Fix several errors in tests run by Travis (#380)

* Declare file encoding to support special characters

* fix missing utf_8_encoder error in Travis tests

* Py 2.7 backwards-compat iterator

* ensure integer argument to torch.nn.functional.pad

* cast match.ceil result as integer

torchaudio/datasets/utils.py

@@ -42,6 +42,12 @@ def unicode_csv_reader(unicode_csv_data, **kwargs):
     csv.field_size_limit(maxInt)
 
     if six.PY2:
+        # Implementation borrowed from docs:
+        # https://docs.python.org/3.0/library/csv.html#examples
+        def utf_8_encoder(unicode_csv_data):
+            for line in unicode_csv_data:
+                yield line.encode('utf-8')
+
         # csv.py doesn't do Unicode; encode temporarily as UTF-8:
         csv_reader = csv.reader(utf_8_encoder(unicode_csv_data), **kwargs)
         for row in csv_reader:
@@ -338,6 +344,10 @@ class _ThreadedIterator(threading.Thread):
             raise StopIteration
         return next_item
 
+    # Required for Python 2.7 compatibility
+    def next(self):
+        return self.__next__()
+
 
 def bg_iterator(iterable, maxsize):
     return _ThreadedIterator(iterable, maxsize=maxsize)

torchaudio/functional.py

+# -*- coding: utf-8 -*-
+
 from __future__ import absolute_import, division, print_function, unicode_literals
 
 import math
@@ -1093,12 +1095,12 @@ def _compute_nccf(waveform, sample_rate, frame_time, freq_low):
     EPSILON = 10 ** (-9)
 
     # Number of lags to check
-    lags = math.ceil(sample_rate / freq_low)
+    lags = int(math.ceil(sample_rate / freq_low))
 
     frame_size = int(math.ceil(sample_rate * frame_time))
 
     waveform_length = waveform.size()[-1]
-    num_of_frames = math.ceil(waveform_length / frame_size)
+    num_of_frames = int(math.ceil(waveform_length / frame_size))
 
     p = lags + num_of_frames * frame_size - waveform_length
     waveform = torch.nn.functional.pad(waveform, (0, p))
@@ -1147,7 +1149,7 @@ def _find_max_per_frame(nccf, sample_rate, freq_high):
     to the first half of lags, then the latter is taken.
     """
 
-    lag_min = math.ceil(sample_rate / freq_high)
+    lag_min = int(math.ceil(sample_rate / freq_high))
 
     # Find near enough max that is smallest
 

torchaudio/transforms.py

+# -*- coding: utf-8 -*-
+
 from __future__ import absolute_import, division, print_function, unicode_literals
 from warnings import warn
 import math