essay.sh

#!/bin/bash
#
# essay.sh
#
# Bash script to find, extract, cleanup, and deliver 
# graduate admissions essays from new student applications.
#
# sjnovick 2015
#

# paths
REPORTDIR="/home/user/reports"
REPORTDEST="/mnt/OfficeShare/reports/"
SPEEDEHOMEDIR="/home/user"
ARCHIVEDIR="/home/user/archive"
LOGFILE="/home/user/essay.log"

# app name variable
APP="APPS*"

# string to determine if essay is present
STRING="ESSAY INCLUDED"

# redirect STDERR to STDOUT and append both to LOGFILE
exec 1>> $LOGFILE 2>&1

# banter for log file
printf "***** Starting: $(date)\n" >> $LOGFILE
#echo "***** Starting: $(date)"

#set counter
z=0

# loop for all apps present
for j in $SPEEDEHOMEDIR/$APP
do
    # count
    let "z++"

    # archive the current app before we get started
    cp $j $ARCHIVEDIR

    # determine if essay(s) exist in current app file
    grep -q "$STRING" $j
    if [ $? -eq 0 ]
        then
            printf "Essays found in $j\n" >> $LOGFILE

            ### BEGIN REPORT ROUTINE
            DATE2="$(date +"%m-%d-%y_%H%M%S")"        
            REPORT="$REPORTDIR/$DATE2.txt"
            # write neat report header
            printf "" > $REPORT
            printf "******************************\n" >> $REPORT
            printf "START OF REPORT\n" >> $REPORT
            printf "$DATE2\n" >> $REPORT
            printf "Source File:\n" >> $REPORT
            printf "$j\n" >> $REPORT
            printf "******************************\n\n" >> $REPORT
            # set counter
            n=0
            # array declaration
            declare -a ARRAY
            # set internal field separator to new line
            IFS=$'\n'
            # populate array, line by line, with apps that contain essays
            ARRAY=( $(grep -Pzo "ST!189![\s\S]*?ESSAY INCLUDED[\s\S]*?SE!.*" $j) )
            # loop through every line
            for i in ${ARRAY[@]}
                do
                    # check to see if this line is the start of an app
                    echo $i | grep -q "ST!189!"
                    if [ $? -eq 0 ]
                        # start of an app
                        then
                            # write neat app header to report
                            printf "************\n" >> $REPORT
                            printf "START OF APP\n" >> $REPORT
                            printf "************\n" >> $REPORT
                            # drop line
                            i=$(echo $i | sed '/ST!189!/d')
                            printf "$i\n" >> $REPORT
                            # count
                            let "n++"
                        # not the start of an app
                        else
                            # drop unwanted characters from line
                            # sed reference:
                            # sed 's/old/new/' = replace old string with new string
                            # sed '/string/d'  = drop entire line that matches string
                            i=$(echo $i | sed \
                                    -e '/N1!TM!/d' \
                                    -e '/N1!AT!/d' \
                                    -e '/REF!SY!/d' \
                                    -e '/IN1!/d' \
                                    -e '/DANGEROUS CHARACTERS FOUND/d' \
                                    -e '/MSG!DY!/d' \
                                    -e '/SE!/d' \
                                    -e 's/BGN!00!.........!//' \
                                    -e 's/!1812!CT//' \
                                    -e 's/REF!48!//' \
                                    -e 's/DTP!196!CM!//' \
                                    -e 's/IN2!05!//'  \
                                    -e 's/IN2!02!//' \
                                    -e 's/IN2!03!//' \
                                    -e 's/IN2!09!//' \
                                    -e 's/DMG!D8!//' \
                                    -e 's/RQS!!!//' \
                                    -e 's/!!!0!!8//' \
                                    -e 's/MSG!//' \
                                    -e 's/<br\s\/>//' \
                                    -e 's/<br\s\/>//')
                            # write line to report
                            printf "$i\n" >> $REPORT
                    fi
                done
            # write neat report footer
            printf "*************\n" >> $REPORT
            printf "TOTAL APPS: $n\n" >> $REPORT
            printf "**************\n" >> $REPORT
            # remove trailing backslashes and empty lines from report
            sed -i -e 's/\\//' -e '/^$/d' -e '/^\s*$/d' $REPORT
            #sed -i '/^$/d' $REPORT
            # convert report to dos format
            unix2dos -q $REPORT
            #copy report to filer
            cp $REPORT $REPORTDEST
            ### END REPORT ROUTINE
            
            # delete essays from current app file
            sed -i '/^RQS!!!ESSAY.*$/,/^SE.*$/ {/^SE.*$/!d}' "$j"
    else
        printf "No essays found in $j\n" >> $LOGFILE
    fi
done
printf "Applications processed: $z\n" >> $LOGFILE
printf "Essays processed: $n\n" >> $LOGFILE
printf "***** Done: $(date)\n" >> $LOGFILE