def safe_filename(fn):
fn = basename(fn)
if not isinstance(fn, unicode):
try:
fn = unicode(fn, sys.getfilesystemencoding())
except UnicodeDecodeError:
fn = unicode(fn, 'utf8', 'replace')
return fn.encode('cp437', 'replace')
def _request_files(files=None):
if files:
for elt in files:
yield elt
return
fn = None
if SAVE_REQ_DIR:
with _SRD_lock:
fns = os.listdir(SAVE_REQ_DIR)
if len(fns) > SAVE_REQ_DIR_THRESHOLD:
for x in sorted(fns)[:SAVE_REQ_DIR_THRESHOLD // 2]:
unlink(os.path.join(SAVE_REQ_DIR, x))
fn = os.path.join(SAVE_REQ_DIR,
str(int(time.time() * 1000)) + '--' +
('%06d' % REQ_COUNTER) +
urllib.quote_plus(request.url, safe=''))
import gzip
files = ((stor.stream, stor.mimetype, stor.name)
for _, stor in request.files.iteritems(multi=True))
try:
first = files.next()
files = itertools.chain([first], files)
except StopIteration: # nothing there
LOG.warn('empty "files"')
if request.data:
import re
rgx = re.compile('''filename=['"](.*?)["']''')
files = [(StringIO(request.data), request.headers['Content-Type'],
rgx.search(request.headers.get('Content-Disposition',
'filename=""')).group(1))]
LOG.info('files=%r', files)
i = 0
for stream, mimetype, name in files:
if len(stream.read(1)) == 1: # skip empty files
stream.seek(0)
if fn:
with gzip.open(fn + ('--file%02d' % i) + '.dmp.gz',
'wb') as fh:
while 1:
chunk = stream.read(65536)
if not chunk:
break
fh.write(chunk)
stream.seek(0)
i += 1
yield (stream, mimetype, basename(name))
def convert_to_pdf(files=None, splitted=None, return_errors=None,
dest_typ=None):
'''yields PDF streams'''
if splitted is None:
splitted = request.args.get('splitted', '0') == '1'
if return_errors is None:
return_errors = request.args.get('errors', '0') == '1'
errors = []
for stream, mimetype, name in request_files(files):
name = basename(name)
LOG.debug('to_pdf from=%s, stream=%r (%s)', mimetype, stream, name)
res = None
try:
res = pdf.to_pdf(stream, mimetype, splitted=splitted)
LOG.debug('to_pdf from=%s => %s', mimetype, repr(res)[:100])
except pdf.NonConvertible:
errors.append((stream, mimetype, name))
LOG.warn('%r not convertible to pdf', errors[-1])
except:
LOG.exception('error converting %s to pdf', repr(stream)[:100])
errors.append((stream, mimetype, name))
continue
if not res:
LOG.error('conversion of %r(%s) failed: %r', mimetype, stream, res)
continue
for i, chunk in enumerate(res):
if not chunk:
LOG.warn('chunk %r !?', chunk)
continue
if not pdf.assert_pdf(chunk,
extra='%s(%s): %s/%d' % (stream, mimetype, name, i),
raise_exc=False):
LOG.warn('part %d is NOT PDF!', i)
errors.append((stream, mimetype, name))
length = (len(chunk) if (pdf.MIN_PDF_SIZE > 0
and hasattr(chunk, '__len__'))
else -1)
LOG.debug('to_pdf yields %s (%d)', repr(chunk)[:100], length)
if length < 0 or length > pdf.MIN_PDF_SIZE:
yield chunk
else:
errors.append((stream, mimetype, name))
LOG.info('to_pdf done, errcount=%d', len(errors))
if return_errors and errors:
yield errors
