Colin's git trees - cmccabe-bin/blob - pdfgrep.sh

   1 #!/bin/bash
   2
   3 #
   4 # pdfgrep
   5 #
   6 # A script for searching through PDFs to find some words.
   7 #
   8 # I guess that everyone knows that PDFs aren't greppable. However, if you use
   9 # this script, you can grep through PDFs for various key words, just as if
  10 # they were text.
  11 #
  12 # usage: pdfgrep [pattern] [PDFs to search ...]
  13 #
  14 # Colin McCabe
  15 #
  16
  17 die() {
  18     echo $1
  19     exit 1
  20 }
  21
  22 ispdf() {
  23     echo ${1} | grep -q -i -E '.pdf$'
  24 }
  25
  26 which pdftotext &> /dev/null || die "you must have pdftotext installed"
  27 which fold &> /dev/null || die "you must have fold installed"
  28
  29 PATTERN=$1
  30 shift
  31 #echo "PATTERN=${PATTERN}"
  32
  33 TMPDIR=`mktemp -d -t pdfgrep.XXXXXXXXXX` || exit 1
  34 trap "rm -rf ${TMPDIR}; exit" INT TERM EXIT
  35
  36 for PDF in "$@"; do
  37     if ispdf "${PDF}"; then
  38         #echo "PDF = $PDF"
  39         pdftotext "${PDF}" "${TMPDIR}/${PDF}.pre.txt"
  40         fold -s -w 120 "${TMPDIR}/${PDF}.pre.txt" > "${TMPDIR}/${PDF}.txt"
  41         grep --with-filename -i --color=always \
  42             ${PATTERN} "${TMPDIR}/${PDF}.txt" | \
  43             sed "s ^${TMPDIR}/  "
  44         rm -f "${TMPDIR}/${PDF}.txt"
  45     fi
  46 done
  47
  48 exit 0