peck-filter

#!/bin/bash
#
# Filter the the output of a peck search by a second set of search criteria
#
# Written by FFS
#
# Changelog:
#
#       2015-01-10 Filter a search
#
# -----------------------------------------------------------

# Help screen
if [ "$1" = "-h" -o "$1" = "--help" -o "$1" = "help" ] ; then
  echo -e "\n\t\t* * * Red Hen Commandline Search Filter * * *"
  echo -e "\n\tFilter a peck search result by a second set of search criteria."
  echo -e "\n\tSee peck -h and peck -h2 for basic instructions."
  echo -e "\n\tSyntax:"
  echo -e "\n\t\t`basename $0` <peck search result csv> <file type> <\"regex search terms\"> <primary tag> <output file> [clobber]"
  echo -e "\n\tExamples (start with a peck search results and refine it):"
  echo -e "\n\t\t`basename $0` ~/MyResults1.csv seg \"a\|[a-zA-Z]+/JJ\" POS_01 ~/FilteredResults.csv"
  echo -e "\n\tThe script produces results that match both set of criteria."
  echo -e "\n\tSee also peck, peck-clip, peck-intersect, and peck-seg.\n"
   exit
fi

# Search result file
if [ -z "$1" ]
  then echo -e "\nPlease include the filename of peck search result to filter; see `basename $0` -h for usage.\n" ; exit
  else INFIL="$1"
fi

# File type
if [ -z "$2" ]
  then echo -e "\n\tPlease include file type, a search term, a tag type, and an output file name!\n" ; exit
  else EXT="$2"
fi ; N=0

# Regex search term
if [ -z "$3" ]
  then echo -e "\n\tPlease include file type, a search term, a tag type, and an output file name!\n" ; exit
  else CUE="$3"
fi

# Current directory
DIR=`pwd` DAT=${DIR##*/}

# Primary tag
if [ -n "$4" ]
  then TAG="$4"
  else TAG="POS_01"
fi

# Filtered output file
if [ -n "$5" ]
  then OUTFIL="$5"
    if [ -f "$OUTFIL" -a -z "$6" ]
      then OUTFIL=${OUTFIL}_${DAT}_$(date +%Y-%m-%d-%H%M%S).csv ; echo -e "\n\tOutput file renamed $OUTFIL to avoid overwriting."
    fi
  else echo -e "\n\tPlease enter an output file name after the search term.\n" ; exit
fi ; N=0 max=0

# Clobber?
if [ -n "$6" ] ; then
  if [ "$6" = "clobber" ] ; then
    if [ -f "$OUTFIL" ] ; then rm "$OUTFIL" ; fi
  fi
fi

# Working file
WORK=/tmp/peck-filter-$$

# Welcome
echo -e "\n\tRed Hen is getting ready to filter the peck search results in $INFIL with a second set of criteria -"
echo -e "\n\tRed Hen goes looking for '$CUE' among $TAG annotations in $EXT files on $DAT ...\n"

# Number of lines
LEN="$( cat $INFIL | wc -l )"

#set -xv

# Loop through the lines, skipping the header
for NUM in `seq 2 $LEN` ; do unset FIL LIN TIM

  LIN="$( sed -n "$NUM p" $INFIL )"
  IFS="|" read FIL o o o TIM o <<< "$LIN"

  # The first filetype searched was ocr, round down to the nearest minute
  if [ "${FIL#*.}" = "ocr" ] ; then TIM=${TIM:0:12} ; fi

  # Perform the search
  HIT="$( peck ${FIL%.*}.$EXT "$CUE" $TAG $WORK clobber | grep $TIM )"
  if [ "$HIT" ] ; then echo -e "$LIN\n$HIT\n" ; echo "$LIN" >> $OUTFIL ; fi

done

# Deduplicate
if [ -f $OUTFIL ] ; then sort -u $OUTFIL | sponge $OUTFIL ; else echo -e "\n\tThe search produced no results\n" ; exit ; fi

# Add field header for R
LIN="$( sed -n "1 p" $INFIL )" ; sed -i "1i$LIN" $OUTFIL

# Display the results
#cat $OUTFIL

# Clean up
rm -f $WORK

# Receipt
echo -e "\n\tRed Hen generated filtered results for $INFIL in $OUTFIL.\n"

# EOF