#!/bin/bash
# PAPERWEB - GPLv3 licence
# Copyright 2016 Yves Gablin
#  This program is free software: you can redistribute it and/or modify
#  it under the terms of the GNU General Public License as published by
#  the Free Software Foundation, either version 3 of the License, or
#  (at your option) any later version.
#  This program is distributed in the hope that it will be useful,
#  but WITHOUT ANY WARRANTY; without even the implied warranty of
#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#  GNU General Public License for more details.
#  You should have received a copy of the GNU General Public License
#  along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
#   -h This help.
# Query:
#   -Q [-d <pipe-separated list of YYYYMMDD/YYYYMM/YYYY prefixes to search>]
#      [-l <pipe-separated list of labels to match (all must be found)>]
#      [-k <pipe-separated list of keywords to match (all must be found)>]
#      [-i] : case-insensitive search
#   At least -d, -l, or -k must be used.
#   Result (WSN):
#     OUTPUT   = "[" [ DOC_LIST ] "]" .
#     DOC_LIST = DOC { "," DOC } .
#     DOC      = "{""folder"":" FOLDER ",""labels"":[" [ LABELS ] "],""count"":" COUNT ",""type"":" TYPE "}" .
#     FOLDER   = json_string .
#     LABELS   = json_string { "," json_string } .
#     COUNT    = json_number .
#     TYPE     = """pdf""" | """pages""" .
# Retrieve a document's thumbnails:
#   -T <date> : the folder-name of the document
#   Result (WSN):
#     OUTPUT   = "[" THUMBS "]" .
#     THUMBS   = CONTENTS { "," CONTENTS } .
#     CONTENTS = "{""mime"":" MIME ",""data"":" B64_DATA ",""width"":" WIDTH ",""height"":" HEIGHT "}" .
#     MIME     = json_string .
#     B64_DATA = json_string .
#     WIDTH    = json_number .
#     HEIGHT   = json_number .
# Retrieve a document's page/PDF and metadata:
#   -D <date> -p <page number>
#   Result (WSN):
#     OUTPUT   = CONTENTS .
#   Width and height should be ignored for PDF contents.
# Retrieve a raw document's page/PDF without metadata:
#   -R <date> -p <page number>
#   Result: file contents.

##########  CONFIGURATION ########## 

BASE='/PATH/TO/PAPERWORK/BASE/DIRECTORY'

##### NO CHANGE PAST THIS LINE #####

# INIT
mode=
dates=()
labels=()
words=()
q_ci=
doc=
page=

# READ COMMAND LINE PARAMETERS
while getopts hQd:l:k:iT:D:p:R: opt; do case "$opt" in
h) sed -n '2,/^$/s/.//p' "$0"; exit 0 ;;
Q) mode=Q ;;
d) IFS='|' read -a dates < <(tr -dc '|[:digit:]' <<<"$OPTARG") ;;
l) IFS='|' read -a labels < <(tr -d ',"[:cntrl:]' <<<"$OPTARG") ;;
k) IFS='|' read -a words < <(tr -d ',"[:cntrl:]' <<<"$OPTARG") ;;
i) q_ci=true ;;
T|D|R)
   mode=$opt; doc=$(tr -dc '[:digit:]_' <<<"$OPTARG") ;;
p) page=$(tr -dc '[:digit:]' <<<"$OPTARG") ;;
esac; done

# CHECK COMMAND LINE PARAMETERS
case "$mode" in
Q) for ((i=${#dates[*]}-1;i>=0;i--)); do [ ${#dates[i]} -ge 4 ] || unset dates[i]; done
   [ -n "${dates[*]}${labels[*]}${words[*]}" ] || exit 2 ;;
T) [ -n "$doc" -a -d "$BASE/$doc" ] || exit 3 ;;
D|R)
   [ -n "$doc" -a -d "$BASE/$doc" ] || exit 3
   [ -f "$BASE/$doc/doc.pdf" -o -f "$BASE/$doc/paper.$page.jpg" ] || exit 3 ;;
*) exit 1 ;;
esac

# RUN
function json_string() {
  printf '"%s"' "$(sed 's#[\\/"]#\\&#g;s#\t#\\t#g')"
}

cd "$BASE"
case "$mode" in
Q)
  found="$(
    case ${#dates[*]} in
    0) find . -mindepth 1 -maxdepth 1 -type d -printf '%P/|\n' ;;
    1) find ${dates[0]}* -maxdepth 0 -type d -printf '%p/|\n' ;;
    *) eval find {$(IFS=, eval echo '"${dates[*]}"')}\* -maxdepth 0 -type d -printf "'%p/|\\n'" ;;
    esac 2>/dev/null)"
  if [ ${#labels[*]} -gt 0 ]; then
    # This is an AND
    for l in "${labels[@]}"; do
      found="$(
        eval grep -lF${q_ci:+i}e "$l" ${found//|/labels} 2>/dev/null \
        | sed 's#labels$#|#')"
    done
    ## That would be an OR (for later, perhaps)
    #found="$(
    #  eval grep -lF${q_ci:+i}f <(printf '%s\n' "${labels[@]}") ${found//|/labels} 2>/dev/null \
    #  | sed 's#labels$#|#')"
  fi
  if [ ${#words[*]} -gt 0 ]; then
    # This is an AND
    tmpf="$(
      eval grep -Ho "'>[^<>]*</span'" ${found//|/paper.*.words} 2>/dev/null \
      | sed 's#</span$##;s#&lt;#<#g;s#&gt;#>#g;s#&amp;#\&#g')"
    for w in "${words[@]}"; do
      tmpf="$(grep -hF${q_ci:+i}e "$w" <<<"$tmpf")"
    done
    found="$(cut -d: -f1 <<<"$tmpf" | sed 's#paper\.[0-9]*\.words$#|#' | sort -u)"
    ## That would be an OR (for later, perhaps)
    #found="$(
    #  eval grep -Ho "'>[^<>]*</span'" ${found//|/paper.*.words} 2>/dev/null \
    #  | sed 's#</span$##;s#&lt;#<#g;s#&gt;#>#g;s#&amp;#\&#g' \
    #  | grep -hF${q_ci:+i}f <(printf '%s\n' "${words[@]}") \
    #  | cut -d: -f1 \
    #  | sed 's#paper\.[0-9]*\.words$#|#' \
    #  | sort -u)"
  fi
  printf '['
  while IFS=/ read folder nil; do if [ -n "$folder" ]; then
    [ -f $folder/doc.pdf ] && type=pdf || type=pages
    count=$(/bin/ls -1 $folder/paper.*.thumb.jpg 2>/dev/null | wc -l)
    labs="$(
      while IFS=, read lab nil; do printf ','; json_string <<<"$lab"; done < <(sort -df $folder/labels))"
    printf ',{"folder":%s,"labels":[%s],"count":%d,"type":"%s"}' \
      "$(json_string <<<"$folder")" "${labs:1}" $count "$type"
  fi; done < <(sort -r <<<"$found") | sed 's/^.//'
  printf ']'
  ;;
T)
  printf '['
  /bin/ls -1 $doc/paper.*.thumb.jpg | sort -t. -k2,2n \
  | while read t; do
    read w h < <(file -b "$t" | sed -r 's/.*, ([0-9]+)x([0-9]+),[^,]*$/\1 \2/')
    printf ',{"mime":"image\/jpeg","data":%s,"width":%d,"height":%d}' \
      "$(base64 --wrap=0 "$t" | json_string)" $w $h
  done | sed 's/^.//'
  printf ']'
  ;;
D)
  if [ -f $doc/doc.pdf ]; then
    p=$doc/doc.pdf
    w=0; h=0
  else
    p=$doc/paper.$page.jpg
    read w h < <(file -b "$p" | sed -r 's/.*, ([0-9]+)x([0-9]+),[^,]*$/\1 \2/')
  fi
  mime=$(file -bi $p | cut -d';' -f1)
  printf '{"mime":%s,"data":%s,"width":%d,"height":%d}' \
    "$(json_string <<<"$mime")" "$(base64 --wrap=0 "$p" | json_string)" $w $h
  ;;
R)
  if [ -f $doc/doc.pdf ]; then
    cat $doc/doc.pdf
  else
    cat $doc/paper.$page.jpg
  fi
  ;;
esac