openmedialibrary/oml/media/pdf.py

150 lines
4.2 KiB
Python
Raw Normal View History

2014-05-04 17:26:43 +00:00
# -*- coding: utf-8 -*-
# vi:si:et:sw=4:sts=4:ts=4
from __future__ import division
import sys
import tempfile
import subprocess
import os
import shutil
from glob import glob
from pyPdf import PdfFileReader
import stdnum.isbn
import settings
from utils import normalize_isbn, find_isbns
2014-05-18 03:01:24 +00:00
import logging
logger = logging.getLogger('oml.meta.pdf')
2014-05-04 17:26:43 +00:00
def cover(pdf):
if sys.platform == 'darwin':
return ql_cover(pdf)
else:
return page(pdf, 1)
def ql_cover(pdf):
tmp = tempfile.mkdtemp()
cmd = [
'qlmanage',
'-t',
'-s',
'1024',
'-o',
tmp,
pdf
]
p = subprocess.Popen(cmd)
p.wait()
2014-05-20 00:08:28 +00:00
image = glob('%s/*' % tmp)
if image:
image = image[0]
with open(image, 'rb') as fd:
data = fd.read()
else:
logger.debug('qlmanage did not create cover for %s', pdf)
data = None
2014-05-04 17:26:43 +00:00
shutil.rmtree(tmp)
return data
def page(pdf, page):
image = tempfile.mkstemp('.jpg')[1]
cmd = [
'gs', '-q',
'-dBATCH', '-dSAFER', '-dNOPAUSE', '-dNOPROMPT',
'-dMaxBitmap=500000000',
'-dAlignToPixels=0', '-dGridFitTT=2',
'-sDEVICE=jpeg', '-dTextAlphaBits=4', '-dGraphicsAlphaBits=4',
'-r72',
'-dUseCropBox',
'-dFirstPage=%d' % page,
'-dLastPage=%d' % page,
'-sOutputFile=%s' % image,
pdf
]
p = subprocess.Popen(cmd)
p.wait()
with open(image, 'rb') as fd:
data = fd.read()
os.unlink(image)
return data
def info(pdf):
data = {}
with open(pdf, 'rb') as fd:
try:
pdfreader = PdfFileReader(fd)
2014-05-18 23:24:04 +00:00
data['pages'] = pdfreader.numPages
2014-05-04 17:26:43 +00:00
info = pdfreader.getDocumentInfo()
if info:
for key in info:
if info[key]:
data[key[1:].lower()] = info[key]
xmp =pdfreader.getXmpMetadata()
if xmp:
for key in dir(xmp):
if key.startswith('dc_'):
value = getattr(xmp, key)
if isinstance(value, dict) and 'x-default' in value:
value = value['x-default']
elif isinstance(value, list):
value = [v.strip() for v in value if v.strip()]
_key = key[3:]
if value and _key not in data:
data[_key] = value
except:
2014-05-18 03:01:24 +00:00
logger.debug('FAILED TO PARSE %s', pdf, exc_info=1)
2014-05-04 17:26:43 +00:00
if 'identifier' in data:
value = normalize_isbn(data['identifier'])
if stdnum.isbn.is_valid(value):
2014-05-21 00:02:21 +00:00
data['isbn'] = [value]
2014-05-04 17:26:43 +00:00
del data['identifier']
'''
cmd = ['pdfinfo', pdf]
p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
stdout, stderr = p.communicate()
for line in stdout.strip().split('\n'):
parts = line.split(':')
key = parts[0].lower().strip()
if key:
data[key] = ':'.join(parts[1:]).strip()
for key in data.keys():
if not data[key]:
del data[key]
'''
text = extract_text(pdf)
data['textsize'] = len(text)
if settings.server['extract_text']:
if not 'isbn' in data:
isbn = extract_isbn(text)
if isbn:
2014-05-21 00:02:21 +00:00
data['isbn'] = [isbn]
2014-05-04 17:26:43 +00:00
return data
'''
#possbile alternative with gs
tmp = tempfile.mkstemp('.txt')[1]
cmd = ['gs', '-dBATCH', '-dNOPAUSE', '-sDEVICE=txtwrite', '-dFirstPage=3', '-dLastPage=5', '-sOutputFile=%s'%tmp, pdf]
'''
def extract_text(pdf):
if sys.platform == 'darwin':
2014-05-16 17:08:10 +00:00
cmd = ['/usr/bin/mdimport', '-d2', pdf]
2014-05-04 17:26:43 +00:00
else:
cmd = ['pdftotext', pdf, '-']
p = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
stdout, stderr = p.communicate()
if sys.platform == 'darwin':
2014-05-20 00:08:28 +00:00
if 'kMDItemTextContent' in stderr:
stdout = stderr.split('kMDItemTextContent = "')[-1].split('\n')[0][:-2]
else:
stdout = ''
2014-05-04 17:26:43 +00:00
return stdout.strip()
def extract_isbn(text):
isbns = find_isbns(text)
if isbns:
return isbns[0]