def merge_pdf(inputs):
inputs = [to_stream(inp) for inp in inputs]
assert all(inputs), 'No inputs? %r' % inputs
if not inputs:
return None
with mktempfn('-M.pdf', 'm-') as out_fn:
try:
simple_merge(out_fn, inputs)
except Exception, e:
LOG.error('simple merge cannot merge %s: %s', inputs, e)
shaved = []
for inp in inputs:
with mktempfn('-shaved.pdf', delete_asap=False) as shvd:
shaved.append(pdftk_merge(shvd, [inp]))
try:
simple_merge(out_fn, map(to_stream, shaved))
except Exception, e:
LOG.error('simple merge cannot merge shaved %s: %s',
shaved, e)
pdftk_merge(out_fn, shaved)
map(unlink, shaved)
def to_filename(stream, opened_ok=True):
if isinstance(stream, basestring):
if file_exists(stream):
return stream
with mktempfn('-TF.pdf', 'tof-', delete_asap=False) as fn:
open(fn, 'wb').write(stream)
return fn
if opened_ok and hasattr(stream, 'name') and file_exists(stream.name):
return stream.name
if hasattr(stream, 'seek'):
stream.seek(0, 0)
else:
LOG.warn('no seek on %r!', stream)
with mktempfn('-TF.pdf', 'tof-', delete_asap=False) as fn:
with open(fn, 'wb') as fh:
while True:
chunk = stream.read(65536)
if not chunk:
break
fh.write(chunk)
return fn
def to_stream(stream):
if not hasattr(stream, 'seek'):
if file_exists(stream):
stream = MMFile(stream, 'rb', delete_on_close=False)
else:
with mktempfn('-TS.pdf', 'tos-') as fn:
open(fn, 'wb').write(stream)
stream = MMFile(fn, 'rb', delete_on_close=True)
stream.seek(0, 2)
if stream.tell() == 0:
LOG.warn('zero pdf !? (%r)', repr(stream)[:100])
stream.seek(0, 0)
return stream
def simple_split(stream):
stream_s = repr(stream)[:100]
pages = list(get_pages(stream))
for page in pages:
out = PdfFileWriter()
out.addPage(page)
writer = timeouter(5, threaded=False)(out.write)
with mktempfn('-SS.pdf', 'ss-') as fn:
with open(fn, 'wb') as fh:
writer(fh)
LOG.debug('split_pdf(%s) yields %s', stream_s, fn)
yield MMFile(fn, 'rb', delete_on_close=True)
def get_pages(stream):
stream_s = repr(stream)[:100]
from pyPdf import PdfFileReader
# LOG.debug('stream: %s (%s)', stream_s, dir(stream))
if not hasattr(stream, 'seek'):
with mktempfn('-GP.pdf', 'gp-') as fn:
with open(fn, 'wb') as fh:
fh.write(stream)
stream = MMFile(fn, 'rb', delete_on_close=True)
stream.seek(0, 2)
if stream.tell() == 0:
LOG.warn('zero pdf !? (%r)', stream_s)
return
stream.seek(0, 0)
for i, page in enumerate(PdfFileReader(stream).pages):
LOG.debug('get_pages(%s) yields %d (%s)', stream_s, i,
repr(page)[:100])
yield page
LOG.info('%r is %d pages', stream_s, i)
def clean_pdf(pdffn):
LOG.info('cleaning %r', pdffn)
with mktempfn('-pp.ps') as psfn:
psfn = pdftops(psfn, [pdffn])
pdffn = pstopdf(pdffn, [psfn])
return pdffn
else:
LOG.warn('no seek on %r!', stream)
with mktempfn('-TF.pdf', 'tof-', delete_asap=False) as fn:
with open(fn, 'wb') as fh:
while True:
chunk = stream.read(65536)
if not chunk:
break
fh.write(chunk)
return fn
if '__main__' == __name__:
logging.basicConfig(level=logging.DEBUG)
todo, args = sys.argv[1], sys.argv[2:]
if 'merge' == todo:
sys.stdout.write(merge_pdf(args).getvalue())
elif 'split' == todo:
import zipfile
with mktempfn('.zip') as zfn:
zfh = zipfile.ZipFile(zfn, 'w', zipfile.ZIP_DEFLATED)
for fh in split_pdf(args[0]):
zfh.writestr(os.path.basename(fh.name), fh.getvalue())
zfh.close()
with open(zfn, 'rb') as zfh:
while True:
chunk = zfh.read(65536)
if not chunk:
break
sys.stdout.write(chunk)