#!/bin/ksh

# Usage:  getcomix.ksh [ -c n ]
# 
#    where n is the number of days' worth of comix (counting
#    backwards from the current date) you wish to view.
#    NOTE: -c does not cross month boundaries.  If you want to
#    view three days' worth of comix on the first of the month,
#    you'll have to copy the script to a temporary file and
#    modify THISDAY and THISMONTH manually.
#
# wget, lynx and xv are assumed to work and be in the PATH.

#unset http_proxy

GETDATE=""
PATH=$PATH:/opt/gnu/bin:/opt/sfw/bin:/usr/sfw/bin   # Just to be sure
TMPDIR=~/pix
COMICS="ad bt db ft fu mb td tmhel luc tm tmcat ba be bo pr ss tmloo nq cle poc hc wes tmhou prc"
UMCREATORS=''
UMWASH=''
UMCOMIX="arlonjanis dilbert getfuzzy bignate luann forbetter pibgorn chickweed peanuts soup2nutz sheldon barkeaterlake frazz brevity"
RBMACOMIX="Baby_Blues Bizarro Blondie Crankshaft Family_Circus Funky_Winkerbean Lockhorns Mallard_Fillmore Mutts Piranha Sally_Forth 6Chix Mgoose Zippy_the_Pinhead Zits Fast_Track"
CHRONCOMIX="9_Chickweed_Lane BC Baldo Ballard_Street Boffo Boondocks Cathy Cleats Dilbert Dinette_Set Doonesbury Drabble For_Better_Or_Worse Foxtrot Fred_Basset Fuzzy Heart Jumble Kudzu La_Cucaracha Luann Monty Non_Sequitur_pan One_Big_Happy Peanutsb Pearls_Before_Swine Preteena Rhymes_with_Orange Rose Rubes Safe_Havens Shermans_Lagoon Shoe Speed_Bump Tank_McNamara Willy_n_Ethel Wizard_of_Id"
YEAR=$(date +%Y)
SHORTYEAR=$(date +%y)
COUNT=0
NUMDAYS=1
NOHUP=''

# Set up days in month

set -A DAYS_IN_MONTH 0 31 28 31 30 31 30 31 31 30 31 30 31

# Set up leap year
# Note that this is not Y2100 compliant code

YEAR_DIV_4=$(($YEAR / 4))

if [[ $((YEAR_DIV_4 * 4)) -eq $YEAR ]]
then
   DAYS_IN_MONTH[2]=29
fi

function strip_leading_zeroes {

    # Kludge to keep, e.g., 08, from being interpreted as octal in
    # IN_MONTH and IN_DAY

    if [[ IN_MONTH -lt 10 ]]
    then
        IN_MONTH=$(print $IN_MONTH | /bin/cut -c2)
    fi

    if [[ IN_DAY -lt 10 ]]
    then
        IN_DAY=$(print $IN_DAY | /bin/cut -c2)
    fi

}

function normalize_date {

    IN_STR=$1
    IN_MONTH=$(print $IN_STR | cut -c1-2)
    IN_DAY=$(print $IN_STR | cut -c3-4)

    # strip leading zeroes

    OUT_MONTH=$((IN_MONTH + 0))
    OUT_DAY=$((IN_DAY + 0))

    NORMAL_DATE="$YEAR/$OUT_MONTH/$OUT_DAY"

}

function daysoff {

   IN_DATE=$1
   DAYS_OFF=$2

   IN_MONTH=$(print $IN_DATE | /bin/cut -c1-2)
   IN_DAY=$(print $IN_DATE | /bin/cut -c3-4)

   strip_leading_zeroes 

#   echo DEBUG In daysoff:  IN_DATE = $IN_DATE
#   echo                    DAYS_OFF = $DAYS_OFF
#   echo                    IN_MONTH = $IN_MONTH
#   echo                    IN_DAY = $IN_DAY

   if [[ $IN_DAY -gt $DAYS_OFF ]]   # It's not early in the month -- simple case
   then
      OUT_DAY=$((IN_DAY - DAYS_OFF))
      OUT_MONTH=$IN_MONTH
      RESULT=$(printf "%02d%02d" $OUT_MONTH $OUT_DAY)
      return
   fi

   if [[ $IN_MONTH -gt 1 ]]   # It's not January
   then
      OUT_MONTH=$(($IN_MONTH - 1))
      OUT_DAY_TEMP=$(($IN_DAY + ${DAYS_IN_MONTH[$OUT_MONTH]}))
      OUT_DAY=$(($OUT_DAY_TEMP - $DAYS_OFF))
      RESULT=$(printf "%02d%02d" $OUT_MONTH $OUT_DAY)
      return
   fi

   if [[ $IN_MONTH -eq 1 ]]   # It's January, too early in the month
   then
      print "ERROR:  Cannot cross year boundaries"
      return
   fi

   print "ERROR:  impossible case in function daysoff, probably a month problem"

}

function getdate {

   THISDAY=$(date +%d)
   THISMONTH=$(date +%m)

# Remove leading zeroes --  if not, printf interprets them as octal numbers.

   THISDAY=$((THISDAY + 0))
   THISMONTH=$((THISMONTH + 0))

   NEWDATE=$(printf "%02d%02d" $THISMONTH $THISDAY)

   while [[ $NUMDAYS -gt 0 ]]
   do
      GETDATE="$NEWDATE $GETDATE"
      daysoff $NEWDATE 1
      NEWDATE=$RESULT
      NUMDAYS=$((NUMDAYS - 1))
   done

   if [[ -z $GETDATE ]]
   then
      GETDATE=$(printf "%02d%02d" $THISMONTH $THISDAY)
      print GETDATE is $GETDATE
   fi

}

function loadshare {

    SUBPROCESSES=$(/usr/bin/ptree $$ | /bin/wc -l)

    if [[ $SUBPROCESSES -gt 100 ]]
    then
        echo $SUBPROCESSES subprocesses found ... sleeping $SLEEPSECS seconds
        SLEEPSECS=$((SUBPROCESSES / 100))
        sleep $SLEEPSECS
#    else
#        echo DEBUG $SUBPROCESSES subprocesses found
    fi
}

# usage: getups -c <n>
#        getups [ MMDD [ MMDD ... ] ]
# 
# where <n> is a count backwards from the current date (including the
# current date -- current month only) and MMDD [MMDD ...] is a list of
# dates with leading zeroes.  If omitted, todays date is the default.

mkdir $TMPDIR 2>/dev/null
cd $TMPDIR

while getopts :c: opt
do
   case $opt in 
        c) COUNT=1
           NUMDAYS=$OPTARG
           ;;
   \?) print "usage: getups -c n"
        print "       getups \[ MMDD \[ MMDD ... \] \] "
        return 1
   esac
done
shift $(($OPTIND - 1))
TENTATIVEDATES=$*

getdate 

# First, get UPS comix:

for STRIP in $COMICS
do
   for ISSUE in $GETDATE
   do
      $NOHUP lynx --dump "http://images.ucomics.com/comics/$STRIP/$YEAR/$STRIP$SHORTYEAR$ISSUE.gif" > $STRIP$SHORTYEAR$ISSUE.gif &
      print GETTING $STRIP$SHORTYEAR$ISSUE.gif
      loadshare
   done
done

# Now get RBMA comics

for STRIP in $RBMACOMIX
do
   for ISSUE in $GETDATE
   do
      $NOHUP wget -q --referer=http://www.chron.com --output-document=0$STRIP$SHORTYEAR$ISSUE.gif http://cst.rbma.com/content/$STRIP?date=$YEAR$ISSUE &
      print GETTING $STRIP$SHORTYEAR$ISSUE.gif
      loadshare
   done
done

# Next, get UM comix:

for STRIP in $UMCOMIX
do
   for ISSUE in $GETDATE
   do
      ISSUE="$YEAR$ISSUE"

    # First, grab the line containing the name of the comic

      $NOHUP lynx -source "http://www.unitedmedia.com/comics/$STRIP/archive/$STRIP-$ISSUE.html" | grep -i "<IMG SRC=\"/comics/$STRIP/archive/images/$STRIP" > $STRIP.line

    # Next, change everything up to .../comics/$STRIP into URL format
    # (preserving the rest as the filename)

      sed "s@^.*=\"/comics/$STRIP@http://www.unitedmedia.com/comics/$STRIP@" < $STRIP.line > $STRIP.line2

    # Finally, eliminate everything after ".gif" or ".jpg" -- what's left is the URL

     grep '\.gif' $STRIP.line2 >/dev/null 2>&1
     if [[ $? -eq 0 ]]
     then
        sed 's@\.gif\".*$@.gif@' < $STRIP.line2 > $STRIP.url
     else
        sed 's@\.jpg\".*$@.jpg@' < $STRIP.line2 > $STRIP.url
     fi

    # Now actually get it.

      print GETTING $STRIP$ISSUE.gif
      $NOHUP lynx -source $(cat $STRIP.url) > $STRIP$ISSUE.gif &
      loadshare
   done
done

# UM Creators strip(s)

for STRIP in $UMCREATORS
do
   for ISSUE in $GETDATE
   do
      daysoff $ISSUE 7
      OLDISSUE=$YEAR$RESULT

      lynx -source "http://www.unitedmedia.com/creators/$STRIP/archive/$STRIP-$OLDISSUE.html" | grep "<img src=\"/creators/$STRIP/archive/images/$STRIP[0-9]*\.gif\"" > $STRIP.line
      sed "s@^.*<img src=\"/creators/$STRIP@http://www.unitedmedia.com/creators/$STRIP@" < $STRIP.line > $STRIP.line2
      sed 's@\.gif\" BORD.*$@.gif@' < $STRIP.line2 > $STRIP.url
      print GETTING $STRIP$OLDISSUE.gif
      $NOHUP lynx -source $(cat $STRIP.url) > $STRIP$OLDISSUE.gif &
      loadshare
   done
done

# UM Washington Post Writers Group strip(s)

for STRIP in $UMWASH
do
   for ISSUE in $GETDATE
   do
      daysoff $ISSUE 7
      OLDISSUE=$YEAR$RESULT

#      print DEBUG "http://www.unitedmedia.com/wash/$STRIP/archive/$STRIP-$OLDISSUE.html"
#      lynx -source "http://www.unitedmedia.com/wash/$STRIP/archive/$STRIP-$OLDISSUE.html" > $STRIP-$OLDISSUE.html
      lynx -source "http://www.unitedmedia.com/wash/$STRIP/archive/$STRIP-$OLDISSUE.html" | grep "<img src=\"/wash/$STRIP/archive/images/$STRIP[0-9]*\.gif\"" > $STRIP.line
      sed "s@^.*<img src=\"/wash/$STRIP@http://www.unitedmedia.com/wash/$STRIP@" < $STRIP.line > $STRIP.line2
      sed 's@\.gif\" BORD.*$@.gif@' < $STRIP.line2 > $STRIP.url
      print GETTING $STRIP$OLDISSUE.gif
      $NOHUP lynx -source $(cat $STRIP.url) > $STRIP$OLDISSUE.gif &
      loadshare
   done
done

# Finally, get the ones direct from Chron

for ISSUE in $GETDATE
do
    normalize_date $ISSUE
#    lynx -source "http://www.chron.com/content/chronicle/comics/archive/buildcp.hts?v=3.0&page=1&quality=high&cpp=80&c=19&c=2&c=101&c=75&c=84&c=85&c=32&c=104&c=12&c=13&c=43&c=74&c=109&c=27&c=44&c=110&c=18&c=38&c=49&c=112&c=71&c=31&c=117&c=28&c=118&c=83&c=121&c=37&c=1&c=72&c=124&c=67&c=129&c=4&c=147&c=30&c=122&c=46&c=137&c=134&c=153&c=154&c=155&c=87&c=82&c=63&c=86&c=136&c=33&c=128&c=135&c=78&c=90&date=$NORMAL_DATE" > chronsource.txt
    lynx -source "http://www.chron.com/content/chronicle/comics/archive/buildcp.mpl?v=3.0&page=1&quality=high&cpp=80&c=19&c=2&c=101&c=75&c=84&c=153&c=104&c=12&c=85&c=13&c=82&c=154&c=74&c=90&c=27&c=44&c=110&c=111&c=18&c=38&c=49&c=112&c=71&c=68&c=63&c=31&c=86&c=117&c=28&c=118&c=78&c=155&c=121&c=37&c=1&c=152&c=72&c=87&c=124&c=79&c=136&c=67&c=33&c=128&c=129&c=135&c=4&c=147&c=53&c=30&c=77&c=46&c=137&c=134" > chronsource.txt
    grep '^<IMG SRC=http://images.chron.com/content/chronicle/comics/images/' chronsource.txt > chronlines.txt
    for STRIP in $CHRONCOMIX
    do
        grep $STRIP chronlines.txt | sed "s@^<IMG SRC=@@" | sed "s@ width=.*@@" | while read FILE
        do
            $NOHUP wget -q --referer=http://www.chron.com --output-document=0$STRIP$SHORTYEAR$ISSUE.gif $FILE &
            print GETTING 0$STRIP$SHORTYEAR$ISSUE.gif
            loadshare
        done
    done
done
        

print WAITING FOR JOBS ...

wait

print DONE WITH JOBS
print

print "Renaming .gif and .jpg ASCII files to *.ASCII"

/bin/file *.[gG][iI][fF] *.[jJ][pP][gG] | grep -i ascii > ascii.files  2>/dev/null
awk -F: '{print $1}' ascii.files | while read a
do
   print Renaming $a
   /bin/mv $a $a.ASCII
done

# Now rename the zero length files, if any

for a in *.[gG][iI][fF] *.[jJ][pP][gG]
do
    if [[ -a $a && ! -s $a ]]
    then
        print Renaming zero-length file $a
        /bin/mv $a $a.ZERO
    fi
done

#VIEW=''
#
#ls *.[gG][iI][fF] >/dev/null 2>&1
#if [[ $? -eq 0 ]]
#then
#   VIEW='*.[gG][iI][fF]'
#fi
#
#ls  *.[jJ][pP][gG] >/dev/null 2>&1
#if [[ $? -eq 0 ]]
#then
#   VIEW="$VIEW *.[jJ][pP][gG]"
#fi
#
#print
#print -n Viewing files ...
#
#xv $VIEW
print

print done.

cd $OLDPWD
#print "REMOVING $TMPDIR"
#sleep 5
#/bin/rm -r $TMPDIR
print "REMOVING extraneous files"
#sleep 5

/bin/rm -f $TMPDIR/*.ZERO $TMPDIR/*.ASCII $TMPDIR/chron*.txt $TMPDIR/*.url $TMPDIR/*.line $TMPDIR/*.line2 $TMPDIR/ascii*