.arkindex.yml

@@ -4,9 +4,9 @@ version: 2
 workers:
   - slug: thumbnails-generator
     name: Thumbnails Generator
-    type: generator
+    type: Utils
     docker:
-      build: Dockerfile
+      command: worker-thumbnails-generator
     configuration:
       width: 900
       height: 400

.gitlab-ci.yml

@@ -14,7 +14,7 @@ variables:
   DEBIAN_FRONTEND: non-interactive
 
 test:
-  image: python:slim
+  image: python:3.12-slim
 
   stage: test
   cache:
@@ -41,7 +41,7 @@ test:
     - tox -- --junitxml=test-report.xml --durations=50
 
 lint:
-  image: python:slim
+  image: python:3.12-slim
 
   cache:
     paths:
@@ -134,3 +134,4 @@ publish-worker:
         - demo
         # Publish worker on https://arkindex.teklia.com
         - prod
+        - elephant

.pre-commit-config.yaml

 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.3.1
+    rev: v0.8.2
     hooks:
       # Run the linter.
       - id: ruff
@@ -9,7 +9,7 @@ repos:
       # Run the formatter.
       - id: ruff-format
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0
+    rev: v5.0.0
     hooks:
       - id: check-ast
       - id: check-docstring-first
@@ -27,10 +27,16 @@ repos:
       - id: check-toml
       - id: requirements-txt-fixer
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.2.6
+    rev: v2.3.0
     hooks:
       - id: codespell
         args: ['--write-changes']
   - repo: meta
     hooks:
       - id: check-useless-excludes
+  - repo: https://gitlab.teklia.com/tools/pre-commit-hooks
+    rev: 0.1.0
+    hooks:
+      - id: long-test-files
+        args: ['1000']
+        files: '^tests\/(.*\/)?test_[^\/]*\.py$'
\ No newline at end of file

Dockerfile

@@ -11,8 +11,4 @@ COPY worker_thumbnails_generator worker_thumbnails_generator
 COPY requirements.txt setup.py pyproject.toml ./
 RUN pip install . --no-cache-dir
 
-# Add archi local CA
-RUN curl https://assets.teklia.com/teklia_dev_ca.pem > /usr/local/share/ca-certificates/arkindex-dev.crt && update-ca-certificates
-ENV REQUESTS_CA_BUNDLE /etc/ssl/certs/ca-certificates.crt
-
 CMD ["worker-thumbnails-generator"]

pyproject.toml

@@ -4,6 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "worker_thumbnails_generator"
+license = { file = "LICENSE" }
 version = "0.1.0"
 description = "Worker to generate thumbnails for Arkindex elements"
 dynamic = ["dependencies"]

requirements.txt

-arkindex-base-worker==0.3.7rc4
+arkindex-base-worker==0.4.0

tests/conftest.py

@@ -45,7 +45,7 @@ def _setup_environment(responses, monkeypatch) -> None:
     monkeypatch.setattr(BaseWorker, "setup_api_client", mock_setup_api_client)
 
 
-@pytest.fixture()
+@pytest.fixture
 def mock_worker() -> ThumbnailsGenerator:
     worker = ThumbnailsGenerator()
     worker.args = Namespace(dev=False)
@@ -63,7 +63,7 @@ def mock_worker() -> ThumbnailsGenerator:
     return worker
 
 
-@pytest.fixture()
+@pytest.fixture
 def page_1_payload() -> dict:
     return {
         "id": "page_1",
@@ -89,7 +89,7 @@ def page_1_payload() -> dict:
     }
 
 
-@pytest.fixture()
+@pytest.fixture
 def page_2_payload() -> dict:
     return {
         "id": "page_2",
@@ -115,7 +115,7 @@ def page_2_payload() -> dict:
     }
 
 
-@pytest.fixture()
+@pytest.fixture
 def page_3_payload() -> dict:
     return {
         "id": "page_3",
@@ -141,7 +141,7 @@ def page_3_payload() -> dict:
     }
 
 
-@pytest.fixture()
+@pytest.fixture
 def pages_payload(
     page_1_payload: dict, page_2_payload: dict, page_3_payload: dict
 ) -> list[dict]:

tests/test_worker.py

@@ -175,6 +175,13 @@ def test_process_element_no_images_found(caplog, mock_worker):
         folder=False,
         response=[],
     )
+    mock_worker.api_client.add_response(
+        "ListElementChildren",
+        id=folder.id,
+        folder=False,
+        recursive=True,
+        response=[],
+    )
 
     mock_worker.process_element(folder)
 

worker_thumbnails_generator/worker.py

@@ -14,32 +14,23 @@ class ThumbnailsGenerator(ElementsWorker):
     def configure(self) -> None:
         super().configure()
 
-        if self.user_configuration:
-            logger.info("Loading user_configuration")
-            self.config.update(self.user_configuration)
-
         self.thumbnail_width: int = self.config["width"]
         self.thumbnail_height: int = self.config["height"]
         self.first_n: int = self.config.get("first_n")
 
-        corpus = self.request("RetrieveCorpus", id=self.corpus_id)
-        self.corpus_types = {
-            element_type["slug"]: element_type for element_type in corpus["types"]
-        }
+        self.list_corpus_types()
 
     def _list_folder_elements(self, folder):
         """
         List all elements in a folder that could have images and could be used as thumbnails.
         """
-        # First try with the first page of results returned by a non-recursive list, much faster than recursive
-        results = self.list_element_children(folder, folder=False)
-        yield from results
+        # First try with a non-recursive list, much faster than recursive
+        yield from self.list_element_children(folder, folder=False)
 
         # Fallback to going through all of the non-folder elements recursively.
-        # If there were no results at all, we do not even try to make the other query,
-        # since we already know there will be no results on this one too.
-        if results:
-            yield from self.list_element_children(folder, folder=False, recursive=True)
+        # This request may return some elements even though the previous one did not,
+        # since this folder might only contain folder elements that have non-folder elements.
+        yield from self.list_element_children(folder, folder=False, recursive=True)
 
     def get_first_images(self, folder):
         """
@@ -129,7 +120,7 @@ class ThumbnailsGenerator(ElementsWorker):
 
 def main() -> None:
     ThumbnailsGenerator(
-        description="Worker to generate thumbnails for Arkindex elements"
+        description="Worker to generate thumbnails for Arkindex elements: Utility process"
     ).run()