Merge branch 'master' into 'bump-python-gitlab'

# Conflicts:
#   requirements.txt

VERSION

-0.2.1-beta2
+0.2.2-beta2

arkindex_worker/cache.py

@@ -156,6 +156,7 @@ class CachedTranscriptionEntity(Model):
     entity = ForeignKeyField(CachedEntity, backref="transcription_entities")
     offset = IntegerField(constraints=[Check("offset >= 0")])
     length = IntegerField(constraints=[Check("length > 0")])
+    worker_version_id = UUIDField()
 
     class Meta:
         primary_key = CompositeKey("transcription", "entity")

arkindex_worker/worker/base.py

@@ -62,6 +62,7 @@ class BaseWorker(object):
         logger.info(f"Worker will use {self.work_dir} as working directory")
 
         self.process_information = None
+        self.user_configuration = None
         self.support_cache = support_cache
         # use_cache will be updated in configure() if the cache is supported and if there
         # is at least one available sqlite database either given or in the parent tasks
@@ -160,6 +161,15 @@ class BaseWorker(object):
         # Load all required secrets
         self.secrets = {name: self.load_secret(name) for name in required_secrets}
 
+        # Load worker run configuration when available and not in dev mode
+        if os.environ.get("ARKINDEX_WORKER_RUN_ID") and not self.args.dev:
+            worker_run = self.request(
+                "RetrieveWorkerRun", id=os.environ["ARKINDEX_WORKER_RUN_ID"]
+            )
+            self.user_configuration = worker_run.get("configuration")
+            if self.user_configuration:
+                logger.info("Loaded user configuration from WorkerRun")
+
         task_id = os.environ.get("PONOS_TASK")
         paths = None
         if self.support_cache and self.args.database is not None:

arkindex_worker/worker/element.py

@@ -8,7 +8,34 @@ from arkindex_worker.cache import CachedElement, CachedImage
 from arkindex_worker.models import Element
 
 
+class MissingTypeError(Exception):
+    """
+    A required element type was not found in a corpus.
+    """
+
+
 class ElementMixin(object):
+    def check_required_types(self, corpus_id: str, *type_slugs: str) -> bool:
+        """
+        Check that a corpus has a list of required element types,
+        and raise an exception if any of them are missing.
+        """
+        assert len(type_slugs), "At least one element type slug is required."
+        assert all(
+            isinstance(slug, str) for slug in type_slugs
+        ), "Element type slugs must be strings."
+
+        corpus = self.request("RetrieveCorpus", id=corpus_id)
+        available_slugs = {element_type["slug"] for element_type in corpus["types"]}
+        missing_slugs = set(type_slugs) - available_slugs
+
+        if missing_slugs:
+            raise MissingTypeError(
+                f'Element type(s) {", ".join(sorted(missing_slugs))} were not found in the {corpus["name"]} corpus ({corpus["id"]}).'
+            )
+
+        return True
+
     def create_sub_element(self, element, type, name, polygon):
         """
         Create a child element on the given element through API

arkindex_worker/worker/entity.py

@@ -21,7 +21,7 @@ class EntityType(Enum):
 
 class EntityMixin(object):
     def create_entity(
-        self, element, name, type, corpus=None, metas=None, validated=None
+        self, element, name, type, corpus=None, metas=dict(), validated=None
     ):
         """
         Create an entity on the given corpus through API
@@ -111,6 +111,7 @@ class EntityMixin(object):
                 "entity": entity,
                 "length": length,
                 "offset": offset,
+                "worker_version_id": self.worker_version_id,
             },
         )
         # TODO: Report transcription entity creation
@@ -118,15 +119,13 @@ class EntityMixin(object):
         if self.use_cache:
             # Store transcription entity in local cache
             try:
-                to_insert = [
-                    {
-                        "transcription": transcription,
-                        "entity": entity,
-                        "offset": offset,
-                        "length": length,
-                    }
-                ]
-                CachedTranscriptionEntity.insert_many(to_insert).execute()
+                CachedTranscriptionEntity.create(
+                    transcription=transcription,
+                    entity=entity,
+                    offset=offset,
+                    length=length,
+                    worker_version_id=self.worker_version_id,
+                )
             except IntegrityError as e:
                 logger.warning(
                     f"Couldn't save created transcription entity in local cache: {e}"

requirements.txt

 arkindex-client==1.0.6
 peewee==3.14.4
-Pillow==8.2.0
+Pillow==8.3.1
 python-gitlab==2.9.0
 python-gnupg==0.4.7
 sh==1.14.2

tests/test_base_worker.py

@@ -119,18 +119,56 @@ def test_cli_arg_verbose_given(mocker, mock_config_api):
     logger.setLevel(logging.NOTSET)
 
 
-def test_configure_dev_mode(mocker, mock_user_api, mock_worker_version_api):
+def test_configure_dev_mode(
+    mocker, monkeypatch, mock_user_api, mock_worker_version_api
+):
     """
     Configuring a worker in developer mode avoid retrieving process information
     """
     worker = BaseWorker()
     mocker.patch.object(sys, "argv", ["worker", "--dev"])
+    monkeypatch.setenv(
+        "ARKINDEX_WORKER_RUN_ID", "aaaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"
+    )
     worker.configure()
 
     assert worker.args.dev is True
     assert worker.process_information is None
     assert worker.worker_version_id == "12341234-1234-1234-1234-123412341234"
     assert worker.is_read_only is True
+    assert worker.user_configuration is None
+
+
+def test_configure_worker_run(mocker, monkeypatch, responses, mock_config_api):
+    worker = BaseWorker()
+    mocker.patch.object(sys, "argv", ["worker"])
+    run_id = "aaaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"
+    monkeypatch.setenv("ARKINDEX_WORKER_RUN_ID", run_id)
+    responses.add(
+        responses.GET,
+        f"http://testserver/api/v1/imports/workers/{run_id}/",
+        json={"id": run_id, "configuration": {"a": "b"}},
+    )
+    worker.configure()
+
+    assert worker.user_configuration == {"a": "b"}
+
+
+def test_configure_worker_run_missing_conf(
+    mocker, monkeypatch, responses, mock_config_api
+):
+    worker = BaseWorker()
+    mocker.patch.object(sys, "argv", ["worker"])
+    run_id = "aaaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa"
+    monkeypatch.setenv("ARKINDEX_WORKER_RUN_ID", run_id)
+    responses.add(
+        responses.GET,
+        f"http://testserver/api/v1/imports/workers/{run_id}/",
+        json={"id": run_id},
+    )
+    worker.configure()
+
+    assert worker.user_configuration is None
 
 
 def test_load_missing_secret():

tests/test_cache.py

@@ -58,7 +58,7 @@ def test_create_tables(tmp_path):
 CREATE TABLE "elements" ("id" TEXT NOT NULL PRIMARY KEY, "parent_id" TEXT, "type" VARCHAR(50) NOT NULL, "image_id" TEXT, "polygon" text, "initial" INTEGER NOT NULL, "worker_version_id" TEXT, FOREIGN KEY ("image_id") REFERENCES "images" ("id"))
 CREATE TABLE "entities" ("id" TEXT NOT NULL PRIMARY KEY, "type" VARCHAR(50) NOT NULL, "name" TEXT NOT NULL, "validated" INTEGER NOT NULL, "metas" text, "worker_version_id" TEXT NOT NULL)
 CREATE TABLE "images" ("id" TEXT NOT NULL PRIMARY KEY, "width" INTEGER NOT NULL, "height" INTEGER NOT NULL, "url" TEXT NOT NULL)
-CREATE TABLE "transcription_entities" ("transcription_id" TEXT NOT NULL, "entity_id" TEXT NOT NULL, "offset" INTEGER NOT NULL CHECK (offset >= 0), "length" INTEGER NOT NULL CHECK (length > 0), PRIMARY KEY ("transcription_id", "entity_id"), FOREIGN KEY ("transcription_id") REFERENCES "transcriptions" ("id"), FOREIGN KEY ("entity_id") REFERENCES "entities" ("id"))
+CREATE TABLE "transcription_entities" ("transcription_id" TEXT NOT NULL, "entity_id" TEXT NOT NULL, "offset" INTEGER NOT NULL CHECK (offset >= 0), "length" INTEGER NOT NULL CHECK (length > 0), "worker_version_id" TEXT NOT NULL, PRIMARY KEY ("transcription_id", "entity_id"), FOREIGN KEY ("transcription_id") REFERENCES "transcriptions" ("id"), FOREIGN KEY ("entity_id") REFERENCES "entities" ("id"))
 CREATE TABLE "transcriptions" ("id" TEXT NOT NULL PRIMARY KEY, "element_id" TEXT NOT NULL, "text" TEXT NOT NULL, "confidence" REAL NOT NULL, "worker_version_id" TEXT NOT NULL, FOREIGN KEY ("element_id") REFERENCES "elements" ("id"))"""
 
     actual_schema = "\n".join(

tests/test_elements_worker/test_elements.py

 # -*- coding: utf-8 -*-
 import json
-import os
-import tempfile
 from argparse import Namespace
 from uuid import UUID
 
@@ -11,58 +9,102 @@ from apistar.exceptions import ErrorResponse
 from arkindex_worker.cache import CachedElement, CachedImage
 from arkindex_worker.models import Element
 from arkindex_worker.worker import ElementsWorker
+from arkindex_worker.worker.element import MissingTypeError
 
 from . import BASE_API_CALLS
 
 
-def test_list_elements_elements_list_arg_wrong_type(monkeypatch, mock_elements_worker):
-    _, path = tempfile.mkstemp()
-    with open(path, "w") as f:
-        json.dump({}, f)
+def test_check_required_types_argument_types(mock_elements_worker):
+    corpus_id = "12341234-1234-1234-1234-123412341234"
+    worker = ElementsWorker()
+
+    with pytest.raises(AssertionError) as e:
+        worker.check_required_types(corpus_id)
+    assert str(e.value) == "At least one element type slug is required."
+
+    with pytest.raises(AssertionError) as e:
+        worker.check_required_types(corpus_id, "lol", 42)
+    assert str(e.value) == "Element type slugs must be strings."
+
 
-    monkeypatch.setenv("TASK_ELEMENTS", path)
+def test_check_required_types(monkeypatch, tmp_path, mock_elements_worker, responses):
+    elements_path = tmp_path / "elements.json"
+    elements_path.write_text("[]")
+    monkeypatch.setenv("TASK_ELEMENTS", str(elements_path))
+
+    corpus_id = "12341234-1234-1234-1234-123412341234"
+    responses.add(
+        responses.GET,
+        f"http://testserver/api/v1/corpus/{corpus_id}/",
+        json={
+            "id": corpus_id,
+            "name": "Some Corpus",
+            "types": [{"slug": "folder"}, {"slug": "page"}],
+        },
+    )
+    worker = ElementsWorker()
+    worker.configure()
+
+    assert worker.check_required_types(corpus_id, "page")
+    assert worker.check_required_types(corpus_id, "page", "folder")
+
+    with pytest.raises(MissingTypeError) as e:
+        assert worker.check_required_types(corpus_id, "page", "text_line", "act")
+    assert (
+        str(e.value)
+        == "Element type(s) act, text_line were not found in the Some Corpus corpus (12341234-1234-1234-1234-123412341234)."
+    )
+
+
+def test_list_elements_elements_list_arg_wrong_type(
+    monkeypatch, tmp_path, mock_elements_worker
+):
+    elements_path = tmp_path / "elements.json"
+    elements_path.write_text("{}")
+
+    monkeypatch.setenv("TASK_ELEMENTS", str(elements_path))
     worker = ElementsWorker()
     worker.configure()
-    os.unlink(path)
 
     with pytest.raises(AssertionError) as e:
         worker.list_elements()
     assert str(e.value) == "Elements list must be a list"
 
 
-def test_list_elements_elements_list_arg_empty_list(monkeypatch, mock_elements_worker):
-    _, path = tempfile.mkstemp()
-    with open(path, "w") as f:
-        json.dump([], f)
+def test_list_elements_elements_list_arg_empty_list(
+    monkeypatch, tmp_path, mock_elements_worker
+):
+    elements_path = tmp_path / "elements.json"
+    elements_path.write_text("[]")
 
-    monkeypatch.setenv("TASK_ELEMENTS", path)
+    monkeypatch.setenv("TASK_ELEMENTS", str(elements_path))
     worker = ElementsWorker()
     worker.configure()
-    os.unlink(path)
 
     with pytest.raises(AssertionError) as e:
         worker.list_elements()
     assert str(e.value) == "No elements in elements list"
 
 
-def test_list_elements_elements_list_arg_missing_id(monkeypatch, mock_elements_worker):
-    _, path = tempfile.mkstemp()
-    with open(path, "w") as f:
+def test_list_elements_elements_list_arg_missing_id(
+    monkeypatch, tmp_path, mock_elements_worker
+):
+    elements_path = tmp_path / "elements.json"
+    with elements_path.open("w") as f:
         json.dump([{"type": "volume"}], f)
 
-    monkeypatch.setenv("TASK_ELEMENTS", path)
+    monkeypatch.setenv("TASK_ELEMENTS", str(elements_path))
     worker = ElementsWorker()
     worker.configure()
-    os.unlink(path)
 
     elt_list = worker.list_elements()
 
     assert elt_list == []
 
 
-def test_list_elements_elements_list_arg(monkeypatch, mock_elements_worker):
-    _, path = tempfile.mkstemp()
-    with open(path, "w") as f:
+def test_list_elements_elements_list_arg(monkeypatch, tmp_path, mock_elements_worker):
+    elements_path = tmp_path / "elements.json"
+    with elements_path.open("w") as f:
         json.dump(
             [
                 {"id": "volumeid", "type": "volume"},
@@ -73,10 +115,9 @@ def test_list_elements_elements_list_arg(monkeypatch, mock_elements_worker):
             f,
         )
 
-    monkeypatch.setenv("TASK_ELEMENTS", path)
+    monkeypatch.setenv("TASK_ELEMENTS", str(elements_path))
     worker = ElementsWorker()
     worker.configure()
-    os.unlink(path)
 
     elt_list = worker.list_elements()
 
@@ -103,9 +144,9 @@ def test_list_elements_element_arg(mocker, mock_elements_worker):
     assert elt_list == ["volumeid", "pageid"]
 
 
-def test_list_elements_both_args_error(mocker, mock_elements_worker):
-    _, path = tempfile.mkstemp()
-    with open(path, "w") as f:
+def test_list_elements_both_args_error(mocker, mock_elements_worker, tmp_path):
+    elements_path = tmp_path / "elements.json"
+    with elements_path.open("w") as f:
         json.dump(
             [
                 {"id": "volumeid", "type": "volume"},
@@ -120,7 +161,7 @@ def test_list_elements_both_args_error(mocker, mock_elements_worker):
         return_value=Namespace(
             element=["anotherid", "againanotherid"],
             verbose=False,
-            elements_list=open(path),
+            elements_list=elements_path.open(),
             database=None,
             dev=False,
         ),
@@ -128,7 +169,6 @@ def test_list_elements_both_args_error(mocker, mock_elements_worker):
 
     worker = ElementsWorker()
     worker.configure()
-    os.unlink(path)
 
     with pytest.raises(AssertionError) as e:
         worker.list_elements()
@@ -847,7 +887,7 @@ def test_create_elements(responses, mock_elements_worker_with_cache, tmp_path):
     assert created_ids == [{"id": "497f6eca-6276-4993-bfeb-53cbbbba6f08"}]
 
     # Check that created elements were properly stored in SQLite cache
-    assert os.path.isfile(tmp_path / "db.sqlite")
+    assert (tmp_path / "db.sqlite").is_file()
 
     assert list(CachedElement.select()) == [
         CachedElement(

tests/test_elements_worker/test_entities.py

@@ -213,7 +213,7 @@ def test_create_entity(responses, mock_elements_worker):
     assert json.loads(responses.calls[-1].request.body) == {
         "name": "Bob Bob",
         "type": "person",
-        "metas": None,
+        "metas": {},
         "validated": None,
         "corpus": "12341234-1234-1234-1234-123412341234",
         "worker_version": "12341234-1234-1234-1234-123412341234",
@@ -247,7 +247,7 @@ def test_create_entity_with_cache(responses, mock_elements_worker_with_cache):
     assert json.loads(responses.calls[-1].request.body) == {
         "name": "Bob Bob",
         "type": "person",
-        "metas": None,
+        "metas": {},
         "validated": None,
         "corpus": "12341234-1234-1234-1234-123412341234",
         "worker_version": "12341234-1234-1234-1234-123412341234",
@@ -261,7 +261,7 @@ def test_create_entity_with_cache(responses, mock_elements_worker_with_cache):
             type="person",
             name="Bob Bob",
             validated=False,
-            metas=None,
+            metas={},
             worker_version_id=UUID("12341234-1234-1234-1234-123412341234"),
         )
     ]
@@ -449,6 +449,7 @@ def test_create_transcription_entity(responses, mock_elements_worker):
         "entity": "11111111-1111-1111-1111-111111111111",
         "offset": 5,
         "length": 10,
+        "worker_version_id": "12341234-1234-1234-1234-123412341234",
     }
 
 
@@ -504,6 +505,7 @@ def test_create_transcription_entity_with_cache(
         "entity": "11111111-1111-1111-1111-111111111111",
         "offset": 5,
         "length": 10,
+        "worker_version_id": "12341234-1234-1234-1234-123412341234",
     }
 
     # Check that created transcription entity was properly stored in SQLite cache
@@ -513,5 +515,6 @@ def test_create_transcription_entity_with_cache(
             entity=UUID("11111111-1111-1111-1111-111111111111"),
             offset=5,
             length=10,
+            worker_version_id=UUID("12341234-1234-1234-1234-123412341234"),
         )
     ]

worker-{{cookiecutter.slug}}/requirements.txt

-arkindex-base-worker==0.2.0
+arkindex-base-worker==0.2.1