arkindex-base-worker 0.5.2a1__py3-none-any.whl → 0.5.2a2__py3-none-any.whl

{arkindex_base_worker-0.5.2a1.dist-info → arkindex_base_worker-0.5.2a2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: arkindex-base-worker
-Version: 0.5.2a1
+Version: 0.5.2a2
 Summary: Base Worker to easily build Arkindex ML workflows
 Author-email: Teklia <contact@teklia.com>
 Maintainer-email: Teklia <contact@teklia.com>
@@ -23,6 +23,7 @@ Requires-Dist: humanize==4.15.0
 Requires-Dist: peewee~=3.17
 Requires-Dist: Pillow==11.3.0
 Requires-Dist: python-gnupg==0.5.6
+Requires-Dist: python-magic==0.4.27
 Requires-Dist: shapely==2.0.6
 Requires-Dist: teklia-toolbox==0.1.12
 Requires-Dist: zstandard==0.25.0

{arkindex_base_worker-0.5.2a1.dist-info → arkindex_base_worker-0.5.2a2.dist-info}/RECORD

@@ -1,10 +1,10 @@
-arkindex_base_worker-0.5.2a1.dist-info/licenses/LICENSE,sha256=NVshRi1efwVezMfW7xXYLrdDr2Li1AfwfGOd5WuH1kQ,1063
+arkindex_base_worker-0.5.2a2.dist-info/licenses/LICENSE,sha256=NVshRi1efwVezMfW7xXYLrdDr2Li1AfwfGOd5WuH1kQ,1063
 arkindex_worker/__init__.py,sha256=Sdt5KXn8EgURb2MurYVrUWaHbH3iFA1XLRo0Lc5AJ44,250
 arkindex_worker/cache.py,sha256=XpEXMSnbhYCvrJquwA9XXqZo-ajMLpaCxKG5wH3Gp6Y,10959
-arkindex_worker/image.py,sha256=sGE8to5iykXv25bpkftOEWzlh5NzBZSKy4lSRoHYHPU,20929
+arkindex_worker/image.py,sha256=9KeZHWNIDkwNJZR0y-mbyD_pvKfrgdktMB32jZqSMYk,20927
 arkindex_worker/models.py,sha256=DgKvAB_2e1cPcuUavZkyTkV10jBK8y083oVklB9idSk,10855
 arkindex_worker/utils.py,sha256=Eqg5pGAuOmuwMT3EhKTQDMek7wHC1KzZL7XXqYVVfHY,10977
-arkindex_worker/worker/__init__.py,sha256=SzD0s1_m6gMV02EUF-NeciqZdVPA4dpXI84tSj-g494,17869
+arkindex_worker/worker/__init__.py,sha256=tM_ynAARmtuJw5YWb_jI0AD5KNXbWN1K-VDiixIp7O4,18009
 arkindex_worker/worker/base.py,sha256=-R_aLMJHbR6X1uM-U0zExsF_KLy5Wl3WJ_YMGO9We0I,22153
 arkindex_worker/worker/classification.py,sha256=qvykymkgd4nGywHCxL8obo4egstoGsmWNS4Ztc1qNWQ,11024
 arkindex_worker/worker/corpus.py,sha256=MeIMod7jkWyX0frtD0a37rhumnMV3p9ZOC1xwAoXrAA,2291
@@ -14,7 +14,7 @@ arkindex_worker/worker/entity.py,sha256=Aj6EOfzHEm7qQV-Egm0YKLZgCrLS_3ggOKTY81M2
 arkindex_worker/worker/image.py,sha256=L6Ikuf0Z0RxJk7JarY5PggJGrYSHLaPK0vn0dy0CIaQ,623
 arkindex_worker/worker/metadata.py,sha256=keZdOdUthSH2hAw9iet5pN7rzWihTUYjZHRGTEjaltw,6843
 arkindex_worker/worker/process.py,sha256=9TEHpMcBax1wc6PrWMMrdXe2uNfqyVj7n_dAYZRBGnY,1854
-arkindex_worker/worker/task.py,sha256=nYfMSFm_d-4t8y4PO4HjFBnLsZf7IsDjkS7-A2Pgnac,1525
+arkindex_worker/worker/task.py,sha256=HASQU5LYVtgvCnRCLFC6iH7h7v6q_usZNZ-r_Wkv9A8,3306
 arkindex_worker/worker/training.py,sha256=tyQOHcwv--_wdYz6CgLEe1YM7kwwwKN30LvGTsnWd78,10923
 arkindex_worker/worker/transcription.py,sha256=sw718R119tsLNY8inPMVeIilvFJo94fMbMtYgH0zTM8,21250
 examples/standalone/python/worker.py,sha256=Zr4s4pHvgexEjlkixLFYZp1UuwMLeoTxjyNG5_S2iYE,6672
@@ -24,7 +24,7 @@ tests/__init__.py,sha256=DG--S6IpGl399rzSAjDdHL76CkOIeZIjajCcyUSDhOQ,241
 tests/conftest.py,sha256=Tp7YFK17NATwF2yAcBwi0QFNyKSXtLS0VhZ-zZngsQI,24343
 tests/test_base_worker.py,sha256=lwS4X3atS2ktEKd1XdogmN3mbzq-tO206-k_0EDITlw,29302
 tests/test_cache.py,sha256=_wztzh94EwVrb8UvpFqgl2aa2_FLaCcJKaqunCYR5Dw,10435
-tests/test_dataset_worker.py,sha256=iDJM2C4PfQNH0r4_QqSWoPt8BcM0geUUdODtWY0Z9PA,22412
+tests/test_dataset_worker.py,sha256=LmL3ERF1__PUPkTLiAFC0IYglZTv5WQYA42Vm-uhe2w,22023
 tests/test_element.py,sha256=hlj5VSF4plwC7uz9R4LGOOXZJQcHZiYCIDZT5V6EIB8,14334
 tests/test_image.py,sha256=yAM5mMfpQcIurT1KLHmu0AhSX2Qm3YvCu7afyZ3XUdU,28314
 tests/test_merge.py,sha256=REpZ13jkq_qm_4L5URQgFy5lxvPZtXxQEiWfYLMdmF0,7956
@@ -44,7 +44,7 @@ tests/test_elements_worker/test_entity.py,sha256=SNAZEsVVLnqlliOmjkgv_cZhw0bAuJU
 tests/test_elements_worker/test_image.py,sha256=BljMNKgec_9a5bzNzFpYZIvSbuvwsWDfdqLHVJaTa7M,2079
 tests/test_elements_worker/test_metadata.py,sha256=qtTDtlp3VnBkfck7PAguK2dEgTLlr1i1EVnmNTeNf3A,20515
 tests/test_elements_worker/test_process.py,sha256=y4RoVhPfyHzR795fw7-_FXElBcKo3fy4Ew_HI-kxJic,3088
-tests/test_elements_worker/test_task.py,sha256=wTUWqN9UhfKmJn3IcFY75EW4I1ulRhisflmY1kmP47s,5574
+tests/test_elements_worker/test_task.py,sha256=oHwP1fbJftXFA2U4qA3Gb4vX-iJoV-sBvPHnfBBpRrc,8906
 tests/test_elements_worker/test_training.py,sha256=qgK7BLucddRzc8ePbQtY75x17QvGDEq5XCwgyyvmAJE,8717
 tests/test_elements_worker/test_transcription_create.py,sha256=yznO9B_BVsOR0Z_VY5ZL8gJp0ZPCz_4sPUs5dXtixAg,29281
 tests/test_elements_worker/test_transcription_create_with_elements.py,sha256=tmcyglgssEqMnt1Mdy_u6X1m2wgLWTo_HdWst3GrK2k,33056
@@ -55,7 +55,7 @@ worker-demo/tests/conftest.py,sha256=XzNMNeg6pmABUAH8jN6eZTlZSFGLYjS3-DTXjiRN6Yc
 worker-demo/tests/test_worker.py,sha256=3DLd4NRK4bfyatG5P_PK4k9P9tJHx9XQq5_ryFEEFVg,304
 worker-demo/worker_demo/__init__.py,sha256=2BPomV8ZMNf3YXJgloatKeHQCE6QOkwmsHGkO6MkQuM,125
 worker-demo/worker_demo/worker.py,sha256=Rt-DjWa5iBP08k58NDZMfeyPuFbtNcbX6nc5jFX7GNo,440
-arkindex_base_worker-0.5.2a1.dist-info/METADATA,sha256=AwYp_xJZzu6zAtvnvZjeK_W29tzqvRuwYnxwMYcKSIc,1849
-arkindex_base_worker-0.5.2a1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-arkindex_base_worker-0.5.2a1.dist-info/top_level.txt,sha256=-vNjP2VfROx0j83mdi9aIqRZ88eoJjxeWz-R_gPgyXU,49
-arkindex_base_worker-0.5.2a1.dist-info/RECORD,,
+arkindex_base_worker-0.5.2a2.dist-info/METADATA,sha256=LyPpeyvKIadAuqir1cymTwxoWm3XovhF-JmzQ1LW0MI,1885
+arkindex_base_worker-0.5.2a2.dist-info/WHEEL,sha256=qELbo2s1Yzl39ZmrAibXA2jjPLUYfnVhUNTlyF1rq0Y,92
+arkindex_base_worker-0.5.2a2.dist-info/top_level.txt,sha256=-vNjP2VfROx0j83mdi9aIqRZ88eoJjxeWz-R_gPgyXU,49
+arkindex_base_worker-0.5.2a2.dist-info/RECORD,,

{arkindex_base_worker-0.5.2a1.dist-info → arkindex_base_worker-0.5.2a2.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.9.0)
+Generator: setuptools (80.10.1)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

arkindex_worker/image.py

@@ -38,7 +38,7 @@ if TYPE_CHECKING:
     from arkindex_worker.models import Element
 
 # See http://docs.python-requests.org/en/master/user/advanced/#timeouts
-DOWNLOAD_TIMEOUT = (30, 60)
+REQUEST_TIMEOUT = (30, 60)
 
 BoundingBox = namedtuple("BoundingBox", ["x", "y", "width", "height"])
 
@@ -346,7 +346,7 @@ def _retried_request(url, *args, method=requests.get, **kwargs):
         url,
         *args,
         headers={"User-Agent": IIIF_USER_AGENT},
-        timeout=DOWNLOAD_TIMEOUT,
+        timeout=REQUEST_TIMEOUT,
         verify=should_verify_cert(url),
         **kwargs,
     )

arkindex_worker/worker/__init__.py

@@ -424,12 +424,13 @@ class DatasetWorker(DatasetMixin, BaseWorker, TaskMixin):
         failed = 0
         for i, dataset_set in enumerate(dataset_sets, start=1):
             try:
-                assert dataset_set.dataset.state == DatasetState.Complete.value, (
-                    "When processing a set, its dataset state should be Complete."
-                )
-
-                logger.info(f"Retrieving data for {dataset_set} ({i}/{count})")
-                self.download_dataset_artifact(dataset_set.dataset)
+                if dataset_set.dataset.state == DatasetState.Complete.value:
+                    logger.info(f"Retrieving data for {dataset_set} ({i}/{count})")
+                    self.download_dataset_artifact(dataset_set.dataset)
+                else:
+                    logger.warning(
+                        f"The dataset {dataset_set.dataset} has its state set to `{dataset_set.dataset.state}`, its archive will not be downloaded"
+                    )
 
                 logger.info(f"Processing {dataset_set} ({i}/{count})")
                 self.process_set(dataset_set)
@@ -444,7 +445,7 @@ class DatasetWorker(DatasetMixin, BaseWorker, TaskMixin):
 
                 logger.warning(message, exc_info=e if self.args.verbose else None)
 
-        # Cleanup the latest downloaded dataset artifact
+        # Cleanup the latest downloaded dataset artifact (if needed)
         self.cleanup_downloaded_artifact()
 
         message = f"Ran on {count} {pluralize('set', count)}: {count - failed} completed, {failed} failed"

arkindex_worker/worker/task.py

@@ -4,9 +4,16 @@ BaseWorker methods for tasks.
 
 import uuid
 from collections.abc import Iterator
+from http.client import REQUEST_TIMEOUT
+from pathlib import Path
+
+import magic
+import requests
 
 from arkindex.compat import DownloadedFile
+from arkindex_worker import logger
 from arkindex_worker.models import Artifact
+from teklia_toolbox.requests import should_verify_cert
 
 
 class TaskMixin:
@@ -45,3 +52,49 @@ class TaskMixin:
         return self.api_client.request(
             "DownloadArtifact", id=task_id, path=artifact.path
         )
+
+    def upload_artifact(self, path: Path) -> None:
+        """
+        Upload a single file as an Artifact of the current task.
+
+        :param path: Path of the single file to upload as an Artifact.
+        """
+        assert path and isinstance(path, Path) and path.exists(), (
+            "path shouldn't be null, should be a Path and should exist"
+        )
+
+        if self.is_read_only:
+            logger.warning("Cannot upload artifact as this worker is in read-only mode")
+            return
+
+        # Get path relative to task's data directory
+        relpath = str(path.relative_to(self.work_dir))
+
+        # Get file size
+        size = path.stat().st_size
+
+        # Detect content type
+        try:
+            content_type = magic.from_file(path, mime=True)
+        except Exception as e:
+            logger.warning(f"Failed to get a mime type for {path}: {e}")
+            content_type = "application/octet-stream"
+
+        # Create artifact on API to get an S3 url
+        artifact = self.api_client.request(
+            "CreateArtifact",
+            id=self.task_id,
+            body={"path": relpath, "content_type": content_type, "size": size},
+        )
+
+        # Upload the file content to S3
+        s3_put_url = artifact["s3_put_url"]
+        with path.open("rb") as content:
+            resp = requests.put(
+                s3_put_url,
+                data=content,
+                headers={"Content-Type": content_type},
+                timeout=REQUEST_TIMEOUT,
+                verify=should_verify_cert(s3_put_url),
+            )
+            resp.raise_for_status()

tests/test_dataset_worker.py

@@ -435,34 +435,6 @@ def test_run_no_sets(mocker, caplog, mock_dataset_worker):
     ]
 
 
-def test_run_initial_dataset_state_error(
-    mocker, responses, caplog, mock_dataset_worker, default_dataset
-):
-    default_dataset.state = DatasetState.Building.value
-    mocker.patch(
-        "arkindex_worker.worker.DatasetWorker.list_sets",
-        return_value=[Set(name="train", dataset=default_dataset)],
-    )
-
-    with pytest.raises(SystemExit):
-        mock_dataset_worker.run()
-
-    assert len(responses.calls) == len(BASE_API_CALLS) * 2
-    assert [
-        (call.request.method, call.request.url) for call in responses.calls
-    ] == BASE_API_CALLS * 2
-
-    assert [(level, message) for _, level, message in caplog.record_tuples] == [
-        (logging.INFO, "Loaded Worker Fake worker @ 123412 from API"),
-        (logging.INFO, "Modern configuration is not available"),
-        (
-            logging.WARNING,
-            "Failed running worker on Set (train) from Dataset (dataset_id): AssertionError('When processing a set, its dataset state should be Complete.')",
-        ),
-        (logging.ERROR, "Ran on 1 set: 0 completed, 1 failed"),
-    ]
-
-
 def test_run_download_dataset_artifact_api_error(
     mocker,
     tmp_path,
@@ -570,16 +542,18 @@ def test_run_no_downloaded_dataset_artifact_error(
     ]
 
 
+@pytest.mark.parametrize("dataset_state", DatasetState)
 def test_run(
     mocker,
     tmp_path,
     responses,
     caplog,
+    dataset_state,
     mock_dataset_worker,
     default_dataset,
     default_artifact,
 ):
-    default_dataset.state = DatasetState.Complete.value
+    default_dataset.state = dataset_state.value
     mocker.patch(
         "arkindex_worker.worker.DatasetWorker.list_sets",
         return_value=[Set(name="train", dataset=default_dataset)],
@@ -590,55 +564,68 @@ def test_run(
     )
     mock_process = mocker.patch("arkindex_worker.worker.DatasetWorker.process_set")
 
-    archive_path = (
-        FIXTURES_DIR
-        / "extract_parent_archives"
-        / "first_parent"
-        / "arkindex_data.tar.zst"
-    )
-    responses.add(
-        responses.GET,
-        f"http://testserver/api/v1/task/{default_dataset.task_id}/artifacts/",
-        status=200,
-        json=[default_artifact],
-    )
-    responses.add(
-        responses.GET,
-        f"http://testserver/api/v1/task/{default_dataset.task_id}/artifact/dataset_id.tar.zst",
-        status=200,
-        body=archive_path.read_bytes(),
-        content_type="application/zstd",
-    )
+    if dataset_state == DatasetState.Complete:
+        archive_path = (
+            FIXTURES_DIR
+            / "extract_parent_archives"
+            / "first_parent"
+            / "arkindex_data.tar.zst"
+        )
+        responses.add(
+            responses.GET,
+            f"http://testserver/api/v1/task/{default_dataset.task_id}/artifacts/",
+            status=200,
+            json=[default_artifact],
+        )
+        responses.add(
+            responses.GET,
+            f"http://testserver/api/v1/task/{default_dataset.task_id}/artifact/dataset_id.tar.zst",
+            status=200,
+            body=archive_path.read_bytes(),
+            content_type="application/zstd",
+        )
 
     mock_dataset_worker.run()
 
     assert mock_process.call_count == 1
 
-    assert len(responses.calls) == len(BASE_API_CALLS) * 2 + 2
+    # We only download the dataset archive when it is Complete
+    extra_calls = []
+    if dataset_state == DatasetState.Complete:
+        extra_calls = [
+            (
+                "GET",
+                f"http://testserver/api/v1/task/{default_dataset.task_id}/artifacts/",
+            ),
+            (
+                "GET",
+                f"http://testserver/api/v1/task/{default_dataset.task_id}/artifact/dataset_id.tar.zst",
+            ),
+        ]
+
+    assert len(responses.calls) == len(BASE_API_CALLS) * 2 + len(extra_calls)
     assert [
         (call.request.method, call.request.url) for call in responses.calls
-    ] == BASE_API_CALLS * 2 + [
-        (
-            "GET",
-            f"http://testserver/api/v1/task/{default_dataset.task_id}/artifacts/",
-        ),
-        (
-            "GET",
-            f"http://testserver/api/v1/task/{default_dataset.task_id}/artifact/dataset_id.tar.zst",
-        ),
-    ]
+    ] == BASE_API_CALLS * 2 + extra_calls
 
-    assert [(level, message) for _, level, message in caplog.record_tuples] == [
+    logs = [
         (logging.INFO, "Loaded Worker Fake worker @ 123412 from API"),
         (logging.INFO, "Modern configuration is not available"),
         (
-            logging.INFO,
-            "Retrieving data for Set (train) from Dataset (dataset_id) (1/1)",
+            logging.WARNING,
+            f"The dataset Dataset (dataset_id) has its state set to `{dataset_state.value}`, its archive will not be downloaded",
         ),
-        (logging.INFO, "Downloading artifact for Dataset (dataset_id)"),
         (logging.INFO, "Processing Set (train) from Dataset (dataset_id) (1/1)"),
         (logging.INFO, "Ran on 1 set: 1 completed, 0 failed"),
     ]
+    if dataset_state == DatasetState.Complete:
+        logs[2] = (
+            logging.INFO,
+            "Retrieving data for Set (train) from Dataset (dataset_id) (1/1)",
+        )
+        logs.insert(3, (logging.INFO, "Downloading artifact for Dataset (dataset_id)"))
+
+    assert [(level, message) for _, level, message in caplog.record_tuples] == logs
 
 
 def test_run_read_only(

tests/test_elements_worker/test_task.py

@@ -1,6 +1,9 @@
+import tempfile
 import uuid
+from pathlib import Path
 
 import pytest
+from requests import HTTPError
 
 from arkindex.exceptions import ErrorResponse
 from arkindex_worker.models import Artifact
@@ -196,3 +199,112 @@ def test_download_artifact(
     ] == BASE_API_CALLS + [
         ("GET", f"http://testserver/api/v1/task/{TASK_ID}/artifact/dataset_id.tar.zst"),
     ]
+
+
+@pytest.mark.parametrize(
+    ("payload", "error"),
+    [
+        # Path
+        (
+            {"path": None},
+            "path shouldn't be null, should be a Path and should exist",
+        ),
+        (
+            {"path": "not path type"},
+            "path shouldn't be null, should be a Path and should exist",
+        ),
+        (
+            {"path": Path("i_do_no_exist.oops")},
+            "path shouldn't be null, should be a Path and should exist",
+        ),
+    ],
+)
+def test_upload_artifact_wrong_param_path(mock_dataset_worker, payload, error):
+    with pytest.raises(AssertionError, match=error):
+        mock_dataset_worker.upload_artifact(**payload)
+
+
+@pytest.fixture
+def tmp_file(mock_dataset_worker):
+    with tempfile.NamedTemporaryFile(
+        mode="w", suffix=".txt", dir=mock_dataset_worker.work_dir
+    ) as file:
+        file.write("Some content...")
+        file.seek(0)
+
+        yield Path(file.name)
+
+
+def test_upload_artifact_api_error(responses, mock_dataset_worker, tmp_file):
+    responses.add(
+        responses.POST,
+        "http://testserver/api/v1/task/my_task/artifacts/",
+        status=418,
+    )
+
+    with pytest.raises(ErrorResponse):
+        mock_dataset_worker.upload_artifact(path=tmp_file)
+
+    assert len(responses.calls) == len(BASE_API_CALLS) + 1
+    assert [
+        (call.request.method, call.request.url) for call in responses.calls
+    ] == BASE_API_CALLS + [("POST", "http://testserver/api/v1/task/my_task/artifacts/")]
+
+
+def test_upload_artifact_s3_upload_error(
+    responses,
+    mock_dataset_worker,
+    tmp_file,
+):
+    responses.add(
+        responses.POST,
+        "http://testserver/api/v1/task/my_task/artifacts/",
+        json={
+            "id": "11111111-1111-1111-1111-111111111111",
+            "path": tmp_file.name,
+            "size": 15,
+            "content_type": "text/plain",
+            "s3_put_url": "http://example.com/oops.txt",
+        },
+    )
+    responses.add(responses.PUT, "http://example.com/oops.txt", status=500)
+
+    with pytest.raises(HTTPError):
+        mock_dataset_worker.upload_artifact(path=tmp_file)
+
+    assert len(responses.calls) == len(BASE_API_CALLS) + 2
+    assert [
+        (call.request.method, call.request.url) for call in responses.calls
+    ] == BASE_API_CALLS + [
+        ("POST", "http://testserver/api/v1/task/my_task/artifacts/"),
+        ("PUT", "http://example.com/oops.txt"),
+    ]
+
+
+def test_upload_artifact(
+    responses,
+    mock_dataset_worker,
+    tmp_file,
+):
+    responses.add(
+        responses.POST,
+        "http://testserver/api/v1/task/my_task/artifacts/",
+        json={
+            "id": "11111111-1111-1111-1111-111111111111",
+            "path": tmp_file.name,
+            "size": 15,
+            "content_type": "text/plain",
+            "s3_put_url": "http://example.com/test.txt",
+        },
+    )
+    responses.add(responses.PUT, "http://example.com/test.txt")
+
+    mock_dataset_worker.upload_artifact(path=tmp_file)
+
+    assert len(responses.calls) == len(BASE_API_CALLS) + 2
+    assert [
+        (call.request.method, call.request.url) for call in responses.calls
+    ] == BASE_API_CALLS + [
+        ("POST", "http://testserver/api/v1/task/my_task/artifacts/"),
+        ("PUT", "http://example.com/test.txt"),
+    ]