Search PDFs from command line

pdfsearch.sh

#!/usr/bin/env bash
#Search pdfs: a wrapper for pdfgrep and pdftotext
#http://github.com/darcyparker
_usage(){
local _filename=${0##*/}
cat << END_USAGE
Usage: $_filename

  Search current directory:
  $_filename <search pattern>

  Search a specific directory:
  $_filename <directory> <search pattern>
END_USAGE
exit 2
}
_requirements(){
cat << END_REQUIREMENTS
Requirements:

  This requires pdfgrep or (pdftotext and grep)
  pdfgrep is preferred because it reports page numbers accurately

  Tip: On linux "apt-get install pdfgrep"

  Tip: On linux "pdftotext" is a part of poppler-utils or xpdf-utils
       : "apt-get install poppler-utils"
       : "apt-get install xpdf-utils"
END_REQUIREMENTS
exit 1
}

_searchWithPdfToText(){
  find "$1" -name "*.pdf" -type f -print0 | xargs -I{} -0 /usr/bin/env bash -c "pdftotext '{}' - | grep --with-filename --label='{}' --color \"$2\""
  exit $?
}

_searchWithPdfgrep(){
  #find "$1" -name "*.pdf" -type f -print0 | xargs -I{} -0 pdfgrep -H -n "$2" '{}'
  pdfgrep -H -n -r "$2" "$1"
  exit $?
}

_search(){
  #First try pdfgrep
  [ -x "$(command -v pdfgrep)" ] && _searchWithPdfgrep "$@"
  #if there is no pdfgrep, use pdftotext
  [ -x "$(command -v pdftotext)" ] && [ -x "$(command -v grep)" ] && _searchWithPdfToText "$@"
  #if pdfgrep or pdftotext and grep are not available, then show requirements
  _requirements
}
_main(){
  [ $# -eq 0 ] || [ $# -gt 2 ] && _usage "$0" $#
  [ $# -eq 1 ] && local _searchdir=$PWD && local _searchpattern=$1
  [ $# -eq 2 ] && local _searchdir=$1 && local _searchpattern=$2
  _search "$_searchdir" "$_searchpattern"
}
_main "$@"