upload_viollier_raw

#!/bin/bash

configfile=viollier.conf

usage() { echo "Usage: $0 [-c <configfile>] [ -s <subdir> ] [ -q ] TSVFILE " 1>&2; exit $1; }

subdir=$(date '+%Y-%m-%d')
verbose='v'

while getopts "c:s:qh" o; do
	case "${o}" in
		c)	configfile=${OPTARG}
			if [[ ! -r ${configfile} ]]; then
				echo "Cannot read ${configfile}" 1>&2
				usage 1
			fi
			;;
		q)	verbose='' ;;
		s)	subdir=${OPTARG} ;;
		h)	usage 0	;;
		*)	usage 1	;;
	esac
done
shift $((OPTIND-1))


. ${configfile}

: ${fileserver:?}
# ${srvport}
: ${expname:?}
: ${basedir:=$(pwd)}
: ${download:?}
: ${sampleset:?}


target_base=raw_othercenters
target=${target_base}/${subdir}

(( ${#@} == 0 )) && usage 2
[[ "${1}" == '-h' || "${1}" == '--help' ]] && usage 0

if [[ ! -d ${basedir}/${download}/ ]]; then
	echo "missing directory ${download}" >&2
	exit 1
elif [[ ! -d ${basedir}/${sampleset}/ ]]; then
	echo "missing sampleset ${sampleset}" >&2
	exit 1
fi

if [[ ! -r "${1}" ]]; then
	echo "Cannot read file ${1}" >&2
	exit 2
fi

mkdir -p ${basedir}/${download}/${target}
cd ${basedir}
error=0
while read aufxx ethid rest; do
	# remove DOS/Windows-style CR+LF line ending
	auf=${aufxx%%$'\r'}
	# skip empty lines
	if [[ -z ${auf} ]]; then
		continue
	elif [[ ! ${auf} =~ ^[[:digit:]]{8,} ]]; then
		echo "cannot parse ${auf}" >&2
		error=1
		continue
	fi
	aufnum=${BASH_REMATCH[0]}

	if [[ -z ${ethid} ]]; then
		rx="^([[:digit:]]*[[:alpha:]]+[[:alnum:]-]+_)?${auf}[^\t]*"
		ethid=$(grep -hoP "${rx}" ${sampleset}/samples.*.tsv)
		if [[ -z ${ethid} ]]; then
			ethid=$(grep -hoP "${rx}" ${sampleset}/missing.*.txt)
			if [[ -z ${ethid} ]]; then
				echo "(missing ${aufnum})"
			else
				echo "(zero yield for ${aufnum} in  ${ethid//$'\n'/ })"
			fi
			continue
		elif [[ ${ethid} == *$'\n'* ]]; then
			echo -e "(${aufnum}: multiple matches: ${ethid//$'\n'/ }\nUploading all matches)"
		fi
	elif [[ ! ${ethid} =~ ^[[:digit:]]{6,} ]]; then
		echo "cannot parse ${ethid}" >&2
		error=1
		continue
	fi

	[[ -d ${download}/${target}/${aufnum}/ ]] && rm -rf ${download}/${target}/${aufnum}/
	mkdir -p ${download}/${target}/${aufnum}
	# copy each duplicate sample in a separate subdirectory
	for singleethid in ${ethid}
	do
		echo "${singleethid} -> ${aufnum}"
		for b in ${sampleset}/${singleethid}/*; do
			bname="$(basename "${b}")"
			# NOTE make sure to distinguish samples by their internal ETHID, in case that they come from the same batch, e.g.:
			#
			# if '34889065' maps to both '34889065_A9' and '34889065_C9' from the same batch '20211203_HTHMGDRXY':
			# 'sampleset/34889065_A9/20211203_HTHMGDRXY' => '...{target}/34889065/20211203_HTHMGDRXY-34889065_A9'
			# 'sampleset/34889065_C9/20211203_HTHMGDRXY' => '...{target}/34889065/20211203_HTHMGDRXY-34889065_C9'
			#
			cp --link -${verbose}rf ${b} ${download}/${target}/${aufnum}/${bname}-${singleethid} || error=1
		done
	done
done < <( cat ${1}; echo; )

(( error )) && exit 0

exec ./upload_sftp -c ${configfile} ${target_base}