Rename PDF files downloaded from paperlessemployee.com

rename-paystatements-from-paperlessemployee.sh

#!/usr/bin/env bash
#
# rename pdf files from paperlessemployee.com
#

set -euo pipefail

if [[ ${#} -ne 1 ]] && [[ ${#} -ne 2 ]]; then
    echo "usage:
    $(basename "${0}") SelectedPayStatements.pdf [Agency Name]

about:
  rename pay statements PDF files from www.paperlessemployee.com from default name 'SelectedPayStatements.pdf'
  to 'My Agency Pay Statement YYYY-MM-DD to YYYY-MM-DD (X USD).pdf'
  e.g.
  'Super Talent Pay Statement 2023-01-21 to 2023-01-28 (total 1,234.00 USD) (final 1,000.00 USD).pdf'

  Presumes the pay statement is only one pay statement and not multiple pay statements
  within one PDF file.
"
    exit
fi

pdf_file=${1-}
agency_name=${2-}

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

if [[ "${pdf_file:(-4)}" != '.pdf' ]]; then
    echo "ERROR must pass a file ending with .pdf, given '${pdf_file}'" >&2
    exit 1
fi

function remove_leading_0 () {
    # remove all leading '0' from the passed $1 string
    echo -n "${1}" | sed -Ee 's/^0+//g'
}

function is_positive_number () {
    # is the passed number greater than 0?
    declare -i num=$(remove_leading_0  "${1}")
    [[ ${num} -gt 0 ]]
}

function trim_whitespace () {
    # delete any whitespace surrounding the piped text
    sed -e 's#[\t ]\+\(.\+\)[\t ]\+#\1#'
}

function remove_blank_lines () {
    # delete any blank lines in the piped text
    sed -Ee '/^[[:space:]]*$/d'
}

function print_line () {
    # print line number $1 from the stdin text (piped data)
    sed -n "${1}p"
}

if ! which pdftotext &>/dev/null; then
    echo "ERROR cannot find 'pdftotext' in the PATH" >&2
    echo "      pdftotext is part of apt package 'poppler-utils'" >&2
    exit 1
fi

text=$(pdftotext "${pdf_file}" -)
text=$(echo -n "${text}" | remove_blank_lines)

agency_name_append=' '
if [[ -z "${agency_name}" ]]; then
    agency_name_append=''
fi
new_name="${agency_name}${agency_name_append}Pay Statement"

# find the unique string before the date text
# Example text around the week dates:
#
#     Pay Period Begin:
#     07/30/2023
#     Pay Period End:
#     08/05/2023
#
if ! text_dates=$(echo "${text}" | grep -A3 -Ee '^Pay Period Begin:'); then
    echo "ERROR failed to find unique string 'Pay Period Begin' in '${pdf_file}'" >&2
    exit 1
fi
echo -e "text_dates:\n${text_dates}"
echo
text_date_start=$(echo -n "${text_dates}" | print_line 2 | trim_whitespace)
text_date_end=$(echo -n "${text_dates}" | print_line 4 | trim_whitespace)
echo "text_date_start: '${text_date_start}'"
echo "text_date_end: '${text_date_end}'"

#
# date text  07/30/2023
# index      0123456789
#
# re-arrange from 'MM/DD/YYYY' to 'YYYY-MM-DD'
#

ds_y=${text_date_start:6:4}  # Date Start Year
ds_m=${text_date_start:0:2}  # Date Start Month
ds_d=${text_date_start:3:2}  # Date Start Day
is_positive_number ${ds_y} || { echo "ERROR: bad year '${ds_y}' in Date Start string '${text_date_start}'" >&2; exit 1; }
is_positive_number ${ds_m} || { echo "ERROR: bad month '${ds_m}' in Date Start string '${text_date_start}'" >&2; exit 1; }
is_positive_number ${ds_d} || { echo "ERROR: bad day '${ds_d}' in Date Start string '${text_date_start}'" >&2; exit 1; }

de_y=${text_date_end:6:4}  # Date End Year
de_m=${text_date_end:0:2}  # Date End Month
de_d=${text_date_end:3:2}  # Date End Day
is_positive_number ${de_y} || { echo "ERROR: bad year '${de_y}' in Date End string '${text_date_end}'" >&2; exit 1; }
is_positive_number ${de_m} || { echo "ERROR: bad month '${de_m}' in Date End string '${text_date_end}'" >&2; exit 1; }
is_positive_number ${de_d} || { echo "ERROR: bad day '${de_d}' in Date End string '${text_date_end}'" >&2; exit 1; }

new_name+=" ${ds_y}-${ds_m}-${ds_d} to ${de_y}-${de_m}-${de_d}"

# find the unique strings that describe pay
# Example of such text:
#
#    TOTAL
#    TAXES
#    TOTAL
#    DEDUCTIONS
#    NET PAY
#    Current:
#    1,234.00
#    123.45
#    12.34
#    1,234.12
#    1,234.12
#    1,234.12
#    YTD:
#    9,999.99
#    999.99
#    99.99
#    9,900.00
#    9,000.00
#    9,000.00
#    0.00
#    111.11
#    222.22
#    3,333.33
#    0.00
#    1,444.44
#    333.33
#    3,210.00
#    LEAVE BALANCE
#

if text_pay_raw=$(echo "${text}" | grep -A12 -Ee '^NET PAY'); then
    # pay total current period
    if text_pay=$(echo -n "${text_pay_raw}" | print_line 3 | trim_whitespace); then
        declare -i text_pay_num=${text_pay%%.??}  # trim trailing '.00' (sub-dollar amount)
        if [[ "${text_pay}" != '' ]] && [[ "${text_pay_num}" -gt 0 ]]; then
            echo "text_pay (total): '${text_pay}'"
            new_name+=" (total ${text_pay} USD)"
        else
            echo "Failed to convert text_pay total" >&2
        fi
    else
        echo "Failed to extract text_pay total from text_pay_raw" >&2
    fi
    # pay final current period (after taxes amount)
    if text_pay=$(echo -n "${text_pay_raw}" | grep -B1 -Ee '^YTD:' | print_line 1 | trim_whitespace); then
        declare -i text_pay_num=${text_pay%%.??}  # trim trailing '.00' (sub-dollar amount)
        if [[ "${text_pay}" != '' ]] && [[ "${text_pay_num}" -gt 0 ]]; then
            echo "text_pay (final): '${text_pay}'"
            new_name+=" (final ${text_pay} USD)"
        else
            echo "Failed to convert text_pay final" >&2
        fi
    else
        echo "Failed to extract text_pay final from text_pay_raw" >&2
    fi
else
    echo "Failed to find NET PAY" >&2
fi

# find the unique string that describe hours
# Example of such text:
#
#    Current Hours
#    Current Amt
#    YTD Hours
#    YTD Amt
#    40.00
#    1,234.00
#    123.00
#    99,999.00
#    40.00
#    123.00
#    99,999.00
#    EARNINGS
#

if text_hours_raw=$(echo "${text}" | grep -A5 -Ee '^Current Hours'); then
    text_hours=$(echo "${text_hours_raw}" | print_line 5 | trim_whitespace)
    if [[ "${text_hours}" != '' ]]; then
        echo "text_hours      : '${text_hours}'"
        new_name+=" (${text_hours} hours)"
    else
        echo "Failed to extract text_hours from text_hours_raw" >&2
    fi
fi

new_name+=".pdf"
echo "new_name: '${new_name}'"
echo

pdf_path=$(dirname -- "${pdf_file}")
set -x
mv -v "${pdf_file}" "${pdf_path}/${new_name}"