Initial commit and release of doi2dataset

tests/test_doi2dataset.py

@@ -0,0 +1,50 @@
+import os
+import sys
+
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
+
+from doi2dataset import NameProcessor, Phase, sanitize_filename, validate_email_address
+
+
+def test_phase_check_year():
+    """Test that check_year correctly determines if a year is within the phase boundaries."""
+    phase = Phase("TestPhase", 2000, 2010)
+    # Within boundaries
+    assert phase.check_year(2005) is True
+    # Outside boundaries
+    assert phase.check_year(1999) is False
+    assert phase.check_year(2011) is False
+    # Boundary cases
+    assert phase.check_year(2000) is True
+    assert phase.check_year(2010) is True
+
+def test_sanitize_filename():
+    """Test the sanitize_filename function to convert DOI to a valid filename."""
+    doi = "10.1234/abc.def"
+    expected = "10_1234_abc_def"
+    result = sanitize_filename(doi)
+    assert result == expected
+
+def test_split_name_with_comma():
+    """Test splitting a full name that contains a comma."""
+    full_name = "Doe, John"
+    given, family = NameProcessor.split_name(full_name)
+    assert given == "John"
+    assert family == "Doe"
+
+def test_split_name_without_comma():
+    """Test splitting a full name that does not contain a comma."""
+    full_name = "John Doe"
+    given, family = NameProcessor.split_name(full_name)
+    assert given == "John"
+    assert family == "Doe"
+
+def test_validate_email_address_valid():
+    """Test that a valid email address is correctly recognized."""
+    valid_email = "john.doe@iana.org"
+    assert validate_email_address(valid_email) is True
+
+def test_validate_email_address_invalid():
+    """Test that an invalid email address is correctly rejected."""
+    invalid_email = "john.doe@invalid_domain"
+    assert validate_email_address(invalid_email) is False

tests/test_fetch_doi_mock.py

@@ -0,0 +1,57 @@
+import json
+import os
+import sys
+
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
+
+import pytest
+
+from doi2dataset import MetadataProcessor
+
+
+class FakeResponse:
+    """
+    A fake response object to simulate an API response.
+    """
+    def __init__(self, json_data, status_code=200):
+        self._json = json_data
+        self.status_code = status_code
+
+    def json(self):
+        return self._json
+
+    def raise_for_status(self):
+        pass
+
+@pytest.fixture
+def fake_openalex_response():
+    """
+    Load the saved JSON response from the file 'srep45389.json'
+    located in the same directory as this test file.
+    """
+    json_path = os.path.join(os.path.dirname(__file__), "srep45389.json")
+    with open(json_path, "r", encoding="utf-8") as f:
+        data = json.load(f)
+    return data
+
+def test_fetch_doi_data_with_file(mocker, fake_openalex_response):
+    """
+    Test fetching DOI metadata by simulating the API call with a locally saved JSON response.
+
+    The APIClient.make_request method is patched to return a fake response based on
+    the contents of 'srep45389.json', so that no actual network request is performed.
+    """
+    doi = "10.1038/srep45389"
+    fake_response = FakeResponse(fake_openalex_response, 200)
+
+    # Patch the make_request method of APIClient to return our fake_response.
+    mocker.patch("doi2dataset.APIClient.make_request", return_value=fake_response)
+
+    # Instantiate MetadataProcessor without upload and progress.
+    processor = MetadataProcessor(doi=doi, upload=False)
+
+    # Call _fetch_data(), which should now return our fake JSON data.
+    data = processor._fetch_data()
+
+    # Verify that the fetched data matches the fake JSON data.
+    assert data == fake_openalex_response