тест на загрузку sql датасета из конфига + разделение тестовых фиксту…

…р на in memory базу датасета и базу датасета в файле

para_tri_dataset/paraphrase_dataset/para_phraser_plus/test/test_sql_dataset.py

@@ -4,7 +4,9 @@
 from sqlalchemy.orm import sessionmaker
 
 from para_tri_dataset.alchemy_utils import Database
+from para_tri_dataset.config import Config
 
+from para_tri_dataset.paraphrase_dataset import get_dataset_from_config
 from para_tri_dataset.paraphrase_dataset.para_phraser_plus import ParaPhraserPlusSQLDataset
 from para_tri_dataset.paraphrase_dataset.para_phraser_plus.base import ParaPhraserPlusPhrase
 
@@ -29,67 +31,97 @@ def paraphrases(phrases):
     ]
 
 
-@pytest.fixture
-def database(paraphrases) -> Database:
-    engine = create_engine("sqlite:///")
-    Session = sessionmaker(engine, expire_on_commit=False)
-    db = Database(engine, Session)
+def create_database(db: Database, dataset_paraphrases):
     db.create_all(Base)
 
     with db.session_scope() as session:
-        for group_id, phrases in enumerate(paraphrases):
+        for group_id, phrases in enumerate(dataset_paraphrases):
             session.add_all([ParaphraserPlusDataset(id=p.id, text=p.text, group_id=group_id) for p in phrases])
 
         session.commit()
 
+
+@pytest.fixture
+def dataset_config(tmp_path, paraphrases):
+    dataset_filepath = tmp_path / 'test-paraphraser-plus.sqlite'
+    db_url = f"sqlite:///{dataset_filepath.absolute()}"
+
+    engine = create_engine(db_url)
+    Session = sessionmaker(engine, expire_on_commit=False)
+
+    db = Database(engine, Session)
+    create_database(db, paraphrases)
+
+    cfg = Config(name="paraphrase_plus_sql", data={'db_url': db_url}, nested_configs=[])
+    yield cfg
+
+    dataset_filepath.unlink()
+
+
+@pytest.fixture
+def in_memory_database(paraphrases) -> Database:
+    engine = create_engine("sqlite:///")
+    Session = sessionmaker(engine, expire_on_commit=False)
+    db = Database(engine, Session)
+
+    create_database(db, paraphrases)
     return db
 
 
-def test_get_phrase_by_id(database, phrases):
-    dataset = ParaPhraserPlusSQLDataset(database)
+@pytest.fixture
+def in_memory_dataset(in_memory_database) -> ParaPhraserPlusSQLDataset:
+    return ParaPhraserPlusSQLDataset(in_memory_database)
+
+
+def test_load_from_config(dataset_config):
+    _ = ParaPhraserPlusSQLDataset.from_config(dataset_config)
+    _ = get_dataset_from_config(dataset_config)
+
+    with pytest.raises(ValueError):
+        dataset_config.name = "foo"
+        _ = get_dataset_from_config(dataset_config)
+
+
+def test_get_phrase_by_id(in_memory_dataset, phrases):
 
     for phrase in phrases:
-        dataset_phrase = dataset.get_phrase_by_id(phrase.id)
+        dataset_phrase = in_memory_dataset.get_phrase_by_id(phrase.id)
         assert phrase == dataset_phrase
 
 
-def test_iterate_phrases(database, phrases):
-    dataset = ParaPhraserPlusSQLDataset(database)
+def test_iterate_phrases(in_memory_dataset, phrases):
 
-    for orig_phrase, dataset_phrase in zip(phrases, dataset.iterate_phrases()):
+    for orig_phrase, dataset_phrase in zip(phrases, in_memory_dataset.iterate_phrases()):
         assert orig_phrase == dataset_phrase
 
-    for orig_phrase, dataset_phrase in zip(phrases[1:], dataset.iterate_phrases(offset=1)):
+    for orig_phrase, dataset_phrase in zip(phrases[1:], in_memory_dataset.iterate_phrases(offset=1)):
         assert orig_phrase == dataset_phrase
 
 
-def test_iterate_phrases_ids(database, phrases):
-    dataset = ParaPhraserPlusSQLDataset(database)
+def test_iterate_phrases_ids(in_memory_dataset, phrases):
 
-    for orig_phrase, phrase_id in zip(phrases, dataset.iterate_phrases_id()):
+    for orig_phrase, phrase_id in zip(phrases, in_memory_dataset.iterate_phrases_id()):
         assert orig_phrase.id == phrase_id
 
-    for orig_phrase, phrase_id in zip(phrases[1:], dataset.iterate_phrases_id(offset=1)):
+    for orig_phrase, phrase_id in zip(phrases[1:], in_memory_dataset.iterate_phrases_id(offset=1)):
         assert orig_phrase.id == phrase_id
 
 
-def test_get_paraphrases(database, paraphrases):
-    dataset = ParaPhraserPlusSQLDataset(database)
+def test_get_paraphrases(in_memory_dataset, paraphrases):
 
     for phrases in paraphrases:
         for i in range(len(phrases)):
             phrase, phrase_paraphrases = phrases[i], [p for j, p in enumerate(phrases) if j != i]
-            dataset_paraphrases = dataset.get_paraphrases(phrase)
+            dataset_paraphrases = in_memory_dataset.get_paraphrases(phrase)
 
             assert phrase_paraphrases == dataset_paraphrases
 
 
-def test_get_paraphrases_id(database, paraphrases):
-    dataset = ParaPhraserPlusSQLDataset(database)
+def test_get_paraphrases_id(in_memory_dataset, paraphrases):
 
     for phrases in paraphrases:
         for i in range(len(phrases)):
             phrase_id, phrase_paraphrases_id = phrases[i].id, [p.id for j, p in enumerate(phrases) if j != i]
-            dataset_paraphrases_id = dataset.get_paraphrases_id(phrase_id)
+            dataset_paraphrases_id = in_memory_dataset.get_paraphrases_id(phrase_id)
 
             assert phrase_paraphrases_id == dataset_paraphrases_id