def versions(self):
versions = cache.get('document_%s_versions' % self.id)
if versions is None:
versions = list(
self._blob.derived_documents.all().exclude(
file_type='png')
)
versions.append(self)
for version in versions:
version.original_document = self
version.priority = type_to_priorty(version.type)
versions = sorted(versions, key=lambda doc:doc.priority, reverse=True)
#cache.set('document_%s_versions' % self.id, versions, 60 * 20)
return versions
def create_pdf(doc):
log.info('PDF Conversion start')
blob = doc._blob
orig_file = blob.file
log.info('Starting conversion of %s to PDF' % blob)
# Check for derived files of PDF type
if blob.documents.filter(file_type='pdf'):
log.info('%s is a PDF, no need to convert' % blob)
return False
elif blob.derived_documents.filter(file_type='pdf'):
log.info('%s has derived PDF, no need to convert' % blob)
return False
tempd_loc = tempfile.mkdtemp()
temp_file, tempf_loc = tempfile.mkstemp(dir=tempd_loc)
temp_file = os.fdopen(temp_file, "wb")
for chunk in orig_file.chunks():
temp_file.write(chunk)
temp_file.close()
try:
port = get_free_port()
log.info("Launching unoconv with port %s" % port)
proc = subprocess.Popen(UNOCONV_CALL % (port, tempd_loc, tempf_loc),
#stderr=subprocess.PIPE,
#stdout=subprocess.PIPE,
shell=True)
# Create and start a watchdog thread
t = threading.Timer(TIMEOUT, timeout, [proc, port])
t.start()
stderr, stdout = None, None#proc.communicate()
proc.wait()
t.cancel()
if proc.returncode != 0:
error = subprocess.CalledProcessError(proc.returncode,
UNOCONV_CALL % (
port, tempd_loc, tempf_loc))
error.output = "%s %s" % (stderr, stdout)
raise error
log.info("unoconv (port %s) output: %s %s" % (port, stderr, stdout))
except subprocess.CalledProcessError as e:
log.error('unoconv (port %s) returned a non-zero exit status: %s' % (port, e.output))
os.unlink(tempf_loc)
files = os.listdir(tempd_loc)
for pdf in files:
if pdf.lower().endswith('.pdf'):
break
else:
shutil.rmtree(tempd_loc, True)
raise ConversionError('Unable to find PDF file')
pdf = os.path.abspath(os.path.join(tempd_loc, pdf))
pdf = open(pdf, 'rb')
filename = os.path.basename(orig_file.name)
filename = os.path.splitext(filename)[0] + '.pdf'
doc = DerivedDocument(derived_from=blob)
doc.file = UploadedFile(pdf, filename, 'application/pdf', 0, None)
doc.index = type_to_priorty('pdf')
# # Do one last check before saving the blob, just incase this task got fired
# # twice in quick succession.
if blob.derived_documents.filter(file_type='pdf'):
return False
else:
doc.save()
pdf.close()
shutil.rmtree(tempd_loc, True)
log.info("Convertion of %s complete" % blob)
return True
