Modernize and merge tokenizer tests for punctuation

spacy/tests/tokenizer/test_only_punct.py

@@ -1,9 +0,0 @@
-from __future__ import unicode_literals
-
-
-def test_only_pre1(en_tokenizer):
-    assert len(en_tokenizer("(")) == 1
-
-
-def test_only_pre2(en_tokenizer):
-    assert len(en_tokenizer("((")) == 2

spacy/tests/tokenizer/test_post_punct.py

@@ -1,43 +0,0 @@
-from __future__ import unicode_literals
-import pytest
-
-
-@pytest.fixture
-def close_puncts():
-    return [')', ']', '}', '*']
-
-
-def test_close(close_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in close_puncts:
-        string = word_str + p
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 2
-        assert tokens[1].string == p
-        assert tokens[0].string == word_str
-
-
-def test_two_different_close(close_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in close_puncts:
-        string = word_str + p + "'"
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 3
-        assert tokens[0].string == word_str
-        assert tokens[1].string == p
-        assert tokens[2].string == "'"
-
-
-def test_three_same_close(close_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in close_puncts:
-        string = word_str + p + p + p
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 4
-        assert tokens[0].string == word_str
-        assert tokens[1].string == p
-
-
-def test_double_end_quote(en_tokenizer):
-    assert len(en_tokenizer("Hello''")) == 2
-    assert len(en_tokenizer("''")) == 1

spacy/tests/tokenizer/test_pre_punct.py

@@ -1,46 +0,0 @@
-from __future__ import unicode_literals
-
-import pytest
-
-
-@pytest.fixture
-def open_puncts():
-    return ['(', '[', '{', '*']
-
-
-def test_open(open_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in open_puncts:
-        string = p + word_str
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 2
-        assert tokens[0].orth_ == p
-        assert tokens[1].orth_ == word_str
-
-
-def test_two_different_open(open_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in open_puncts:
-        string = p + "`" + word_str
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 3
-        assert tokens[0].orth_ == p
-        assert tokens[1].orth_ == "`"
-        assert tokens[2].orth_ == word_str
-
-
-def test_three_same_open(open_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for p in open_puncts:
-        string = p + p + p + word_str
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 4
-        assert tokens[0].orth_ == p
-        assert tokens[3].orth_ == word_str
-
-
-def test_open_appostrophe(en_tokenizer):
-    string = "'The"
-    tokens = en_tokenizer(string)
-    assert len(tokens) == 2
-    assert tokens[0].orth_ == "'"

spacy/tests/tokenizer/test_punct.py

@@ -0,0 +1,109 @@
+from __future__ import unicode_literals
+
+import pytest
+
+PUNCT_OPEN = ['(', '[', '{', '*']
+PUNCT_CLOSE = [')', ']', '}', '*']
+PUNCT_PAIRED = [('(', ')'),  ('[', ']'), ('{', '}'), ('*', '*')]
+
+
+@pytest.mark.parametrize('text', ["(", "((", "<"])
+def test_tokenizer_only_punct(en_tokenizer, text):
+    tokens = en_tokenizer(text)
+    assert len(tokens) == len(text)
+
+
+@pytest.mark.parametrize('punct', PUNCT_OPEN)
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_open_punct(en_tokenizer, punct, text):
+    tokens = en_tokenizer(punct + text)
+    assert len(tokens) == 2
+    assert tokens[0].text == punct
+    assert tokens[1].text == text
+
+
+@pytest.mark.parametrize('punct', PUNCT_CLOSE)
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_close_punct(en_tokenizer, punct, text):
+    tokens = en_tokenizer(text + punct)
+    assert len(tokens) == 2
+    assert tokens[0].text == text
+    assert tokens[1].text == punct
+
+
+@pytest.mark.parametrize('punct', PUNCT_OPEN)
+@pytest.mark.parametrize('punct_add', ["`"])
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_two_diff_open_punct(en_tokenizer, punct, punct_add, text):
+    tokens = en_tokenizer(punct + punct_add + text)
+    assert len(tokens) == 3
+    assert tokens[0].text == punct
+    assert tokens[1].text == punct_add
+    assert tokens[2].text == text
+
+
+@pytest.mark.parametrize('punct', PUNCT_CLOSE)
+@pytest.mark.parametrize('punct_add', ["'"])
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_two_diff_close_punct(en_tokenizer, punct, punct_add, text):
+    tokens = en_tokenizer(text + punct + punct_add)
+    assert len(tokens) == 3
+    assert tokens[0].text == text
+    assert tokens[1].text == punct
+    assert tokens[2].text == punct_add
+
+
+@pytest.mark.parametrize('punct', PUNCT_OPEN)
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_same_open_punct(en_tokenizer, punct, text):
+    tokens = en_tokenizer(punct + punct + punct + text)
+    assert len(tokens) == 4
+    assert tokens[0].text == punct
+    assert tokens[3].text == text
+
+
+@pytest.mark.parametrize('punct', PUNCT_CLOSE)
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_same_close_punct(en_tokenizer, punct, text):
+    tokens = en_tokenizer(text + punct + punct + punct)
+    assert len(tokens) == 4
+    assert tokens[0].text == text
+    assert tokens[1].text == punct
+
+
+@pytest.mark.parametrize('text', ["'The"])
+def test_tokenizer_splits_open_appostrophe(en_tokenizer, text):
+    tokens = en_tokenizer(text)
+    assert len(tokens) == 2
+    assert tokens[0].text == "'"
+
+
+@pytest.mark.parametrize('text', ["Hello''"])
+def test_tokenizer_splits_double_end_quote(en_tokenizer, text):
+    tokens = en_tokenizer(text)
+    assert len(tokens) == 2
+    tokens_punct = en_tokenizer("''")
+    assert len(tokens_punct) == 1
+
+
+@pytest.mark.parametrize('punct_open,punct_close', PUNCT_PAIRED)
+@pytest.mark.parametrize('text', ["Hello"])
+def test_tokenizer_splits_open_close_punct(en_tokenizer, punct_open, punct_close, text):
+    tokens = en_tokenizer(punct_open + text + punct_close)
+    assert len(tokens) == 3
+    assert tokens[0].text == punct_open
+    assert tokens[1].text == text
+    assert tokens[2].text == punct_close
+
+
+@pytest.mark.parametrize('punct_open,punct_close', PUNCT_PAIRED)
+@pytest.mark.parametrize('punct_open_add,punct_close_add', [("`", "'")])
+@pytest.mark.parametrize('text', ["Hello"])
+def test_two_different(en_tokenizer, punct_open, punct_close, punct_open_add, punct_close_add, text):
+    tokens = en_tokenizer(punct_open_add + punct_open + text + punct_close + punct_close_add)
+    assert len(tokens) == 5
+    assert tokens[0].text == punct_open_add
+    assert tokens[1].text == punct_open
+    assert tokens[2].text == text
+    assert tokens[3].text == punct_close
+    assert tokens[4].text == punct_close_add

spacy/tests/tokenizer/test_surround_punct.py

@@ -1,32 +0,0 @@
-from __future__ import unicode_literals
-import pytest
-
-
-@pytest.fixture
-def paired_puncts():
-    return [('(', ')'),  ('[', ']'), ('{', '}'), ('*', '*')]
-
-
-def test_token(paired_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for open_, close_ in paired_puncts:
-        string = open_ + word_str + close_
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 3
-        assert tokens[0].orth_ == open_
-        assert tokens[1].orth_ == word_str
-        assert tokens[2].orth_ == close_
-
-
-def test_two_different(paired_puncts, en_tokenizer):
-    word_str = 'Hello'
-    for open_, close_ in paired_puncts:
-        string = "`" + open_ + word_str + close_ + "'"
-        tokens = en_tokenizer(string)
-        assert len(tokens) == 5
-        assert tokens[0].orth_ == "`"
-        assert tokens[1].orth_ == open_
-        assert tokens[2].orth_ == word_str
-        assert tokens[2].orth_ == word_str
-        assert tokens[3].orth_ == close_
-        assert tokens[4].orth_ == "'"