get_names_and_accessions

#!/bin/bash
set -e          # Will EXIT the script if any command returns a non-zero exit status.
set -E          # Make ERR trapping work inside functions too.
set -u          # Variables must be pre-defined before using them.
set -o pipefail # If a pipe fails, returns the error code for the failed pipe even if it isn't the last command in a series of pipes.

# If an error occurs, report the line number to stderr (see https://olivergondza.github.io/2019/10/01/bash-strict-mode.html).
trap 's=$?; echo >&2 "$0: Error on line "$LINENO": $BASH_COMMAND"; exit $s' ERR

input_list=$1
input_base=$(basename $input_list)
input_dir=$(dirname $input_list)

function build_name() {
    prepend=$1
    echo "$input_dir/$prepend.$input_base"
}

if [[ ! -f "names_to_ID.hmdb_metabolites.txt" ]]; then

    if [[ ! -f "hmdb_metabolites.xml" ]]; then
        wget https://hmdb.ca/system/downloads/current/hmdb_metabolites.zip
        unzip hmdb_metabolites.zip
    fi



    grep -P '^  <name>' hmdb_metabolites.xml | sed 's/  <name>//' | sed 's/<\/name>//'  > names.txt
    grep -P '^  <accession>' hmdb_metabolites.xml | sed 's/  <accession>//' | sed 's/<\/accession>//'  > accessions.txt
    paste names.txt accessions.txt | sort > names_to_ID.hmdb_metabolites.txt
    rm names.txt
    rm accessions.txt

fi

awk 'BEGIN{FS="\t"; OFS="\t"} {print NR,$0}' $input_list | sort -k2 > $( build_name numbered )
sort $input_list  > $( build_name sorted )

join -a 2 -t$'\t' names_to_ID.hmdb_metabolites.txt $( build_name sorted ) > $( build_name name_sorted.combined )

join -1 2 -2 1 -t$'\t' $( build_name numbered ) $( build_name name_sorted.combined ) | sort -t$'\t' -k2 > $( build_name original_order )

awk 'BEGIN{FS="\t"; OFS="\t"} {print $1,$3}' $( build_name original_order ) > $( build_name clean_original_order )