#!/usr/bin/python
# -*- coding: utf-8 -*-
#
# bwscan2pdf V 0.4
# Copyright (c) by Thomas Lauckner, 2009
#
# This work is licensed cc-by-sa
# See http://creativecommons.org/licenses/by-sa/3.0/
#
# Changelog:
#
# V 0.4 The script uses temporary files now.
# V 0.3 Changed gimp script from threshold to photocopy for better output
# V 0.2 tesseract option included
# V 0.1 Initial release
#
# TODO:
# * Progress bar "#" -> "#######"
# * Confirm the input files are images (in the right format???)
# * Be flexible: Use any image format "convert" accepts
# * Optimize for speed

import sys
import os
import getopt
import tempfile

# We have to use a temporary directory
# since the output of tesseract cannot be controlled.
# => Put all output in this temporary directory and everything is fine :)
TEMPDIR=tempfile.mkdtemp(prefix="bwscan")

def usage():
	print "Usage: %s -s -d -o -e -n -r -t files"  % sys.argv[0]
	print "Usage: %s --single --double --odd --even --normal --reverse --tesseract files" % sys.argv[0]

def defaults():
  single="TRUE"
  tesseract="TRUE"
  return (single,tesseract)


def not_implemented():
	print "This option is not implemented yet."

def merge(file1,file2):
  file(file2,'a').write("\n"+file(file1,'r').read())

try:
	opts, args = getopt.getopt(sys.argv[1:], "sdoernt", ["single", "double", "odd", "even", "reverse", "normal", "tesseract"])
except getopt.GetoptError:
	usage()
	sys.exit(2)

def tempFilename(suffixstr=""):
  (openedfd,name)=tempfile.mkstemp(suffix=suffixstr,prefix="bwscan",dir=TEMPDIR)
  os.close(openedfd)
  return name

single="NONE"
even="NONE"
reverse="NONE"
tesseract="NONE"
for opt,arg in opts:
	if opt in ("-s", "--single"):
		if single=="NONE":
			single="TRUE"
		else:
			usage()
			sys.exit(2)
	elif opt in ("-d", "--double"):
		if single=="NONE":
			single="FALSE"
		else:
			usage()
			sys.exit(2)
	elif opt in ("-o", "--odd"):
		if even=="NONE":
			even="FALSE"
		else:
			usage()
			sys.exit(2)
	elif opt in ("-e", "--even"):
		if even=="NONE":
			even="TRUE"
		else:
			usage()
			sys.exit(2)
	elif opt in ("-r", "--reverse"):
		if reverse=="NONE":
			reverse="TRUE"
		else:
			usage()
			sys.exit(2)
	elif opt in ("-n", "--normal"):
		if reverse=="NONE":
			reverse="FALSE"
		else:
			usage()
			sys.exit(2)
	if opt in ("-t", "--tesseract"):
	  tesseract="TRUE"

# Fall back to default
if single=="NONE":
  print "Option 'single/double' is not set. Falling back to default."
  (single,tesseract)=defaults()

if single=="TRUE":
	reverse="FALSE"
if reverse=="NONE":
	usage()
	sys.exit(2)
elif reverse=="TRUE" and even=="NONE":
	usage()
	sys.exit(2)

filenames_in=args
filenames_sorted=[]
filenames_temp_bwpng={}
filenames_temp_pdf={}
filenames_temp_tif={}
temp_final_textfile=tempFilename(".txt")
temp_final_pdf=tempFilename(".pdf")

if single=="TRUE":
	filenames_sorted=filenames_in
elif reverse=="FALSE":
	# Sort files, assuming that the odd pages were scanned first
	odd_pages,rem=divmod(len(filenames_in),2)
	odd_pages+=rem
	for oddindex in range(odd_pages):
	  filenames_sorted.append(filenames_in[oddindex])
	  if (oddindex+odd_pages < len(filenames_in)):
	    filenames_sorted.append(filenames_in[oddindex+odd_pages])
elif reverse=="TRUE":
	odd_pages,rem=divmod(len(filenames_in),2)
	if rem==1:
		print "The number of files needs to be be even."
		sys.exit(2)
	for oddindex in range(odd_pages):
		filenames_sorted.append(filenames_in[oddindex])
		if even=="FALSE" and oddindex==odd_pages-1:
			continue
		else:
			filenames_sorted.append(filenames_in[len(filenames_in)-oddindex-1])

# Convert images to b/w pdf using gimp and convert
# via gimp-threshold
#gimpscript='''
#gimp -i -d -f -b '(define (threshold infile outfile) (let* ((image (car (gimp-file-load RUN-NONINTERACTIVE infile infile))) (drawable (car (gimp-image-get-active-layer image)))) (gimp-threshold drawable 200 255) (set! drawable (car (gimp-image-get-active-layer image))) (gimp-file-save RUN-NONINTERACTIVE image drawable outfile outfile) (gimp-image-delete image))) (threshold "%s" "%s") (gimp-quit 0)' '''

# via plug-in-photocopy
gimpscript='''
gimp -i -d -f -b '(define (photocopy infile outfile) (let* ((image (car (gimp-file-load RUN-NONINTERACTIVE infile infile))) (drawable (car (gimp-image-get-active-layer image)))) (plug-in-photocopy RUN-NONINTERACTIVE image drawable 50 0.8 .2 .2) (set! drawable (car (gimp-image-get-active-layer image))) (gimp-file-save RUN-NONINTERACTIVE image drawable outfile outfile) (gimp-image-delete image))) (photocopy "%s" "%s") (gimp-quit 0)' '''
os.system("gimp -i -d -f -b '(gimp-quit 0)'")
for filename in filenames_sorted:
  filenames_temp_bwpng[filename]=tempFilename(".png")
  filenames_temp_pdf[filename]=tempFilename(".pdf")
  filenames_temp_tif[filename]=tempFilename(".tif")

#  print "Converting %s to b/w..." % filename
  os.system(gimpscript % (filename,filenames_temp_bwpng[filename]))
#  print "Converting %s to pdf..." % filename
  os.system("convert %s %s" % (filenames_temp_bwpng[filename],filenames_temp_pdf[filename]))
  if tesseract=="TRUE":
#    print "Converting %s for tesseract use..." % filename
    os.system("convert %s %s" % (filename,filenames_temp_tif[filename]))
#    print "Using tesseract OCR on %s..." % filename
    os.system("tesseract %s %s -l deu" % (filenames_temp_tif[filename],filenames_temp_tif[filename]))
print "" # This is to end the "#"-Line (see at the beginning of the big loop)

# Produce a single pdf file using pdftk
filenames_sorted_string=""
for filename in filenames_sorted:
  filenames_sorted_string=filenames_sorted_string+filenames_temp_pdf[filename]+" "

if os.path.exists("bwscan.pdf"):
  stop=1
  answer=raw_input("bwscan.pdf exists. Remove? [y/n] ")
  if (answer=="y"):
    os.remove("bwscan.pdf")
    stop=0
else: stop=0

if stop==0:
  if tesseract=="TRUE":
#    print "Producing bwscan.txt..."
    for filename in filenames_in:
      merge(filenames_temp_tif[filename]+".txt",temp_final_textfile)
#  print "Producing bwscan.pdf..."
  os.system("pdftk %s cat output %s" % (filenames_sorted_string,temp_final_pdf))
  if tesseract=="TRUE":
#    print "Attaching bwscan.txt to bwscan.pdf..."
    os.system("pdftk %s attach_files %s to_page 1 output bwscan.pdf" % (temp_final_pdf,temp_final_textfile))
  else:
    # This has a reason: os.rename() does not work, if the files are on different partitions
    # (e.g. /tmp and /home)
    os.system("pdftk %s cat output bwscan.pdf")
  answer=raw_input("Should I remove the original files? [y/n] ")
  if (answer=="y"):
    for filename in filenames_in:
      print "Removing %s" % filename
      os.remove(filename)
  
#print "Removing garbage..."
for filename,tempfile in filenames_temp_bwpng.items():
  os.remove(tempfile)
for filename,tempfile in filenames_temp_pdf.items():
  os.remove(tempfile)
for filename,tempfile in filenames_temp_tif.items():
  os.remove(tempfile)
  if tesseract=="TRUE":
    os.remove(tempfile+".txt")
os.remove(temp_final_textfile)
os.remove(temp_final_pdf)
os.rmdir(TEMPDIR)