Pdf book from the image files

makePdfBook.py

# -*- mode: python; coding: utf-8 -*-
#
#
import os
from optparse import OptionParser
from matplotlib.backends.backend_pdf import PdfPages
from scipy.misc import imread
import matplotlib.pyplot as plt
import numpy as np

##
#
#
def plotImage(file, grayscale=True):
    if grayscale :
        from skimage import io
        im = io.imread(file, as_grey=True)
        plt.imshow(im, cmap=plt.get_cmap("gray"))
    else:
        im = imread(file)
        plt.imshow(im)

    # no-grid, no-scales and no bbox
    a = plt.gca()
    a.get_xaxis().set_visible(False)
    a.get_yaxis().set_visible(False)
    for spines in a.spines.values():
       spines.set_visible(False)


##
#
#
def fileSignature(filePath):
    magic = {
        'SWF': [ '43', '57', '53' ],
        'ID3': [ '49', '44', '33' ],
        'MP3': [ 'FF', 'FB' ],
        'JPG': [ 'FF', 'D8' ],
        'PNG': [ '89', '50', '4E', '47' ],
        'BMP': [ '42', '4D' ],
        'GIF': [ '47', '49', '46', '38' ],
        'GZIP': [ '1F', '8B' ]
        }
    def buffer2hexlist(buffer, start, count):
        result = []
        for item in range(count):
            result.append('%02X' % ord(buffer[start + item]))
        return result
    #
    #
    with open(filePath, "rb") as fd:
        data = fd.read(8)
        fd.close()
        blist = buffer2hexlist(data, 0, 8)

    for ftype, signature in magic.iteritems():
        flag = True
        for id, val in zip(signature, blist):
            if id != val:
                flag = False
                break
        if flag:
            return ftype
    return 'UNKNOWN'

##
# glab files
#
def filePaths( directory ):
    file_paths = [];

    for root, directories, files in os.walk(directory):
        for filename in files:
            filepath = os.path.join(root, filename)
            sig = fileSignature(filepath)
            if sig in ['JPG', 'PNG', 'GIF']:
                file_paths.append(filepath)
    return file_paths

#
#
#
if __name__ == '__main__':
    parser = OptionParser()
    parser.add_option("-o", "--output", dest="filename", default="pages.pdf",
                       help="write report to FILE", metavar="FILE")
    parser.add_option("-g", "--grayscale",
                      action="store_false", dest="gray", default=True,
                      help="don't print status messages to stdout")
    parser.add_option("-q", "--quiet",
                      action="store_false", dest="verbose", default=True,
                      help="don't print status messages to stdout")

    (options, args) = parser.parse_args()
    files = filePaths( args[0] )
    pp = PdfPages( options.filename )
    for idx in range(len(files)):
        if options.verbose:
            print idx, files[idx]
        plt.subplots()
        plotImage(files[idx], options.gray)
        f = plt.gcf()
        # A4-portlate
        f.set_size_inches(8.27, 11.69)
        pp.savefig(f, bbox_inches='tight')
        plt.close()
    pp.close()

add the grayscale mode.