Convert a Payslip to text format

payslip2text.sh

#!/bin/sh
# payslip2text: Convert <company> payslip(s) to text form

# awk: extract fields
PP='/^If undeliverable/          { printf "%s-%s\n", $8, $9 }
    /^Gross Salary/              { print $3 };
    /^Contr.employer Health Ins/ { print $4 };
    /^Pension fund premium/      { print $4 };
    /^PAWW contribution/         { print $3 };
    /^Taxable wage/              { print $3 };
    /^Wage Tax/                  { print $4 };
    /^Home work allowance/       { print $4 };
    /^Net Wages/                 { print $3 };
    /^Premium Health Ins/        { print $4 };
    /^Payable Amount/            { print $3 };
    /^Bank transfer/             { print $4 }'

# sed: filter (rules semicolon delimited)
# 1. Strip decimals (European notation)
# 2. Replace commas with decimals (US notation)
# 3. Remove trailing minus sign (Payslip characteristic)
FF="s/\.//g;s/,/\./g;s/-$//g"

# process arguments (convert if needed)
echo "Date Gross Emp.Health.Ctr Pension.Prem PAWW.Ctr Taxable Tax WFH.Allow Net Health.Prem Payable Transferred"
for i 
do
  case $i in
    *.pdf) buf=`pdftotext -q -raw $i - 2>/dev/null` ;;
    *.txt) buf=`cat $i` ;;
    *) echo "Warning: Skipping unsupported file-extension: $i" 1>&2; continue ;;
  esac
  [[ $? -eq 0 ]] && printf "%s" "${buf}" | awk "${PP}" | sed "${FF}" | xargs || \
    (echo "Error: Cannot process buffer for $i" 1>&2 && exit -1)
done