PDFMiner sample.

pyparse.py

import sys
from pdfminer.pdfinterp import PDFResourceManager, process_pdf
from pdfminer.pdfdevice import PDFDevice

class ObjectFindDevice(PDFDevice):

    debug = 0

    def __init__(self, rsrcmgr):
        self.rsrcmgr = rsrcmgr
        self.ctm = None
        self.fonts = set()
        self.nof_pages = 0
        self.pages = set()
        return

    def begin_page(self, page, ctm):
        self.nof_pages += 1
        self.pages.add(page)

    def end_page(self, page):
        return
    def begin_figure(self, name, bbox, matrix):
        return
    def end_figure(self, name):
        return

    def paint_path(self, graphicstate, stroke, fill, evenodd, path):
        return
    def render_image(self, name, stream):
        return
    def render_string(self, textstate, seq):
        self.fonts.add(textstate.font)

def convert_point_to_mm(box):
    '''
    72point = 1inch = 25.4mm
    >>> convert_point_to_mm([0, 0, 283.4646, 419.5276])
    [0.0, 0.0, 100.00001166666667, 148.00001444444442]
    '''
    return [(x / 72.0 * 25.4) for x in box]


if __name__ == '__main__':
    rsrcmgr = PDFResourceManager(caching=True)
    device = ObjectFindDevice(rsrcmgr)
    with open('test_data/omotemen.pdf') as fp:
        process_pdf(rsrcmgr, device, fp)
    print('fonts used: {}'.format(len(device.fonts)))
    print('no of pages: {}'.format(device.nof_pages))
    for page in device.pages:
        print(page.mediabox)
        print(convert_point_to_mm(page.mediabox))