From 61f1ca09c279dcc2f7f94d9ffffb0681b3154dbc Mon Sep 17 00:00:00 2001
From: Ines Montani <ines@ines.io>
Date: Thu, 12 Jan 2017 21:58:55 +0100
Subject: [PATCH] Modernise serializer codecs tests

---
 spacy/tests/serialize/test_codecs.py | 65 ++++++++++++----------------
 1 file changed, 28 insertions(+), 37 deletions(-)

diff --git a/spacy/tests/serialize/test_codecs.py b/spacy/tests/serialize/test_codecs.py
index 00177f21a..4e080f600 100644
--- a/spacy/tests/serialize/test_codecs.py
+++ b/spacy/tests/serialize/test_codecs.py
@@ -1,60 +1,51 @@
+# coding: utf-8
 from __future__ import unicode_literals
-import pytest
+
+from ...serialize.packer import _BinaryCodec
+from ...serialize.huffman import HuffmanCodec
+from ...serialize.bits import BitArray
 
 import numpy
-
-from spacy.vocab import Vocab
-from spacy.serialize.packer import _BinaryCodec
-from spacy.serialize.huffman import HuffmanCodec
-from spacy.serialize.bits import BitArray
+import pytest
 
 
-def test_binary():
+def test_serialize_codecs_binary():
     codec = _BinaryCodec()
     bits = BitArray()
-    msg = numpy.array([0, 1, 0, 1, 1], numpy.int32)
-    codec.encode(msg, bits)
+    array = numpy.array([0, 1, 0, 1, 1], numpy.int32)
+    codec.encode(array, bits)
     result = numpy.array([0, 0, 0, 0, 0], numpy.int32)
     bits.seek(0)
     codec.decode(bits, result)
-    assert list(msg) == list(result)
+    assert list(array) == list(result)
 
 
-def test_attribute():
-    freqs = {'the': 10, 'quick': 3, 'brown': 4, 'fox': 1, 'jumped': 5, 'over': 8,
-            'lazy': 1, 'dog': 2, '.': 9}
- 
-    int_map = {'the': 0, 'quick': 1, 'brown': 2, 'fox': 3, 'jumped': 4, 'over': 5,
-               'lazy': 6, 'dog': 7, '.': 8}
+def test_serialize_codecs_attribute():
+    freqs = {'the': 10, 'quick': 3, 'brown': 4, 'fox': 1, 'jumped': 5,
+             'over': 8, 'lazy': 1, 'dog': 2, '.': 9}
+    int_map = {'the': 0, 'quick': 1, 'brown': 2, 'fox': 3, 'jumped': 4,
+               'over': 5, 'lazy': 6, 'dog': 7, '.': 8}
 
     codec = HuffmanCodec([(int_map[string], freq) for string, freq in freqs.items()])
-
     bits = BitArray()
-    
-    msg = numpy.array([1, 7], dtype=numpy.int32)
-    msg_list = list(msg)
-    codec.encode(msg, bits)
+    array = numpy.array([1, 7], dtype=numpy.int32)
+    codec.encode(array, bits)
     result = numpy.array([0, 0], dtype=numpy.int32)
     bits.seek(0)
     codec.decode(bits, result)
-    assert msg_list == list(result)
+    assert list(array) == list(result)
 
 
-def test_vocab_codec():
-    vocab = Vocab()
-    lex = vocab['dog']
-    lex = vocab['the']
-    lex = vocab['jumped']
-
-    codec = HuffmanCodec([(lex.orth, lex.prob) for lex in vocab])
-
+def test_serialize_codecs_vocab(en_vocab):
+    words = ["the", "dog", "jumped"]
+    for word in words:
+        _ = en_vocab[word]
+    codec = HuffmanCodec([(lex.orth, lex.prob) for lex in en_vocab])
     bits = BitArray()
-    
-    ids = [vocab[s].orth for s in ('the', 'dog', 'jumped')]
-    msg = numpy.array(ids, dtype=numpy.int32)
-    msg_list = list(msg)
-    codec.encode(msg, bits)
-    result = numpy.array(range(len(msg)), dtype=numpy.int32)
+    ids = [en_vocab[s].orth for s in words]
+    array = numpy.array(ids, dtype=numpy.int32)
+    codec.encode(array, bits)
+    result = numpy.array(range(len(array)), dtype=numpy.int32)
     bits.seek(0)
     codec.decode(bits, result)
-    assert msg_list == list(result)
+    assert list(array) == list(result)