filter_ids

#!/bin/bash

set -e

# ------------------------------------------------------------------------------
# usage
# ------------------------------------------------------------------------------
function usage() {
  echo -e "\n\tusage:\t\tfilter_ids <opts>"
  echo -e "\n\tdescription:\tfilter ids from a csv file and return the row indices"
  echo -e "\n\toptions:\t"
  echo -e "\t\t--study <study>\t\tstudy name"
  echo -e "\t\t--sub <sub>\t\tsubject id"
  echo -e "\t\t--ses <ses>\t\tsession id"
  echo -e "\t\t--scan <scan>\t\tscan id"
  echo -e "\t\t--airc-id <airc-id>\tairc id"
  echo -e "\t\t--idx-only <0|1>\treturn row indices only (default: 0)"
  echo -e "\t\t-h|--help\t\ttdisplay help message"
  echo ""
}

# ------------------------------------------------------------------------------
# args
# ------------------------------------------------------------------------------
if [[ $# -eq 0 ]]; then
  usage
  exit 0
fi

# defaults
idx_only=0

vars=""
while [[ $# -gt 0 ]]; do
  case "$1" in
    --study)
      study=$2
      if [[ -z ${study} ]]; then
        usage
        error_msg "study value not specified"
        exit 1
      elif [[ ${study} == "all" ]]; then
        study=""
      fi
      vars="${vars} study"
      shift 2
      ;;
    --sub)
      sub=$2
      if [[ -z ${sub} ]]; then
        usage
        error_msg "sub value not specified"
        exit 1
      elif [[ ${sub} == "all" ]]; then
        sub=""
      fi
      vars="${vars} sub"
      shift 2
      ;;
    --ses)
      ses=$2
      if [[ -z ${ses} ]]; then
        usage
        error_msg "ses value not specified"
        exit 1
      elif [[ ${ses} == "all" ]]; then
        ses=""
      fi
      vars="${vars} ses"
      shift 2
      ;;
    --scan)
      scan=$2
      if [[ -z ${scan} ]]; then
        usage
        error_msg "scan value not specified"
        exit 1
      elif [[ ${scan} == "all" ]]; then
        scan=""
      fi
      vars="${vars} scan"
      shift 2
      ;;
    --airc-id)
      airc_id=$2
      if [[ -z ${airc_id} ]]; then
        usage
        error_msg "airc_id value not specified"
        exit 1
      elif [[ ${airc_id} == "all" ]]; then
        airc_id=""
      fi
      vars="${vars} airc_id"
      shift 2
      ;;
    --idx-only)
      idx_only=$2
      if [[ -z ${idx_only} ]]; then
        usage
        error_msg "idx_only value not specified"
        exit 1
      fi
      if [[ ${idx_only} -ne 0 ]] && [[ ${idx_only} -ne 1 ]]; then
        # echo -e "\n\tinvalid idx_only value: ${idx_only}"
        usage
        error_msg "invalid idx_only value (${idx_only}), must be 0 or 1"
        exit 1
      fi
      vars="${vars} idx_only"
      shift 2
      ;;
    -h|--help)
      usage
      exit 0
      ;;
    *)
      echo -e "\n\tunknown option: $1"
      usage
      exit 1
      ;;
  esac
done

# ------------------------------------------------------------------------------
# paths
# ------------------------------------------------------------------------------
root_dir=`get_root_dir kenrod`
in_ids="${root_dir}/software/scripts/eep170030/ids_long-format_study-all.csv"

# ------------------------------------------------------------------------------
# check paths
# ------------------------------------------------------------------------------
check_in_paths ${in_ids}

# ------------------------------------------------------------------------------
# main
# ------------------------------------------------------------------------------
# Read the column names from the header line
header=$(head -n 1 "$in_ids")

# Extract the desired column names dynamically
study_col=$(echo "${header}" | awk -F, '{ for (i = 1; i <= NF; i++) if ($i == "study") print i }')
subj_col=$(echo "${header}" | awk -F, '{ for (i = 1; i <= NF; i++) if ($i == "sub") print i }')
ses_col=$(echo "${header}" | awk -F, '{ for (i = 1; i <= NF; i++) if ($i == "ses") print i }')
scan_col=$(echo "${header}" | awk -F, '{ for (i = 1; i <= NF; i++) if ($i == "scan") print i }')
airc_id_col=$(echo "${header}" | awk -F, '{ for (i = 1; i <= NF; i++) if ($i == "airc_id") print i }')

# Read the CSV file and filter lines based on arguments
if [[ ${idx_only} == 0 ]]; then
  awk \
  -F, \
  -v study=${study} \
  -v subj=${sub} \
  -v ses=${ses} \
  -v scan=${scan} \
  -v airc_id=${airc_id} \
  -v study_col=${study_col} \
  -v subj_col=${subj_col} \
  -v ses_col=${ses_col} \
  -v scan_col=${scan_col} \
  -v airc_id_col=${airc_id_col} \
  ' \
  NR > 1 && \
  ($study_col == study || study == "") && \
  ($subj_col == subj || subj == "") && \
  ($ses_col == ses || ses == "") && \
  ($scan_col == scan || scan == "") && \
  ($airc_id_col == airc_id || airc_id == "") \
  {
      print FNR - 1 "," $0;
  }
  ' "${in_ids}"
elif [[ ${idx_only} -eq 1 ]]; then
  row_numbers=`awk \
  -F, \
  -v study="${study}" \
  -v subj="${sub}" \
  -v ses="${ses}" \
  -v scan="${scan}" \
  -v airc_id="${airc_id}" \
  -v study_col="${study_col}" \
  -v subj_col="${subj_col}" \
  -v ses_col="${ses_col}" \
  -v scan_col="${scan_col}" \
  -v airc_id_col="${airc_id_col}" \
  ' \
  NR > 1 && \
  ($study_col == study || study == "") && \
  ($subj_col == subj || subj == "") && \
  ($ses_col == ses || ses == "") && \
  ($scan_col == scan || scan == "") && \
  ($airc_id_col == airc_id || airc_id == "") \
  {
      print FNR - 1;
  }
  ' "$in_ids" | tr '\n' ' ' | sed 's/ /,/g' | sed 's/,$//g'`
  echo "${row_numbers}"
fi