#!/bin/bash
# Main script for turning collated molecule directories into dspace 
# import packages.
#
# Usage: makeBatches.sh [starting mol idx] [Batch size] [# of batches]
batchMax=$3
batchSz=$2
strtId=$1
dataIn=data/import2
dataOut="data/batches/`date -u --iso-8601=seconds`"
echo "Converting data from $dataIn to $dataOut"
mkdir -p $dataOut
for ((batch=0; batch < batchMax ; batch++))
do
  start=$(( batch*batchSz + strtId ))
  finish=$(($start + batchSz -1))
  echo "Batch #$batch will contain molecules with ids from $start to $finish" 
  batchdir="$dataOut/batch$start-$finish"
  mkdir -p $batchdir
  for ((mol=start; mol <= finish; mol++))
  do
    molName="000000$mol"
    padMolName=${molName:$((${#molName}-6))}
    location=$dataIn/${padMolName:5:1}/${padMolName:4:1}/${padMolName:3:1}/${padMolName}
    if [ -e $location ]
    then
	cp -r "$location" "$batchdir"
	echo -n .
    else
	echo -n x
    fi
  done
  echo;echo "Data extracted for batch #$batch"
  ./extractMetadata.groovy "$batchdir"
  cd $batchdir/..
  tar -cf "batch$start-$finish.tar" "batch$start-$finish"
  cd -
done