#!/bin/sh

# ================================================================
set -e

# For building with autoconf:
# * in-directory build:
#   pwd                  is /path/to/the/tree/c/reg_test
#   path to mlr          is /path/to/the/tree/c/mlr
#   path to reg_test/run is /path/to/the/tree/c/reg_test/run
#
# * out-of-directory ("VPATH") build:
#   pwd                  is /path/to/build/tree/c/reg_test
#   path to mlr          is /path/to/build/tree/c/mlr
#   path to reg_test/run is /path/to/source/tree/c/reg_test/run

# For building without autoconf:
#   pwd                  is /does/not/matter
#   path to mlr          is /path/to/the/tree/c/mlr
#   path to reg_test/run is /path/to/the/tree/c/reg_test/run

ourdir=`dirname $0`
srcdir=$ourdir/../..
pwd=`pwd`

try1=$pwd/../mlr    # For autoconf builds, in-tree or out-of-tree
try2=$srcdir/c/mlr  # For non-autoconf builds
if [ -x "$try1" ]; then
  path_to_mlr=$try1
elif [ -x "$try2" ]; then
  path_to_mlr=$try2
else
  echo "$0: Could not find path to mlr executable." 1>&2
  echo "Try 1: $try1" 1>&2
  echo "Try 2: $try2" 1>&2
  exit 1
fi
echo Using mlr executable $path_to_mlr

indir=$ourdir/input
expdir=$ourdir/expected
outdir=$pwd/output
mkdir -p $outdir

rm -f $outdir/out
touch $outdir/out
echo

num_passed=0

announce() {
	echo >> $outdir/out
	echo "================================================================" >> $outdir/out
	echo "$@" >> $outdir/out
	echo >> $outdir/out
}

# A key feature of this regression script is that it can be invoked from any
# directory. Depending on the directory it's invoked from, the path to the mlr
# executable may vary.  Nonetheless for debugging it's crucial that we echo out
# each command being executed. See also diff -I below.
run_mlr() {
  # Use just "mlr" for info messages
	echo mlr "$@"
	echo mlr "$@" >> $outdir/out
  # Use path to mlr for invoking the command
	$path_to_mlr "$@" >> $outdir/out
	echo >> $outdir/out
	# since set -e
	num_passed=`expr $num_passed + 1`
}

# ================================================================
announce STATELESS MAPPERS

run_mlr cat $indir/abixy
run_mlr cat /dev/null

run_mlr cut -f a,x $indir/abixy
run_mlr cut --complement -f a,x $indir/abixy

run_mlr cut -r    -f c,e         $indir/having-fields-regex.dkvp
run_mlr cut -r    -f '"C","E"'   $indir/having-fields-regex.dkvp
run_mlr cut -r    -f '"c"i,"e"'  $indir/having-fields-regex.dkvp
run_mlr cut -r    -f '"C"i,"E"'  $indir/having-fields-regex.dkvp
run_mlr cut -r -x -f c,e         $indir/having-fields-regex.dkvp
run_mlr cut -r -x -f '"C","E"'   $indir/having-fields-regex.dkvp
run_mlr cut -r -x -f '"C","E"i'  $indir/having-fields-regex.dkvp
run_mlr cut -r -x -f '"c","e"i'  $indir/having-fields-regex.dkvp

run_mlr --csvlite cut -r -f '^Name$,^Date_[0-9].*$' $indir/date1.csv $indir/date2.csv

run_mlr having-fields --at-least  a,b         $indir/abixy
run_mlr having-fields --at-least  a,c         $indir/abixy
run_mlr having-fields --at-least  a,b,i,x,y   $indir/abixy
run_mlr having-fields --which-are a,b,i,x     $indir/abixy
run_mlr having-fields --which-are a,b,i,x,y   $indir/abixy
run_mlr having-fields --which-are a,b,i,y,x   $indir/abixy
run_mlr having-fields --which-are a,b,i,x,w   $indir/abixy
run_mlr having-fields --which-are a,b,i,x,y,z $indir/abixy
run_mlr having-fields --at-most   a,c         $indir/abixy
run_mlr having-fields --at-most   a,b,i,x,y   $indir/abixy
run_mlr having-fields --at-most   a,b,i,x,y,z $indir/abixy

run_mlr having-fields --all-matching  '"^[a-z][a-z][a-z]$"'  $indir/having-fields-regex.dkvp
run_mlr having-fields --any-matching  '"^[a-z][a-z][a-z]$"'  $indir/having-fields-regex.dkvp
run_mlr having-fields --none-matching '"^[a-z][a-z][a-z]$"'  $indir/having-fields-regex.dkvp
run_mlr having-fields --all-matching  '"^[a-z][a-z][a-z]$"i' $indir/having-fields-regex.dkvp
run_mlr having-fields --any-matching  '"^[a-z][a-z][a-z]$"i' $indir/having-fields-regex.dkvp
run_mlr having-fields --none-matching '"^[a-z][a-z][a-z]$"i' $indir/having-fields-regex.dkvp

run_mlr rename b,BEE,x,EKS $indir/abixy
run_mlr rename nonesuch,nonesuch,x,EKS $indir/abixy

run_mlr --csvlite rename -r    '^Date_[0-9].*$,Date'  $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r    '(.*)e(.*),\1EEE\2'    $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r    '"(.*)e(.*)"i,\1EEE\2' $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r -g '"(.*)e(.*)"i,\1EEE\2' $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r    '"e",EEE'              $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r -g '"e",EEE'              $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r    '"e"i,EEE'             $indir/date1.csv $indir/date2.csv
run_mlr --csvlite rename -r -g '"e"i,EEE'             $indir/date1.csv $indir/date2.csv

run_mlr regularize $indir/regularize.dkvp

# ----------------------------------------------------------------
announce TRIVIAL RETAINERS

run_mlr group-by a   $indir/abixy
run_mlr group-by a,b $indir/abixy

run_mlr group-like $indir/het.dkvp

run_mlr tac $indir/abixy
run_mlr tac /dev/null

# ----------------------------------------------------------------
announce HEAD/TAIL/ETC.

run_mlr head -n 2        $indir/abixy-het
run_mlr head -n 2 -g a   $indir/abixy-het
run_mlr head -n 2 -g a,b $indir/abixy-het

run_mlr tail -n 2        $indir/abixy-het
run_mlr tail -n 2 -g a   $indir/abixy-het
run_mlr tail -n 2 -g a,b $indir/abixy-het

run_mlr top -f x,y -n 2        $indir/abixy-het
run_mlr top -f x,y -n 2 -g a   $indir/abixy-het
run_mlr top -f x,y -n 2 -g a,b $indir/abixy-het

run_mlr --seed 12345 sample -k 2        $indir/abixy-het
run_mlr --seed 12345 sample -k 2 -g a   $indir/abixy-het
run_mlr --seed 12345 sample -k 2 -g a,b $indir/abixy-het

run_mlr uniq    -g a   $indir/abixy-het
run_mlr uniq    -g a,b $indir/abixy-het

run_mlr uniq -c -g a   $indir/abixy-het
run_mlr uniq -c -g a,b $indir/abixy-het

run_mlr count-distinct -f a   $indir/small $indir/abixy
run_mlr count-distinct -f a,b $indir/small $indir/abixy

# ----------------------------------------------------------------
announce SORT

run_mlr sort -f a   $indir/abixy
run_mlr sort -r a   $indir/abixy
run_mlr sort -f x   $indir/abixy
run_mlr sort -r x   $indir/abixy
run_mlr sort -nf x  $indir/abixy
run_mlr sort -nr x  $indir/abixy

run_mlr sort -f a,b   $indir/abixy
run_mlr sort -r a,b   $indir/abixy
run_mlr sort -f x,y   $indir/abixy
run_mlr sort -r x,y   $indir/abixy
run_mlr sort -nf x,y  $indir/abixy
run_mlr sort -nr x,y  $indir/abixy

run_mlr sort -f a -nr x $indir/abixy
run_mlr sort -nr y -f a $indir/abixy
run_mlr sort -f a -r b -nf x -nr y $indir/abixy

run_mlr sort -f x $indir/sort-het.dkvp
run_mlr sort -r x $indir/sort-het.dkvp

# ----------------------------------------------------------------
announce TOP

run_mlr top    -n 4 -f x        $indir/abixy-wide
run_mlr top    -n 1 -f x,y      $indir/abixy-wide
run_mlr top    -n 4 -f x   -g a $indir/abixy-wide
run_mlr top    -n 1 -f x,y -g a $indir/abixy-wide
run_mlr top -a -n 4 -f x        $indir/abixy-wide
run_mlr top -a -n 4 -f x   -g a $indir/abixy-wide

# ----------------------------------------------------------------
announce JOIN

run_mlr --opprint join                   -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u                -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join         --ul      -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u      --ul      -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join              --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u           --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --ul      --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u      --ul --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np --ul      -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np --ul      -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np      --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np      --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np --ul --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np --ul --ur -f $indir/joina.dkvp -l l -r r -j o $indir/joinb.dkvp


run_mlr --opprint join                   -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u                -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join         --ul      -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u      --ul      -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join              --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u           --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join         --ul --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u      --ul --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np --ul      -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np --ul      -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np      --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np      --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp

run_mlr --opprint join    --np --ul --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp
run_mlr --opprint join -u --np --ul --ur -f /dev/null -l l -r r -j o $indir/joinb.dkvp


run_mlr --opprint join                   -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u                -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join         --ul      -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u      --ul      -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join              --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u           --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join         --ul --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u      --ul --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join    --np --ul      -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u --np --ul      -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join    --np      --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u --np      --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --opprint join    --np --ul --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null
run_mlr --opprint join -u --np --ul --ur -f $indir/joina.dkvp -l l -r r -j o /dev/null

run_mlr --odkvp join -u -j a -f $indir/join-het.dkvp $indir/abixy-het
run_mlr --odkvp join -u -j a -f $indir/abixy-het     $indir/join-het.dkvp
run_mlr --odkvp join -u --np --ul --ur -j a -f $indir/join-het.dkvp $indir/abixy-het
run_mlr --odkvp join -u --np --ul --ur -j a -f $indir/abixy-het     $indir/join-het.dkvp

# ----------------------------------------------------------------
announce STATS

run_mlr --opprint stats1    -a mean,sum,count,min,max,mode -f i,x,y $indir/abixy
run_mlr --opprint stats1    -a min,p10,p50,mode,p90,max    -f i,x,y $indir/abixy
run_mlr --opprint stats1    -a mean,meaneb,stddev          -f i,x,y $indir/abixy
run_mlr --oxtab   stats1 -s -a mean,sum,count,min,max,mode -f i,x,y $indir/abixy

run_mlr --opprint stats1    -a mean,sum,count,min,max,mode -f i,x,y -g a $indir/abixy
run_mlr --opprint stats1    -a min,p10,p50,mode,p90,max    -f i,x,y -g a $indir/abixy
run_mlr --opprint stats1    -a mean,meaneb,stddev          -f i,x,y -g a $indir/abixy
run_mlr --oxtab   stats1 -s -a mean,sum,count,min,max,mode -f i,x,y -g a $indir/abixy

run_mlr --opprint stats1    -a mean,sum,count,min,max,mode -f i,x,y -g a,b $indir/abixy
run_mlr --opprint stats1    -a min,p10,p50,mode,p90,max    -f i,x,y -g a,b $indir/abixy
run_mlr --opprint stats1    -a mean,meaneb,stddev          -f i,x,y -g a,b $indir/abixy
run_mlr --oxtab   stats1 -s -a mean,sum,count,min,max,mode -f i,x,y -g a,b $indir/abixy

run_mlr --oxtab   stats1 -a mean -f x      $indir/abixy-het
run_mlr --oxtab   stats1 -a mean -f x -g a $indir/abixy-het

run_mlr --opprint stats2       -a linreg-ols,linreg-pca,r2,corr,cov -f x,y,xy,y2,x2,x2        $indir/abixy-wide
run_mlr --opprint stats2       -a linreg-ols,linreg-pca,r2,corr,cov -f x,y,xy,y2,x2,x2 -g a,b $indir/abixy-wide
run_mlr --oxtab   stats2 -s    -a linreg-ols,linreg-pca,r2,corr,cov -f x,y,xy,y2,x2,x2        $indir/abixy-wide-short
run_mlr --oxtab   stats2 -s    -a linreg-ols,linreg-pca,r2,corr,cov -f x,y,xy,y2,x2,x2 -g a,b $indir/abixy-wide-short
run_mlr --opprint stats2 --fit -a linreg-ols,linreg-pca             -f x,y,xy,y2,x2,x2        $indir/abixy-wide-short
run_mlr --opprint stats2 --fit -a linreg-ols,linreg-pca             -f x,y,xy,y2,x2,x2 -g a   $indir/abixy-wide-short

run_mlr --opprint stats2    -a logireg -f x,y      $indir/logi.dkvp
run_mlr --opprint stats2    -a logireg -f x,y -g g $indir/logi.dkvp

run_mlr --oxtab   stats2 -a cov -f x,y      $indir/abixy-het
run_mlr --oxtab   stats2 -a cov -f x,y -g a $indir/abixy-het

run_mlr --opprint step -a rsum,delta,counter -f x,y $indir/abixy
run_mlr --opprint step -a rsum,delta,counter -f x,y -g a $indir/abixy

run_mlr --odkvp   step -a rsum,delta,counter -f x,y      $indir/abixy-het
run_mlr --odkvp   step -a rsum,delta,counter -f x,y -g a $indir/abixy-het

run_mlr --opprint histogram -f x,y --lo 0 --hi 1 --nbins 20 $indir/small

# ----------------------------------------------------------------
announce DSLs

run_mlr filter '$x>.3'    $indir/abixy
run_mlr filter '$x>0.3'   $indir/abixy
run_mlr filter '$x>0.3 && $y>0.3'   $indir/abixy
run_mlr filter '$x>0.3 || $y>0.3'   $indir/abixy
run_mlr filter 'NR>=4 && NR <= 7'   $indir/abixy

run_mlr filter -x '$x>.3'    $indir/abixy
run_mlr filter -x '$x>0.3'   $indir/abixy
run_mlr filter -x '$x>0.3 && $y>0.3'   $indir/abixy
run_mlr filter -x '$x>0.3 || $y>0.3'   $indir/abixy
run_mlr filter -x 'NR>=4 && NR <= 7'   $indir/abixy

run_mlr filter '$nosuchfield>.3'    $indir/abixy

run_mlr put '$x2 = $x**2' $indir/abixy
run_mlr put '$z = -0.024*$x+0.13' $indir/abixy
run_mlr put '$c = $a . $b' $indir/abixy

run_mlr --opprint put '$nr=NR;$fnr=FNR;$nf=NF;$filenum=FILENUM' $indir/abixy $indir/abixy

announce OPERATOR PRECEDENCE AND ASSOCIATIVITY

# filter -v and put -v print the AST
run_mlr put -v '$x=$a+$b+$c'   /dev/null
run_mlr put -v '$x=$a+$b-$c'   /dev/null
run_mlr put -v '$x=$a-$b-$c'   /dev/null
run_mlr put -v '$x=$a-$b+$c'   /dev/null

run_mlr put -v '$x=$a*$b*$c'   /dev/null
run_mlr put -v '$x=$a*$b/$c'   /dev/null
run_mlr put -v '$x=$a/$b/$c'   /dev/null
run_mlr put -v '$x=$a/$b*$c'   /dev/null

run_mlr put -v '$x=$a+$b+$c'   /dev/null
run_mlr put -v '$x=$a+$b*$c'   /dev/null
run_mlr put -v '$x=$a*$b*$c'   /dev/null
run_mlr put -v '$x=$a*$b+$c'   /dev/null

run_mlr put -v '$x=$a+$b+$c'   /dev/null
run_mlr put -v '$x=$a+$b**$c'  /dev/null
run_mlr put -v '$x=$a**$b**$c' /dev/null
run_mlr put -v '$x=$a**$b+$c'  /dev/null

run_mlr put -v '$x=$a.$b.$c'   /dev/null

run_mlr put -v '$x=-$a+$b*$c'  /dev/null
run_mlr put -v '$x=-$a*$b+$c'  /dev/null
run_mlr put -v '$x=$a+-$b*$c'  /dev/null
run_mlr put -v '$x=$a*-$b+$c'  /dev/null
run_mlr put -v '$x=$a+$b*-$c'  /dev/null
run_mlr put -v '$x=$a*$b+-$c'  /dev/null


run_mlr filter -v '$a==1 && $b == 1 && $c == 1'   /dev/null
run_mlr filter -v '$a==1 || $b == 1 && $c == 1'   /dev/null
run_mlr filter -v '$a==1 || $b == 1 || $c == 1'   /dev/null
run_mlr filter -v '$a==1 && $b == 1 || $c == 1'   /dev/null

run_mlr filter -v '$x<$a*$b*$c'   /dev/null
run_mlr filter -v '$x<$a*$b/$c'   /dev/null
run_mlr filter -v '$x<$a/$b/$c'   /dev/null
run_mlr filter -v '$x<$a/$b*$c'   /dev/null

run_mlr filter -v '$x<$a+$b+$c'   /dev/null
run_mlr filter -v '$x<$a+$b*$c'   /dev/null
run_mlr filter -v '$x<$a*$b*$c'   /dev/null
run_mlr filter -v '$x<$a*$b+$c'   /dev/null

run_mlr filter -v '$x<$a+$b+$c'   /dev/null
run_mlr filter -v '$x<$a+$b**$c'  /dev/null
run_mlr filter -v '$x<$a**$b**$c' /dev/null
run_mlr filter -v '$x<$a**$b+$c'  /dev/null

run_mlr filter -v '$x<$a.$b.$c'   /dev/null

run_mlr filter -v '$x<-$a+$b*$c'  /dev/null
run_mlr filter -v '$x<-$a*$b+$c'  /dev/null
run_mlr filter -v '$x<$a+-$b*$c'  /dev/null
run_mlr filter -v '$x<$a*-$b+$c'  /dev/null
run_mlr filter -v '$x<$a+$b*-$c'  /dev/null
run_mlr filter -v '$x<$a*$b+-$c'  /dev/null

run_mlr filter -v '$x =~ "bcd"'       $indir/regex.dkvp
run_mlr filter -v '$x =~ "^bcd"'      $indir/regex.dkvp
run_mlr filter -v '$x =~ "^abc"'      $indir/regex.dkvp
run_mlr filter -v '$x =~ "^abc$"'     $indir/regex.dkvp
run_mlr filter -v '$x =~ "^a.*d$"'    $indir/regex.dkvp
run_mlr filter -v '$x =~ "^a.*"."d$"' $indir/regex.dkvp
run_mlr filter -v '$y =~ "\".."'      $indir/regex.dkvp

run_mlr filter -v '$x =~ "bcd"i'       $indir/regex.dkvp
run_mlr filter -v '$x =~ "^bcd"i'      $indir/regex.dkvp
run_mlr filter -v '$x =~ "^abc"i'      $indir/regex.dkvp
run_mlr filter -v '$x =~ "^abc$"i'     $indir/regex.dkvp
run_mlr filter -v '$x =~ "^a.*d$"i'    $indir/regex.dkvp
run_mlr filter -v '$x =~ "^a.*"."d$"i' $indir/regex.dkvp

run_mlr --csvlite put '$gmt=sec2gmt($sec)' $indir/sec2gmt
run_mlr --csvlite put '$sec=gmt2sec($gmt)' $indir/gmt2sec

run_mlr put '$z=min($x, $y)' $indir/minmax.dkvp
run_mlr put '$z=max($x, $y)' $indir/minmax.dkvp

run_mlr --opprint put '$hms=sec2hms($sec);   $resec=hms2sec($hms);   $diff=$resec-$sec' $indir/sec2xhms
run_mlr --opprint put '$hms=fsec2hms($sec);  $resec=hms2fsec($hms);  $diff=$resec-$sec' $indir/fsec2xhms
run_mlr --opprint put '$hms=sec2dhms($sec);  $resec=dhms2sec($hms);  $diff=$resec-$sec' $indir/sec2xhms
run_mlr --opprint put '$hms=fsec2dhms($sec); $resec=dhms2fsec($hms); $diff=$resec-$sec' $indir/fsec2xhms

announce SUB/GSUB

run_mlr --opprint put '$y = sub($x, "e.*l",        "y123y")'      $indir/sub.dat
run_mlr --opprint put '$y = sub($x, "e.*l"i,       "y123y")'      $indir/sub.dat
run_mlr --opprint put '$y = sub($x, "e.*"."l",     "y123y")'      $indir/sub.dat
run_mlr --opprint put '$y = sub($x, "([hg])e.*l(.)", "y\1y123\2y")' $indir/sub.dat
run_mlr --opprint put '$y = sub($x, "([hg])e.*l.",   "y\1y123\2y")' $indir/sub.dat
run_mlr --opprint put '$y = sub($x, "([hg])e.*l(.)", "y\1y123.y")'  $indir/sub.dat

run_mlr --opprint put '$y = sub($x,  "a",    "aa")'   $indir/gsub.dat
run_mlr --opprint put '$y = gsub($x, "a",    "aa")'   $indir/gsub.dat
run_mlr --opprint put '$y = gsub($x, "A",    "Aa")'   $indir/gsub.dat
run_mlr --opprint put '$y = gsub($x, "a"i,   "Aa")'   $indir/gsub.dat
run_mlr --opprint put '$y = gsub($x, "A"i,   "Aa")'   $indir/gsub.dat
run_mlr --opprint put '$y = gsub($x, "a(.)", "aa\1\1\1")' $indir/gsub.dat

# $ echo 'x=<abcabc_abcabc>'|mlr put '$y=gsub($x,"(.)b(.)","[\1X\2]")'
# x=<abcabc_abcabc>,y=<[aXc][aXc]_[aXc][aXc]>

# $ echo 'x=<abcabc_abcab>'|mlr put '$y=gsub($x,"(.)b(.)","[\1X\2]")'
# x=<abcabc_abcab>,y=<[aXc][aXc]_[aXc][aX>]

# $ echo 'x=<abcabc_abcab'|mlr put '$y=gsub($x,"(.)b(.)","[\1X\2]")'
# x=<abcabc_abcab,y=<[aXc][aXc]_[aXc]ab

# ----------------------------------------------------------------
announce CHAINING

run_mlr cat then cat $indir/short
run_mlr cat then tac $indir/short
run_mlr tac then cat $indir/short
run_mlr tac then tac $indir/short

run_mlr cat then cat then cat $indir/short
run_mlr cat then cat then tac $indir/short
run_mlr cat then tac then cat $indir/short
run_mlr cat then tac then tac $indir/short
run_mlr tac then cat then cat $indir/short
run_mlr tac then cat then tac $indir/short
run_mlr tac then tac then cat $indir/short
run_mlr tac then tac then tac $indir/short

# ----------------------------------------------------------------
announce NUMBER FORMATTING

run_mlr                --opprint stats1 -a sum -f x $indir/ofmt.dat
run_mlr --ofmt '%.3lf' --opprint stats1 -a sum -f x $indir/ofmt.dat
run_mlr --opprint --ofmt '%.3lf' stats1 -a sum -f x $indir/ofmt.dat

# ----------------------------------------------------------------
announce IMPLICIT-HEADER-CSV INPUT

run_mlr --irs crlf                       --no-mmap --icsvlite --ifs ,  --opprint cut -x -f b/ $indir/multi-sep.csv
run_mlr --irs crlf --implicit-csv-header --no-mmap --icsvlite --ifs ,  --opprint cut -x -f 2  $indir/multi-sep.csv

run_mlr --irs crlf                       --no-mmap --icsvlite --ifs /, --opprint cut -x -f b  $indir/multi-sep.csv
run_mlr --irs crlf --implicit-csv-header --no-mmap --icsvlite --ifs /, --opprint cut -x -f 2  $indir/multi-sep.csv

run_mlr --irs crlf                       --mmap    --icsvlite --ifs ,  --opprint cut -x -f b/ $indir/multi-sep.csv
run_mlr --irs crlf --implicit-csv-header --mmap    --icsvlite --ifs ,  --opprint cut -x -f 2  $indir/multi-sep.csv

run_mlr --irs crlf                       --mmap    --icsvlite --ifs /, --opprint cut -x -f b  $indir/multi-sep.csv
run_mlr --irs crlf --implicit-csv-header --mmap    --icsvlite --ifs /, --opprint cut -x -f 2  $indir/multi-sep.csv

run_mlr                       --icsv --ifs ,  --opprint cut -x -f b/ $indir/multi-sep.csv
run_mlr --implicit-csv-header --icsv --ifs ,  --opprint cut -x -f 2  $indir/multi-sep.csv

run_mlr                       --icsv --ifs /, --opprint cut -x -f b  $indir/multi-sep.csv
run_mlr --implicit-csv-header --icsv --ifs /, --opprint cut -x -f 2  $indir/multi-sep.csv


# ----------------------------------------------------------------
announce HET-CSV INPUT

run_mlr --icsvlite --odkvp cat $indir/a.csv
run_mlr --icsvlite --odkvp cat $indir/b.csv
run_mlr --icsvlite --odkvp cat $indir/c.csv
run_mlr --icsvlite --odkvp cat $indir/d.csv
run_mlr --icsvlite --odkvp cat $indir/e.csv
run_mlr --icsvlite --odkvp cat $indir/f.csv
run_mlr --icsvlite --odkvp cat $indir/g.csv

run_mlr --icsvlite --odkvp cat $indir/a.csv $indir/a.csv
run_mlr --icsvlite --odkvp cat $indir/b.csv $indir/b.csv
run_mlr --icsvlite --odkvp cat $indir/c.csv $indir/c.csv
run_mlr --icsvlite --odkvp cat $indir/d.csv $indir/d.csv
run_mlr --icsvlite --odkvp cat $indir/e.csv $indir/e.csv
run_mlr --icsvlite --odkvp cat $indir/f.csv $indir/f.csv
run_mlr --icsvlite --odkvp cat $indir/g.csv $indir/g.csv

run_mlr --icsvlite --odkvp cat $indir/a.csv $indir/b.csv
run_mlr --icsvlite --odkvp cat $indir/b.csv $indir/c.csv
run_mlr --icsvlite --odkvp cat $indir/c.csv $indir/d.csv
run_mlr --icsvlite --odkvp cat $indir/d.csv $indir/e.csv
run_mlr --icsvlite --odkvp cat $indir/e.csv $indir/f.csv
run_mlr --icsvlite --odkvp cat $indir/f.csv $indir/g.csv

run_mlr --icsvlite --odkvp cat $indir/a.csv $indir/b.csv \
  $indir/c.csv $indir/d.csv $indir/e.csv $indir/f.csv $indir/g.csv

run_mlr --icsvlite --odkvp tac $indir/het.csv

# ----------------------------------------------------------------
announce HET-PPRINT INPUT

run_mlr --ipprint --odkvp cat $indir/a.pprint
run_mlr --ipprint --odkvp cat $indir/b.pprint
run_mlr --ipprint --odkvp cat $indir/c.pprint
run_mlr --ipprint --odkvp cat $indir/d.pprint
run_mlr --ipprint --odkvp cat $indir/e.pprint
run_mlr --ipprint --odkvp cat $indir/f.pprint
run_mlr --ipprint --odkvp cat $indir/g.pprint

run_mlr --ipprint --odkvp cat $indir/a.pprint $indir/a.pprint
run_mlr --ipprint --odkvp cat $indir/b.pprint $indir/b.pprint
run_mlr --ipprint --odkvp cat $indir/c.pprint $indir/c.pprint
run_mlr --ipprint --odkvp cat $indir/d.pprint $indir/d.pprint
run_mlr --ipprint --odkvp cat $indir/e.pprint $indir/e.pprint
run_mlr --ipprint --odkvp cat $indir/f.pprint $indir/f.pprint
run_mlr --ipprint --odkvp cat $indir/g.pprint $indir/g.pprint

run_mlr --ipprint --odkvp cat $indir/a.pprint $indir/b.pprint
run_mlr --ipprint --odkvp cat $indir/b.pprint $indir/c.pprint
run_mlr --ipprint --odkvp cat $indir/c.pprint $indir/d.pprint
run_mlr --ipprint --odkvp cat $indir/d.pprint $indir/e.pprint
run_mlr --ipprint --odkvp cat $indir/e.pprint $indir/f.pprint
run_mlr --ipprint --odkvp cat $indir/f.pprint $indir/g.pprint

run_mlr --ipprint --odkvp cat $indir/a.pprint $indir/b.pprint \
  $indir/c.pprint $indir/d.pprint $indir/e.pprint $indir/f.pprint $indir/g.pprint

# ================================================================
announce NULL-FIELD INPUT

run_mlr --icsvlite --odkvp cat $indir/null-fields.csv
run_mlr --inidx --ifs comma --odkvp cat $indir/null-fields.nidx
run_mlr --idkvp --oxtab cat $indir/missings.dkvp

# ================================================================
announce SPACE-PADDING

run_mlr --mmap    --idkvp    --odkvp --ifs space --repifs cat $indir/space-pad.dkvp
run_mlr --no-mmap --idkvp    --odkvp --ifs space --repifs cat $indir/space-pad.dkvp
run_mlr --mmap    --inidx    --odkvp --ifs space --repifs cat $indir/space-pad.nidx
run_mlr --no-mmap --inidx    --odkvp --ifs space --repifs cat $indir/space-pad.nidx
run_mlr --mmap    --icsvlite --odkvp --ifs space --repifs cat $indir/space-pad.pprint
run_mlr --no-mmap --icsvlite --odkvp --ifs space --repifs cat $indir/space-pad.pprint

# ================================================================
announce DOUBLE PS

run_mlr --no-mmap --opprint cat $indir/double-ps.dkvp
run_mlr    --mmap --opprint cat $indir/double-ps.dkvp

# ================================================================
announce MISSING FINAL LF

run_mlr --no-mmap --csvlite cat $indir/truncated.csv
run_mlr           --csvlite cat $indir/truncated.csv
run_mlr --no-mmap --dkvp    cat $indir/truncated.dkvp
run_mlr           --dkvp    cat $indir/truncated.dkvp
run_mlr --no-mmap --nidx    cat $indir/truncated.nidx
run_mlr           --nidx    cat $indir/truncated.nidx
run_mlr --no-mmap --pprint  cat $indir/truncated.pprint
run_mlr           --pprint  cat $indir/truncated.pprint
run_mlr --no-mmap --xtab    cat $indir/truncated.xtab
run_mlr           --xtab    cat $indir/truncated.xtab

# ================================================================
announce UTF-8 alignment

run_mlr --icsvlite --opprint cat $indir/utf8-1.csv
run_mlr --icsvlite --opprint cat $indir/utf8-2.csv
run_mlr --icsvlite --oxtab   cat $indir/utf8-1.csv
run_mlr --icsvlite --oxtab   cat $indir/utf8-2.csv

run_mlr --inidx --ifs space --opprint         cat $indir/utf8-align.nidx
run_mlr --inidx --ifs space --opprint --right cat $indir/utf8-align.nidx
run_mlr --oxtab cat $indir/utf8-align.dkvp 

# ================================================================
announce STDIN

run_mlr --csv cat < $indir/rfc-csv/simple.csv

# ================================================================
announce RFC-CSV

run_mlr --csv cat $indir/rfc-csv/simple.csv
run_mlr --csv cat $indir/rfc-csv/simple-truncated.csv
run_mlr --csv cat $indir/rfc-csv/narrow.csv
run_mlr --csv cat $indir/rfc-csv/narrow-truncated.csv
run_mlr --csv cat $indir/rfc-csv/quoted-comma.csv
run_mlr --csv cat $indir/rfc-csv/quoted-comma-truncated.csv
run_mlr --csv cat $indir/rfc-csv/quoted-crlf.csv
run_mlr --csv cat $indir/rfc-csv/quoted-crlf-truncated.csv
run_mlr --csv cat $indir/rfc-csv/simple-truncated.csv $indir/rfc-csv/simple.csv
run_mlr --csv --ifs semicolon --ofs pipe --irs lf --ors lflf cut -x -f b $indir/rfc-csv/modify-defaults.csv

# ================================================================
announce MULTI-CHARACTER IRS/IFS/IPS FOR DKVP

run_mlr --oxtab --idkvp --mmap    --irs lf   --ifs ,  --ips =  cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --idkvp --mmap    --irs lf   --ifs /, --ips =: cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --idkvp --mmap    --irs crlf --ifs ,  --ips =  cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --idkvp --mmap    --irs crlf --ifs /, --ips =: cut -o -f x,a,i $indir/multi-sep.dkvp

run_mlr --oxtab --no-mmap --irs lf   --ifs ,  --ips =  cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --no-mmap --irs lf   --ifs /, --ips =: cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --no-mmap --irs crlf --ifs ,  --ips =  cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --no-mmap --irs crlf --ifs /, --ips =: cut -o -f x,a,i $indir/multi-sep.dkvp

# ================================================================
announce MULTI-CHARACTER IRS/IFS FOR NIDX

run_mlr --oxtab --inidx --mmap    --irs lf   --ifs ,  cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --mmap    --irs lf   --ifs /, cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --mmap    --irs crlf --ifs ,  cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --mmap    --irs crlf --ifs /, cut -o -f 4,1,3 $indir/multi-sep.dkvp

run_mlr --oxtab --inidx --no-mmap --irs lf   --ifs ,  cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --no-mmap --irs lf   --ifs /, cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --no-mmap --irs crlf --ifs ,  cut -o -f 4,1,3 $indir/multi-sep.dkvp
run_mlr --oxtab --inidx --no-mmap --irs crlf --ifs /, cut -o -f 4,1,3 $indir/multi-sep.dkvp

# ================================================================
announce MULTI-CHARACTER IRS/IFS FOR CSVLITE

run_mlr --oxtab --icsvlite --mmap    --irs lf   --ifs ,  cut -o -f x/,a/,i/ $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --mmap    --irs lf   --ifs /, cut -o -f x,a,i    $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --mmap    --irs crlf --ifs ,  cut -o -f x/,a/,i/ $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --mmap    --irs crlf --ifs /, cut -o -f x,a,i    $indir/multi-sep.csv

run_mlr --oxtab --icsvlite --no-mmap --irs lf   --ifs ,  cut -o -f x/,a/,i/ $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --no-mmap --irs lf   --ifs /, cut -o -f x,a,i    $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --no-mmap --irs crlf --ifs ,  cut -o -f x/,a/,i/ $indir/multi-sep.csv
run_mlr --oxtab --icsvlite --no-mmap --irs crlf --ifs /, cut -o -f x,a,i    $indir/multi-sep.csv

# ================================================================
announce MULTI-CHARACTER SEPARATORS FOR XTAB

run_mlr --mmap    --xtab --ifs crlf --ofs Z cut -x -f b $indir/truncated.xtab-crlf
run_mlr --no-mmap --xtab --ifs crlf --ofs Z cut -x -f b $indir/truncated.xtab-crlf
run_mlr --mmap    --xtab --ips . --ops @ cut -x -f b $indir/dots.xtab
run_mlr --no-mmap --xtab --ips . --ops @ cut -x -f b $indir/dots.xtab
run_mlr --no-mmap --xtab --ips . --ops := cut -x -f b $indir/dots.xtab
run_mlr --xtab --ips ": " --ops '@@@@' put '$sum=int($a+$b)' $indir/multi-ips.dkvp

# ================================================================
announce MULTI-CHARACTER IRS FOR PPRINT

run_mlr --mmap    --pprint --irs crlf --ifs / --ofs @ cut -x -f b $indir/dots.pprint
run_mlr --no-mmap --pprint --irs crlf --ifs / --ofs @ cut -x -f b $indir/dots.pprint

# ================================================================
announce MULTI-CHARACTER IXS SPECIFIERS

run_mlr --oxtab --idkvp --mmap  --irs lf   --ifs '\x2c'  --ips '\075'  cut -o -f x,a,i $indir/multi-sep.dkvp
run_mlr --oxtab --idkvp --mmap  --irs lf   --ifs /, --ips '\x3d\x3a' cut -o -f x,a,i $indir/multi-sep.dkvp

# ================================================================
# A key feature of this regression script is that it can be invoked from any
# directory. Depending on the directory it's invoked from, the path to $outdir
# may vary. Nonetheless for debugging it's crucial that we echo out each
# command being executed -- here we use diff -I to skip the info lines and
# focus the test on program output per se.
diff -I '^mlr ' -C5 $expdir/out $outdir/out

# ================================================================
echo ALL REGRESSION TESTS PASSED
echo Tests completed: $num_passed
