
+}
+
 missing() {
         echo >&2 ""
         echo >&2 "ERROR: '$1' no such file or directory"
@@ -34,21 +60,15 @@ make_dir() {
 }
 
 
-# print error message if no argument is supplied
-if [  $# -ne 2  ] 
-then 
-	display_usage
-	exit 1
-fi
-
 # assign arguments to variables
-DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+DIR="$( dirname $( readlink -f "${BASH_SOURCE[0]}" ) )"
 all_samples="$2"
 PROJECT_DIR="$1"
 OUT_DIR="$PROJECT_DIR/cov"
+PRINT=""
 
 # getopt to use -h flag
-ARGS=$(getopt -o h -n "$0" -- "$@")
+ARGS=$(getopt -o Ph -n "$0" -- "$@")
 # reorganize arguments as returned by getopt
 eval set -- "$ARGS"
 
@@ -61,6 +81,10 @@ while true; do
             display_usage
             exit 1
             ;;
+        -P)
+            shift
+            PRINT="P"
+            ;;
         --)
             shift
             break
@@ -68,6 +92,7 @@ while true; do
     esac
 done
 
+
 ## Control
 [ -d "$PROJECT_DIR" ] || missing "$PROJECT_DIR"
 [ -f "$all_samples" ] || missing "$all_samples"
@@ -75,13 +100,35 @@ done
 # Check required output directories
 [ -d "$OUT_DIR" ] || make_dir "$PROJECT_DIR" "$OUT_DIR"
 
+# Check if all files exist
+missed=""
+while read file
+do
+    if [ ! -f $file ]
+        then
+#	    >&2 echo ""            
+#	    >&2 echo "ERROR: '$file' in $all_samples does not exist."
+	    missing_sample "$file" "$all_samples"
+	    missed="T"
+    fi
+done<$all_samples
+
+[ -z "$missed" ] || { echo ""; echo "Please remove erroneous paths in 'all_samples'"; echo ""; exit ; }
+
+
+##
+## Processing
+##
+
 ## Generate command lines
 while read file
 do
-#    name=$(echo $file | awk -F"/" '{split($NF,a,"."); print a[1]}')
-#    name=$(echo $file | awk -F"/" '{split($NF,a,"/"); print a[1]}')
-	name=$(basename $file)
-    echo $DIR/src/qaTools/qaCompute -c 10 -d -i $file $OUT_DIR/$name.cov
+    if [ $PRINT ]
+	then 
+	    name=$(basename $file)
+	    echo $DIR/src/qaTools/qaCompute -c 10 -d -i $file $OUT_DIR/$name.cov
+	else
+	   $DIR/src/qaTools/qaCompute -c 10 -d -i $file $OUT_DIR/$name.cov 
+    fi
 
 done < $all_samples
-exit

[Luis Pedro Coelho]

I find it a bit worrisome that the commandline that is printed with -P is not the same as the one that is executed if not -P (almost the same, but the basename call makes it potentially different). I'd much prefer something along the lines of

cmdline=....
if [ $print] ; then
    echo $cmdline
else
   eval $cmdline
fi