Skip to content

test_dataset#

test_create_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
21
22
23
24
25
26
27
28
@pytest.mark.usefixtures("data_catalog", "reference_data")
def test_create_dataset(admin_client, dataset_a_json, dataset_signal_handlers):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    assert_nested_subdict({"user": "admin", "organization": "admin"}, res.data["metadata_owner"])
    assert_nested_subdict(dataset_a_json, res.data, ignore=["generate_pid_on_publish"])
    assert res.data["metadata_repository"] == "Fairdata"  # Constant value
    dataset_signal_handlers.assert_call_counts(created=1, updated=0)

test_update_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
31
32
33
34
35
36
37
38
39
40
41
@pytest.mark.usefixtures("data_catalog", "reference_data")
def test_update_dataset(admin_client, dataset_a_json, dataset_b_json, dataset_signal_handlers):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    dataset_signal_handlers.assert_call_counts(created=1, updated=0)
    dataset_signal_handlers.reset()
    id = res.data["id"]
    dataset_b_json["persistent_identifier"] = res.data["persistent_identifier"]
    res = admin_client.put(f"/v3/datasets/{id}", dataset_b_json, content_type="application/json")
    assert res.status_code == 200
    assert_nested_subdict(dataset_b_json, res.data, ignore=["generate_pid_on_publish"])
    dataset_signal_handlers.assert_call_counts(created=0, updated=1)

test_update_dataset_with_project#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
44
45
46
47
48
49
50
51
def test_update_dataset_with_project(
    admin_client, dataset_a_json, dataset_d_json, data_catalog, reference_data
):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    id = res.data["id"]
    res = admin_client.patch(f"/v3/datasets/{id}", dataset_d_json, content_type="application/json")
    assert res.status_code == 200
    assert_nested_subdict(dataset_d_json, res.data)

test_filter_pid#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
def test_filter_pid(
    admin_client, dataset_a_json, dataset_b_json, datacatalog_harvested_json, reference_data
):
    dataset_a_json["generate_pid_on_publish"] = None
    dataset_b_json["generate_pid_on_publish"] = None
    dataset_a_json["persistent_identifier"] = "some_pid"
    dataset_b_json["persistent_identifier"] = "other_pid"
    dataset_a_json["data_catalog"] = datacatalog_harvested_json["id"]
    dataset_b_json["data_catalog"] = datacatalog_harvested_json["id"]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201, res.data
    res = admin_client.post("/v3/datasets", dataset_b_json, content_type="application/json")
    assert res.status_code == 201
    res = admin_client.get("/v3/datasets?persistent_identifier=some_pid")
    assert res.data["count"] == 1

test_search_pid#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
def test_search_pid(
    admin_client, dataset_a_json, dataset_b_json, datacatalog_harvested_json, reference_data
):
    dataset_a_json["generate_pid_on_publish"] = None
    dataset_b_json["generate_pid_on_publish"] = None
    dataset_a_json["persistent_identifier"] = "some_pid"
    dataset_b_json["persistent_identifier"] = "anotherpid"
    dataset_a_json["data_catalog"] = datacatalog_harvested_json["id"]
    dataset_b_json["data_catalog"] = datacatalog_harvested_json["id"]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    res = admin_client.post("/v3/datasets", dataset_b_json, content_type="application/json")
    assert res.status_code == 201
    res = admin_client.get("/v3/datasets?search=some_pid")
    assert res.data["count"] == 1

test_aggregation_and_filters#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
def test_aggregation_and_filters(
    admin_client, dataset_a_json, dataset_b_json, dataset_c_json, data_catalog, reference_data
):
    admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    admin_client.post("/v3/datasets", dataset_b_json, content_type="application/json")
    admin_client.post("/v3/datasets", dataset_c_json, content_type="application/json")
    res = admin_client.get("/v3/datasets")
    assert res.data["count"] == 3

    res = admin_client.get("/v3/datasets/aggregates")
    assert res.data != None
    assert list(sorted(key for key, value in res.data.items())) == [
        "access_type",
        "creator",
        "data_catalog",
        "field_of_science",
        "file_type",
        "infrastructure",
        "keyword",
        "organization",
        "project",
    ]
    aggregates = res.data

    for aggregate in aggregates.values():
        if len(aggregate["hits"]):
            value = (
                aggregate["hits"][0]["value"].get("fi")
                or aggregate["hits"][0]["value"].get("en")
                or aggregate["hits"][0]["value"].get("und")
            )
            count = aggregate["hits"][0]["count"]
            res = admin_client.get(f"/v3/datasets?{aggregate['query_parameter']}={value}")
            assert res.data["count"] == count

test_create_dataset_invalid_catalog#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
124
125
126
127
def test_create_dataset_invalid_catalog(admin_client, dataset_a_json):
    dataset_a_json["data_catalog"] = "urn:nbn:fi:att:data-catalog-does-not-exist"
    response = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert response.status_code == 400

test_create_dataset_invalid_language#

Try creating a dataset with an improperly formatted 'language' field. Each error case has a corresponding error message.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
@pytest.mark.parametrize(
    "value,expected_error",
    [
        ([{"url": "non_existent"}], "Language entries not found for given URLs: non_existent"),
        ([{"foo": "bar"}], "'url' field must be defined for each object in the list"),
        (["FI"], "Each item in the list must be an object with the field 'url'"),
        ("FI", 'Expected a list of items but got type "str".'),
    ],
)
def test_create_dataset_invalid_language(admin_client, dataset_a_json, value, expected_error):
    """
    Try creating a dataset with an improperly formatted 'language' field.
    Each error case has a corresponding error message.
    """
    dataset_a_json["language"] = value

    response = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert response.status_code == 400
    assert response.json()["language"] == [expected_error]

test_delete_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
151
152
153
154
155
156
157
def test_delete_dataset(admin_client, dataset_a_json, data_catalog, reference_data):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    id = res.data["id"]
    assert res.status_code == 201
    assert_nested_subdict(dataset_a_json, res.data, ignore=["generate_pid_on_publish"])
    res = admin_client.delete(f"/v3/datasets/{id}")
    assert res.status_code == 204

test_get_removed_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
def test_get_removed_dataset(admin_client, dataset_a_json, data_catalog, reference_data):
    # create a dataset and check it works
    res1 = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res1.status_code == 201
    id = res1.data["id"]
    res2 = admin_client.get(f"/v3/datasets/{id}")
    assert res2.status_code == 200
    assert_nested_subdict(dataset_a_json, res2.data, ignore=["generate_pid_on_publish"])

    # delete the dataset...
    res3 = admin_client.delete(f"/v3/datasets/{id}")
    assert res3.status_code == 204

    # ...and check that it
    # 1. cannot be found without query parameter
    res4 = admin_client.get(f"/v3/datasets/{id}")
    assert res4.status_code == 404
    # 2. can be found with the query parameter
    res5 = admin_client.get(f"/v3/datasets/{id}?include_removed=True")
    assert res5.status_code == 200
    assert_nested_subdict(dataset_a_json, res5.data, ignore=["generate_pid_on_publish"])

test_list_datasets_with_ordering#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
def test_list_datasets_with_ordering(
    admin_client, dataset_a_json, dataset_b_json, data_catalog, reference_data
):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    dataset_a_id = res.data["id"]
    dataset_a_pid = res.data["persistent_identifier"]
    res = admin_client.post("/v3/datasets", dataset_b_json, content_type="application/json")
    assert res.status_code == 201
    dataset_a_json["persistent_identifier"] = dataset_a_pid
    res = admin_client.put(
        f"/v3/datasets/{dataset_a_id}",
        dataset_a_json,
        content_type="application/json",
    )
    assert res.status_code == 200, res.data
    res = admin_client.get("/v3/datasets?ordering=created")
    assert_nested_subdict(
        {0: dataset_a_json, 1: dataset_b_json},
        dict(enumerate((res.data["results"]))),
        ignore=["generate_pid_on_publish"],
    )

    res = admin_client.get("/v3/datasets?ordering=modified")
    assert_nested_subdict(
        {0: dataset_b_json, 1: dataset_a_json},
        dict(enumerate((res.data["results"]))),
        ignore=["generate_pid_on_publish"],
    )

test_list_datasets_with_default_pagination#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
214
215
216
217
218
219
220
221
222
def test_list_datasets_with_default_pagination(admin_client, dataset_a, dataset_b):
    res = admin_client.get(reverse("dataset-list"))
    assert res.status_code == 200
    assert res.data == {
        "count": 2,
        "next": None,
        "previous": None,
        "results": [ANY, ANY],
    }

test_list_datasets_fields_param#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
225
226
227
228
229
230
231
232
233
def test_list_datasets_fields_param(admin_client, dataset_a, dataset_b):
    res = admin_client.get(reverse("dataset-list"), {"fields": "id,created"})
    assert res.status_code == 200
    assert res.data == {
        "count": 2,
        "next": None,
        "previous": None,
        "results": [{"id": ANY, "created": ANY}, {"id": ANY, "created": ANY}],
    }

test_list_datasets_faulty_fields_param#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
236
237
238
239
def test_list_datasets_faulty_fields_param(admin_client, dataset_a, dataset_b):
    res = admin_client.get(reverse("dataset-list"), {"fields": "id,foo,bar,created"})
    assert res.status_code == 400
    assert res.data == {"fields": "Fields not found in dataset: foo,bar"}

test_list_datasets_with_invalid_query_param#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
242
243
244
245
246
247
248
249
250
251
252
253
254
def test_list_datasets_with_invalid_query_param(admin_client, dataset_a):
    res = admin_client.get(reverse("dataset-list"), {"päginätiön": "false", "offzet": 10})
    assert res.status_code == 400
    assert res.json() == {
        "päginätiön": "Unknown query parameter",
        "offzet": "Unknown query parameter",
    }

    # unknown params should be ok in non-strict mode
    res = admin_client.get(
        reverse("dataset-list"), {"päginätiön": "false", "offzet": 10, "strict": "false"}
    )
    assert res.status_code == 200

test_list_datasets_with_pagination#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
257
258
259
260
261
262
263
264
265
def test_list_datasets_with_pagination(admin_client, dataset_a, dataset_b):
    res = admin_client.get(reverse("dataset-list"), {"pagination": "true"})
    assert res.status_code == 200
    assert res.data == {
        "count": 2,
        "next": None,
        "previous": None,
        "results": [ANY, ANY],
    }

test_list_datasets_with_no_pagination#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
268
269
270
271
def test_list_datasets_with_no_pagination(admin_client, dataset_a, dataset_b):
    res = admin_client.get(reverse("dataset-list"), {"pagination": "false"})
    assert res.status_code == 200
    assert res.data == [ANY, ANY]

test_create_dataset_with_metadata_owner#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
274
275
276
277
278
279
280
281
282
283
284
def test_create_dataset_with_metadata_owner(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json["metadata_owner"] = {
        "organization": "organization-a.fi",
        "user": "metax-user-a",
    }
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")

    assert res.status_code == 201
    assert_nested_subdict(dataset_a_json["metadata_owner"], res.data["metadata_owner"])

test_patch_metadata_owner#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
287
288
289
290
291
292
293
294
295
296
297
298
299
300
def test_patch_metadata_owner(admin_client, dataset_a_json, data_catalog, reference_data):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    new_owner = {
        "organization": "organization-a.fi",
        "user": "metax-user-a",
    }
    dataset_id = res.data["id"]
    res = admin_client.patch(
        f"/v3/datasets/{dataset_id}",
        {"metadata_owner": new_owner},
        content_type="application/json",
    )
    assert res.status_code == 200
    assert_nested_subdict(new_owner, res.data["metadata_owner"])

test_put_dataset_by_user#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
303
304
305
306
307
308
309
310
311
312
313
314
315
316
def test_put_dataset_by_user(
    user_client, dataset_a_json, data_catalog, reference_data, requests_mock
):
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    user = res.data["metadata_owner"]["user"]

    # metadata owner should remain unchanged by put
    dataset_a_json["persistent_identifier"] = res.data["persistent_identifier"]
    res = user_client.put(
        f"/v3/datasets/{res.data['id']}", dataset_a_json, content_type="application/json"
    )
    assert res.status_code == 200
    assert res.data["metadata_owner"]["user"] == user

test_patch_metadata_owner_not_allowed#

End-user cannot use custom metadata owner values.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
def test_patch_metadata_owner_not_allowed(
    user_client, dataset_a_json, data_catalog, reference_data
):
    """End-user cannot use custom metadata owner values."""
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    new_owner = {
        "organization": "organization-a.fi",
        "user": "metax-user-a",
    }
    dataset_id = res.data["id"]
    res = user_client.patch(
        f"/v3/datasets/{dataset_id}",
        {"metadata_owner": new_owner},
        content_type="application/json",
    )
    assert res.status_code == 400

test_owned_dataset#

End-user cannot use custom metadata owner values.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
def test_owned_dataset(
    admin_client, user_client, dataset_a_json, dataset_b_json, data_catalog, reference_data
):
    """End-user cannot use custom metadata owner values."""
    dataset_a_json["state"] = "published"
    dataset_b_json["state"] = "published"
    dataset_b_json["actors"] = [
        {"roles": ["creator", "publisher"], "organization": {"pref_label": {"en": "org"}}}
    ]
    res = admin_client.post("/v3/datasets", dataset_b_json, content_type="application/json")
    assert res.status_code == 201
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    dataset_id = res.data["id"]
    assert res.status_code == 201

    res = user_client.get(
        "/v3/datasets?only_owned_or_shared=false", content_type="application/json"
    )
    assert res.status_code == 200
    assert res.data["count"] == 2

    res = user_client.get(
        "/v3/datasets?only_owned_or_shared=true", content_type="application/json"
    )
    assert res.status_code == 200
    assert res.data["count"] == 1
    assert res.data["results"][0]["id"] == dataset_id

test_create_dataset_with_actor#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
366
367
368
def test_create_dataset_with_actor(dataset_c, data_catalog, reference_data):
    assert dataset_c.status_code == 201
    assert len(dataset_c.data["actors"]) == 2

test_modify_dataset_actor_roles#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
371
372
373
374
375
376
377
378
379
380
381
382
383
def test_modify_dataset_actor_roles(
    admin_client, dataset_c, dataset_c_json, data_catalog, reference_data
):
    assert dataset_c.status_code == 201
    assert len(dataset_c.data["actors"]) == 2
    dataset_c_json["actors"][0]["roles"].append("publisher")
    res = admin_client.put(
        f"/v3/datasets/{dataset_c.data['id']}", dataset_c_json, content_type="application/json"
    )
    assert res.status_code == 200
    assert len(res.data["actors"]) == 2
    assert "publisher" in res.data["actors"][0]["roles"]
    assert "creator" in res.data["actors"][0]["roles"]

test_create_dataset_with_other_identifiers#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
@pytest.mark.django_db
def test_create_dataset_with_other_identifiers(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json["other_identifiers"] = [
        {
            "identifier_type": {
                "url": "http://uri.suomi.fi/codelist/fairdata/identifier_type/code/doi"
            },
            "notation": "foo",
        },
        {"notation": "bar"},
        {"old_notation": "foo", "notation": "bar"},
    ]

    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    assert len(res.data["other_identifiers"]) == 3

test_update_dataset_with_other_identifiers#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
@pytest.mark.django_db
def test_update_dataset_with_other_identifiers(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    # Create a dataset with other_identifiers
    dataset_a_json["other_identifiers"] = [
        {
            "identifier_type": {
                "url": "http://uri.suomi.fi/codelist/fairdata/identifier_type/code/doi"
            },
            "notation": "foo",
        },
        {"notation": "bar"},
        {"old_notation": "foo", "notation": "baz"},
    ]

    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    assert len(res.data["other_identifiers"]) == 3
    assert_nested_subdict(dataset_a_json["other_identifiers"], res.data["other_identifiers"])

    ds_id = res.data["id"]

    # Update other_identifiers
    dataset_a_json["other_identifiers"] = [
        {
            "identifier_type": {
                "url": "http://uri.suomi.fi/codelist/fairdata/identifier_type/code/urn"
            },
            "notation": "foo",
        },
        {
            "identifier_type": {
                "url": "http://uri.suomi.fi/codelist/fairdata/identifier_type/code/doi"
            },
            "notation": "new_foo",
        },
        {"notation": "updated_bar"},
        {"old_notation": "updated_foo", "notation": "updated_baz"},
    ]
    res = admin_client.patch(
        f"/v3/datasets/{ds_id}", dataset_a_json, content_type="application/json"
    )
    assert res.status_code == 200
    assert len(res.data["other_identifiers"]) == 4
    assert_nested_subdict(dataset_a_json["other_identifiers"], res.data["other_identifiers"])

    # Assert that the old other_identifiers don't exist in the db anymore
    new_count = OtherIdentifier.available_objects.all().count()
    assert new_count == 4

test_dataset_put_maximal_and_minimal#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
def test_dataset_put_maximal_and_minimal(
    admin_client, dataset_maximal_json, reference_data, data_catalog
):
    res = admin_client.post(
        "/v3/datasets?include_nulls=true", dataset_maximal_json, content_type="application/json"
    )
    assert res.status_code == 201
    assert_nested_subdict(dataset_maximal_json, res.json())

    minimal_json = {
        "data_catalog": dataset_maximal_json["data_catalog"],
        "title": dataset_maximal_json["title"],
        "generate_pid_on_publish": "URN",
    }
    res = admin_client.put(
        f"/v3/datasets/{res.data['id']}?include_nulls=true",
        minimal_json,
        content_type="application/json",
    )
    assert res.status_code == 200

    # writable fields not in minimal_json should be cleared to falsy values
    assert list(sorted(key for key, value in res.data.items() if value)) == [
        "api_version",
        "created",
        "data_catalog",
        "dataset_versions",
        "draft_revision",
        "generate_pid_on_publish",
        "id",
        "metadata_owner",
        "metadata_repository",
        "modified",
        "state",
        "title",
        "version",
    ]

test_dataset_patch_maximal_and_minimal#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
def test_dataset_patch_maximal_and_minimal(
    admin_client, dataset_maximal_json, reference_data, data_catalog
):
    res = admin_client.post("/v3/datasets", dataset_maximal_json, content_type="application/json")
    assert res.status_code == 201
    maximal_data = res.json()

    minimal_json = {
        "description": {"en": "new description"},
    }
    res = admin_client.patch(
        f"/v3/datasets/{res.data['id']}", minimal_json, content_type="application/json"
    )
    assert res.status_code == 200
    minimal_data = res.json()

    # fields in minimal_json should replace values in maximal, others unchanged
    assert minimal_data == {
        **maximal_data,
        **minimal_json,
        "modified": matchers.DateTimeStr(),  # match any datetime
        "draft_revision": 2,  # increased by 1 on save
    }

test_dataset_put_remove_fileset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
def test_dataset_put_remove_fileset(
    admin_client, dataset_maximal_json, reference_data, data_catalog
):
    dataset_maximal_json.pop("remote_resources")
    FileStorageFactory(storage_service="ida", csc_project="project")
    dataset_maximal_json["fileset"] = {
        "storage_service": "ida",
        "csc_project": "project",
    }
    res = admin_client.post("/v3/datasets", dataset_maximal_json, content_type="application/json")
    assert res.status_code == 201

    # PUT without fileset would remove existing fileset which is not allowed
    minimal_json = {
        "data_catalog": dataset_maximal_json["data_catalog"],
        "title": dataset_maximal_json["title"],
    }
    res = admin_client.put(
        f"/v3/datasets/{res.data['id']}", minimal_json, content_type="application/json"
    )
    assert res.status_code == 400
    assert "not allowed" in res.json()["fileset"]

test_dataset_restricted#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
546
547
548
549
550
551
552
553
554
555
def test_dataset_restricted(admin_client, dataset_a_json, reference_data, data_catalog):
    dataset_a_json["access_rights"]["access_type"] = {
        "url": "http://uri.suomi.fi/codelist/fairdata/access_type/code/restricted"
    }
    dataset_a_json["access_rights"]["restriction_grounds"] = [
        {"url": "http://uri.suomi.fi/codelist/fairdata/restriction_grounds/code/research"}
    ]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    assert_nested_subdict(dataset_a_json, res.json(), ignore=["generate_pid_on_publish"])

test_create_dataset_require_data_catalog#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
558
559
560
561
562
563
564
565
def test_create_dataset_require_data_catalog(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json["state"] = "published"
    dataset_a_json.pop("data_catalog")
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 400
    assert "Dataset has to have a data catalog when publishing" in str(res.data["data_catalog"])

test_create_dataset_draft_without_catalog#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
def test_create_dataset_draft_without_catalog(
    user_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json.pop("data_catalog")
    dataset_a_json["state"] = "draft"
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    assert_nested_subdict(dataset_a_json, res.data)

    # make sure the dataset can also be updated
    res = user_client.patch(
        f"/v3/datasets/{res.data['id']}",
        {"title": {"en": "test"}},
        content_type="application/json",
    )
    assert res.status_code == 200

test_flush_dataset_by_service#

Flush should delete dataset from database.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
def test_flush_dataset_by_service(service_client, dataset_a_json, data_catalog, reference_data):
    """Flush should delete dataset from database."""
    res = service_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    dataset_id = res.data["id"]

    # Not harvested, not catalog admin
    res = service_client.delete(f"/v3/datasets/{dataset_id}?flush=true")
    assert res.status_code == 403

    # Harvested, not catalog admin
    data_catalog.harvested = True
    data_catalog.save()
    res = service_client.delete(f"/v3/datasets/{dataset_id}?flush=true")
    assert res.status_code == 403

    # Not harvested, catalog admin
    data_catalog.harvested = False
    data_catalog.save()
    data_catalog.dataset_groups_admin.add(Group.objects.get(name="test"))
    res = service_client.delete(f"/v3/datasets/{dataset_id}?flush=true")
    assert res.status_code == 403

    # Both harvested and catalog admin -> ok
    data_catalog.harvested = True
    data_catalog.save()
    res = service_client.delete(f"/v3/datasets/{dataset_id}?flush=true")
    assert res.status_code == 204
    assert not Dataset.all_objects.filter(id=dataset_id).exists()

test_flush_dataset_by_user#

Flush should not be allowed for regular user.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
616
617
618
619
620
621
622
623
624
625
626
627
628
def test_flush_dataset_by_user(user_client, dataset_a_json, data_catalog, reference_data):
    """Flush should not be allowed for regular user."""
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    id = res.data["id"]
    res = user_client.delete(f"/v3/datasets/{id}?flush=true")
    assert res.status_code == 403
    assert Dataset.available_objects.filter(id=id).exists()

    # Delete without flush should only soft delete
    res = user_client.delete(f"/v3/datasets/{id}")
    assert res.status_code == 204
    assert Dataset.all_objects.filter(id=id).exists()
    assert not Dataset.available_objects.filter(id=id).exists()

test_flush_draft#

Flush should be allowed for drafts.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
631
632
633
634
635
636
637
638
def test_flush_draft(user_client, dataset_a_json, data_catalog, reference_data):
    """Flush should be allowed for drafts."""
    dataset_a_json["state"] = "draft"
    res = user_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    id = res.data["id"]
    res = user_client.delete(f"/v3/datasets/{id}?flush=true")
    assert res.status_code == 204
    assert not Dataset.available_objects.filter(id=id).exists()

ida_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
641
642
643
644
645
646
@pytest.fixture
def ida_dataset(data_catalog, reference_data):
    ida_storage = factories.FileStorageFactory(storage_service="ida", csc_project="project")
    dataset = factories.DatasetFactory(data_catalog=data_catalog)
    factories.FileSetFactory(dataset=dataset, storage=ida_storage)
    return dataset

ida_dataset_other#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
649
650
651
652
653
654
@pytest.fixture
def ida_dataset_other(data_catalog, reference_data):
    ida_storage = factories.FileStorageFactory(storage_service="ida", csc_project="other_project")
    dataset = factories.DatasetFactory(data_catalog=data_catalog)
    factories.FileSetFactory(dataset=dataset, storage=ida_storage)
    return dataset

pas_dataset#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
657
658
659
660
661
662
@pytest.fixture
def pas_dataset(data_catalog, reference_data):
    pas_storage = factories.FileStorageFactory(storage_service="pas", csc_project="project")
    dataset = factories.DatasetFactory(data_catalog=data_catalog)
    factories.FileSetFactory(dataset=dataset, storage=pas_storage)
    return dataset

test_filter_by_storage_service#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
def test_filter_by_storage_service(admin_client, ida_dataset, ida_dataset_other, pas_dataset):
    res = admin_client.get(
        "/v3/datasets?storage_services=ida&pagination=false", content_type="application/json"
    )
    assert {d["id"] for d in res.data} == {str(ida_dataset.id), str(ida_dataset_other.id)}

    res = admin_client.get(
        "/v3/datasets?storage_services=pas&pagination=false", content_type="application/json"
    )
    assert {d["id"] for d in res.data} == {str(pas_dataset.id)}

    res = admin_client.get(
        "/v3/datasets?storage_services=ida,pas&pagination=false", content_type="application/json"
    )
    assert {d["id"] for d in res.data} == {
        str(pas_dataset.id),
        str(ida_dataset.id),
        str(ida_dataset_other.id),
    }

test_filter_by_csc_project#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
def test_filter_by_csc_project(admin_client, ida_dataset, ida_dataset_other, pas_dataset):
    res = admin_client.get(
        "/v3/datasets?csc_projects=project&pagination=false", content_type="application/json"
    )
    assert {d["id"] for d in res.data} == {str(ida_dataset.id), str(pas_dataset.id)}

    res = admin_client.get(
        "/v3/datasets?csc_projects=other_project&pagination=false", content_type="application/json"
    )
    assert {d["id"] for d in res.data} == {str(ida_dataset_other.id)}

    res = admin_client.get(
        "/v3/datasets?csc_projects=project,other_project&pagination=false",
        content_type="application/json",
    )
    assert {d["id"] for d in res.data} == {
        str(pas_dataset.id),
        str(ida_dataset.id),
        str(ida_dataset_other.id),
    }

test_filter_by_has_files#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
708
709
710
711
712
713
714
715
716
717
718
def test_filter_by_has_files(
    admin_client, dataset_a, dataset_with_files, data_catalog, reference_data
):
    res = admin_client.get(
        "/v3/datasets?has_files=false&pagination=false", content_type="application/json"
    )
    assert [d["id"] for d in res.data] == [dataset_a.dataset_id]
    res = admin_client.get(
        "/v3/datasets?has_files=true&pagination=false", content_type="application/json"
    )
    assert [d["id"] for d in res.data] == [str(dataset_with_files.id)]

test_empty_description#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
721
722
723
724
725
726
727
728
729
730
731
def test_empty_description(admin_client, dataset_a_json, data_catalog, reference_data):
    dataset_a_json["state"] = "draft"
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    dataset_id = res.data["id"]
    res = admin_client.patch(
        f"/v3/datasets/{dataset_id}?include_nulls=true",
        {"description": {"fi": "", "en": None}},
        content_type="application/json",
    )
    assert res.status_code == 200
    assert res.json()["description"] == None

test_api_version#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
734
735
736
737
738
739
740
741
def test_api_version(admin_client, dataset_a_json, data_catalog, reference_data):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    dataset_id = res.data["id"]
    assert res.data["api_version"] == 3
    Dataset.objects.filter(id=dataset_id).update(api_version=2)
    res = admin_client.patch(f"/v3/datasets/{dataset_id}", {}, content_type="application/json")
    assert res.status_code == 200
    assert res.data["api_version"] == 3

test_dataset_last_modified_by#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
744
745
746
747
748
749
750
751
752
753
def test_dataset_last_modified_by(admin_client, user_client, user):
    dataset = DatasetFactory(metadata_owner=MetadataProviderFactory(user=user))

    admin_client.patch(f"/v3/datasets/{dataset.id}", {}, content_type="application/json")
    dataset.refresh_from_db()
    assert dataset.last_modified_by.username == "admin"

    user_client.patch(f"/v3/datasets/{dataset.id}", {}, content_type="application/json")
    dataset.refresh_from_db()
    assert dataset.last_modified_by == user

test_dataset_expanded_catalog#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
756
757
758
759
760
761
762
763
764
765
766
def test_dataset_expanded_catalog(admin_client, dataset_a_json, data_catalog, reference_data):
    res1 = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res1.status_code == 201

    res2 = admin_client.get(f"/v3/datasets/{res1.data['id']}")
    assert res2.status_code == 200
    assert isinstance(res2.data["data_catalog"], str)

    res3 = admin_client.get(f"/v3/datasets/{res1.data['id']}?expand_catalog=true")
    assert res3.status_code == 200
    assert isinstance(res3.data["data_catalog"], dict)

test_many_actors#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
769
770
771
772
773
774
775
776
def test_many_actors(admin_client, dataset_a_json, data_catalog, reference_data):
    org = {"organization": {"pref_label": {"en": "organization"}}}
    dataset_a_json["actors"] = [
        {"person": {"name": f"Test person {i}"}, **org, "roles": ["creator"]} for i in range(100)
    ]
    dataset_a_json["actors"][0]["roles"].append("publisher")
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201

test_dataset_search_entry#

Check that search entry contains correct values from dataset.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
779
780
781
782
783
784
785
786
787
788
789
790
791
792
def test_dataset_search_entry(admin_client, dataset_a_json, data_catalog, reference_data):
    """Check that search entry contains correct values from dataset."""
    dataset_a_json["other_identifiers"] = [{"notation": "doi:other_identifier"}]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    dataset_id = res.data["id"]

    entry = SearchEntry.objects.get(object_id=dataset_id)
    assert res.data["persistent_identifier"] in entry.title  # pid
    assert "Test dataset" in entry.title  # title
    assert "test subjects (persons)" in entry.title  # theme
    assert "Test dataset desc" in entry.description  # description
    assert "keyword another_keyword" in entry.description  # keywords
    assert "doi:other_identifier" in entry.content  # entity.notation

test_create_dataset_with_extra_fields#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
795
796
797
798
799
800
801
802
803
804
805
806
def test_create_dataset_with_extra_fields(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json["test-abcd"] = "non valid field"
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 400
    assert str(res.data["test-abcd"][0]) == "Unexpected field"
    # with strict=false the dataset creation should get through
    res = admin_client.post(
        "/v3/datasets?strict=false", dataset_a_json, content_type="application/json"
    )
    assert res.status_code == 201

test_dataset_citation#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
809
810
811
812
813
814
815
def test_dataset_citation(admin_client):
    minimal_json = {"title": {"en": "hello world"}, "bibliographic_citation": "   some text here "}
    res = admin_client.post("/v3/datasets", minimal_json, content_type="application/json")
    assert res.status_code == 201
    assert res.data == matchers.DictContaining(
        {"title": {"en": "hello world"}, "bibliographic_citation": "some text here"}
    )

test_get_dataset_include_nulls#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
818
819
820
821
822
823
824
825
826
827
def test_get_dataset_include_nulls(admin_client, dataset_a_json, data_catalog, reference_data):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201
    dataset_id = res.data["id"]
    res = admin_client.get(f"/v3/datasets/{dataset_id}", content_type="application/json")
    assert "deprecated" not in res.data
    res = admin_client.get(
        f"/v3/datasets/{dataset_id}?include_nulls=true", content_type="application/json"
    )
    assert res.data["deprecated"] == None

test_missing_required_fields#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
def test_missing_required_fields(
    admin_client, data_catalog, data_catalog_harvested, reference_data
):
    # title
    dataset = {"state": "published"}
    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")
    assert res.status_code == 400
    assert "This field is required." in str(res.data["title"])

    dataset = {
        **dataset,
        "title": {"en": "test"},
        "actors": [{"roles": [], "organization": {"pref_label": {"en": "Test Org"}}}],
    }
    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")

    data = res.json()
    assert res.status_code == 400
    assert "Dataset has to have a data catalog when publishing." in data["data_catalog"]
    assert "Dataset has to have access rights when publishing." in data["access_rights"]
    assert "Dataset has to have a description when publishing." in data["description"]
    assert "An actor with creator role is required." in data["actors"]
    assert "Exactly one actor with publisher role is required." in data["actors"]
    assert "All actors in a published dataset should have at least one role." in data["actors"]

    # data_catalog(ida), access_rights, description, actors
    dataset = {
        **dataset,
        "data_catalog": "urn:nbn:fi:att:data-catalog-ida",
        "access_rights": {
            "access_type": {"url": "http://uri.suomi.fi/codelist/fairdata/access_type/code/open"},
        },
        "description": {"en": "test"},
        "actors": [
            {
                "roles": ["creator", "publisher"],
            }
        ],
    }

    # actors[].organization
    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")
    assert res.status_code == 400
    assert (
        str(res.data["actors"][0]["organization"])
        == str(res.data["actors"][0]["organization"])
        == "This field is required"
    )

    dataset = {
        **dataset,
        "actors": [
            {
                "roles": ["creator", "publisher"],
                "organization": {"pref_label": {"en": "test org"}},
            }
        ],
    }

    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")
    assert res.status_code == 400
    assert "Dataset has to have a license when publishing." in str(res.data["access_rights"])

    # access_rights.license
    dataset["access_rights"]["license"] = [
        {"url": "http://uri.suomi.fi/codelist/fairdata/license/code/CC-BY-4.0"}
    ]

    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")
    assert res.status_code == 400
    assert "required by the catalog when publishing" in str(res.data["generate_pid_on_publish"])

    # data_catalog(harvested), access_rights, description, actors, license
    harvested_dataset = {**dataset, "data_catalog": "urn:nbn:fi:att:data-catalog-harvested"}
    harvested_res = admin_client.post(
        "/v3/datasets", harvested_dataset, content_type="application/json"
    )

    assert harvested_res.status_code == 400
    assert "Dataset has to have a persistent identifier" in str(
        harvested_res.data["persistent_identifier"]
    )

    # harvested_catalog: persistent_identifier
    harvested_dataset = {**harvested_dataset, "persistent_identifier": "PID"}
    harvested_res = admin_client.post(
        "/v3/datasets", harvested_dataset, content_type="application/json"
    )
    assert harvested_res.status_code == 201

    # continue with ida-dataset
    # ida-catalog: pid_type
    dataset = {**dataset, "generate_pid_on_publish": "URN"}
    res = admin_client.post("/v3/datasets", dataset, content_type="application/json")
    assert res.status_code == 201

test_missing_restriction_grounds#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
def test_missing_restriction_grounds(admin_client, dataset_a_json, data_catalog, reference_data):
    dataset_a_json["access_rights"]["access_type"] = {
        "url": "http://uri.suomi.fi/codelist/fairdata/access_type/code/restricted"
    }
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 400
    assert (
        "Dataset access rights has to contain restriction grounds if access type is not 'Open'."
        in str(res.data["access_rights"])
    )

    dataset_a_json["access_rights"]["restriction_grounds"] = [
        {"url": "http://uri.suomi.fi/codelist/fairdata/restriction_grounds/code/research"}
    ]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201

test_unnecessary_restriction_grounds#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
945
946
947
948
949
950
951
952
953
954
955
956
957
def test_unnecessary_restriction_grounds(
    admin_client, dataset_a_json, data_catalog, reference_data
):
    dataset_a_json["access_rights"]["restriction_grounds"] = [
        {"url": "http://uri.suomi.fi/codelist/fairdata/restriction_grounds/code/research"}
    ]
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 400
    assert "Open datasets do not accept restriction grounds." in str(res.data["access_rights"])

    dataset_a_json["access_rights"].pop("restriction_grounds")
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    assert res.status_code == 201

test_get_dataset_no_multiple_objects_error#

Check that multiple groups in dataset_groups_admin don't produce MultipleObjectsReturned error in query.

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
960
961
962
963
964
965
966
967
def test_get_dataset_no_multiple_objects_error(service_client):
    """Check that multiple groups in dataset_groups_admin don't produce MultipleObjectsReturned error in query."""
    data_catalog = factories.DataCatalogFactory()
    data_catalog.dataset_groups_admin.add(Group.objects.get(name="test"))
    data_catalog.dataset_groups_admin.add(Group.objects.create(name="somegroup"))
    dataset = factories.PublishedDatasetFactory(data_catalog=data_catalog)
    res = service_client.patch(f"/v3/datasets/{dataset.id}", {}, content_type="application/json")
    assert res.status_code == 200

test_not_a_list#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
def test_not_a_list(admin_client, dataset_a_json, data_catalog, reference_data, monkeypatch):
    res = admin_client.post("/v3/datasets", dataset_a_json, content_type="application/json")
    provenances = {
        "spatial": {
            "geographic_name": "Tapiola",
            "reference": {"url": "http://www.yso.fi/onto/yso/p105747"},
        },
    }

    dataset_id = res.data["id"]
    res = admin_client.patch(
        f"/v3/datasets/{dataset_id}",
        {"provenance": provenances},
        content_type="application/json",
    )
    assert res.status_code == 400
    assert "Expected a list" in res.json()["provenance"]["non_field_errors"][0]

test_remove_rights#

Source code in tests/unit/apps/core/api/dataset/test_dataset.py 
989
990
991
992
993
994
995
996
997
998
def test_remove_rights(admin_client, dataset_a_json, data_catalog, reference_data):
    dataset = factories.DatasetFactory()
    dataset_id = dataset.id
    res = admin_client.patch(
        f"/v3/datasets/{dataset_id}",
        {"access_rights": None},
        content_type="application/json",
    )
    assert res.status_code == 200
    assert "access_rights" not in res.data