D lang lister

Add a dlang module that retrieve origins from an http api endpoint.
Each origin is a git based project url on github.com, gitlab.com or
bitbucket.com.

setup.py

@@ -65,6 +65,7 @@ setup(
         lister.cran=swh.lister.cran:register
         lister.crates=swh.lister.crates:register
         lister.debian=swh.lister.debian:register
+        lister.dlang=swh.lister.dlang:register
         lister.fedora=swh.lister.fedora:register
         lister.gitea=swh.lister.gitea:register
         lister.github=swh.lister.github:register

swh/lister/dlang/__init__.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+
+"""
+Dlang lister
+=============
+
+D is a general-purpose programming language with static typing, systems-level access,
+and C-like syntax.
+
+The `Dlang`_ lister list origins from its packages manager registry `DUB`_.
+
+The registry provides an `http api endpoint`_ that helps in getting the packages index
+with name,  url, versions and dates.
+
+As of July 2023 `DUB`_ list 2364 package names.
+
+Origins retrieving strategy
+---------------------------
+
+To build a list of origins we make a GET request to an `http api endpoint`_ that returns
+a Json array of objects.
+The origin url for each package is constructed with the information of corresponding
+`repository` entry which represents Git based projects hosted on Github, GitLab or
+Bitbucket.
+
+Page listing
+------------
+
+There is only one page listing all origins url.
+
+Origins from page
+-----------------
+
+The lister is stateless and yields all origins url from one page. It is a list of package
+url with last update information.
+
+Running tests
+-------------
+
+Activate the virtualenv and run from within swh-lister directory::
+
+   pytest -s -vv --log-cli-level=DEBUG swh/lister/dlang/tests
+
+Testing with Docker
+-------------------
+
+Change directory to swh/docker then launch the docker environment::
+
+   docker compose up -d
+
+Then schedule a dlang listing task::
+
+   docker compose exec swh-scheduler swh scheduler task add -p oneshot list-dlang
+
+You can follow lister execution by displaying logs of swh-lister service::
+
+   docker compose logs -f swh-lister
+
+.. _Dlang: https://dlang.org/
+.. _DUB: https://code.dlang.org/
+.. _http api endpoint: https://code.dlang.org/api/packages/dump"
+"""
+
+
+def register():
+    from .lister import DlangLister
+
+    return {
+        "lister": DlangLister,
+        "task_modules": ["%s.tasks" % __name__],
+    }

swh/lister/dlang/lister.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import logging
+from typing import Any, Dict, Iterator, List, Optional
+
+import iso8601
+
+from swh.scheduler.interface import SchedulerInterface
+from swh.scheduler.model import ListedOrigin
+
+from ..pattern import CredentialsType, StatelessLister
+
+logger = logging.getLogger(__name__)
+
+# Aliasing the page results returned by `get_pages` method from the lister.
+DlangListerPage = List[Dict[str, Any]]
+
+
+class DlangLister(StatelessLister[DlangListerPage]):
+    """List D lang origins."""
+
+    LISTER_NAME = "dlang"
+    VISIT_TYPE = "git"  # D lang origins url are Git repositories
+    INSTANCE = "dlang"
+
+    BASE_URL = "https://code.dlang.org"
+    PACKAGES_DUMP_URL = BASE_URL + "/api/packages/dump"
+    KINDS = {
+        "github": "https://github.com",
+        "gitlab": "https://gitlab.com",
+        "bitbucket": "https://bitbucket.com",
+    }
+
+    KIND_URL_PATTERN = "{url}/{owner}/{project}"
+
+    def __init__(
+        self,
+        scheduler: SchedulerInterface,
+        credentials: Optional[CredentialsType] = None,
+        max_origins_per_page: Optional[int] = None,
+        max_pages: Optional[int] = None,
+        enable_origins: bool = True,
+    ):
+        super().__init__(
+            scheduler=scheduler,
+            credentials=credentials,
+            instance=self.INSTANCE,
+            url=self.PACKAGES_DUMP_URL,
+            max_origins_per_page=max_origins_per_page,
+            max_pages=max_pages,
+            enable_origins=enable_origins,
+        )
+        self.session.headers.update({"Accept": "application/json"})
+
+    def get_pages(self) -> Iterator[DlangListerPage]:
+        """Yield an iterator which returns 'page'
+
+        It uses the api endpoint provided by `https://registry.dlang.io/packages`
+        to get a list of package names with an origin url that corresponds to Git
+        repository.
+
+        There is only one page that list all origins urls.
+        """
+        response = self.http_request(self.url)
+        yield response.json()
+
+    def get_origins_from_page(self, page: DlangListerPage) -> Iterator[ListedOrigin]:
+        """Iterate on all pages and yield ListedOrigin instances"""
+        assert self.lister_obj.id is not None
+
+        for entry in page:
+            repo: Dict[str, Any] = entry["repository"]
+            kind: str = repo["kind"]
+
+            if kind not in self.KINDS:
+                logging.error("Can not build a repository url with %r" % repo)
+                continue
+
+            repo_url = self.KIND_URL_PATTERN.format(
+                url=self.KINDS[kind], owner=repo["owner"], project=repo["project"]
+            )
+
+            last_update = iso8601.parse_date(entry["stats"]["updatedAt"])
+
+            yield ListedOrigin(
+                lister_id=self.lister_obj.id,
+                visit_type=self.VISIT_TYPE,
+                url=repo_url,
+                last_update=last_update,
+            )

swh/lister/dlang/tasks.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+from celery import shared_task
+
+from swh.lister.dlang.lister import DlangLister
+
+
+@shared_task(name=__name__ + ".DlangListerTask")
+def list_dlang(**lister_args):
+    """Lister task for D lang packages registry"""
+    return DlangLister.from_configfile(**lister_args).run().dict()
+
+
+@shared_task(name=__name__ + ".ping")
+def _ping():
+    return "OK"

swh/lister/dlang/tests/test_lister.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+import iso8601
+
+from swh.lister.dlang.lister import DlangLister
+
+expected_origins = [
+    {
+        "url": "https://github.com/katyukha/TheProcess",
+        "last_update": "2023-07-12T14:42:46.231Z",
+    },
+    {
+        "url": "https://gitlab.com/AntonMeep/silly",
+        "last_update": "2023-07-12T01:32:31.235Z",
+    },
+]
+
+
+def test_dlang_lister(datadir, requests_mock_datadir, swh_scheduler):
+    lister = DlangLister(scheduler=swh_scheduler)
+    res = lister.run()
+
+    assert res.pages == 1
+    assert res.origins == 1 + 1
+
+    scheduler_origins = swh_scheduler.get_listed_origins(lister.lister_obj.id).results
+
+    assert len(scheduler_origins) == len(expected_origins)
+    assert {
+        (
+            scheduled.visit_type,
+            scheduled.url,
+            scheduled.last_update,
+        )
+        for scheduled in scheduler_origins
+    } == {
+        ("git", expected["url"], iso8601.parse_date(expected["last_update"]))
+        for expected in expected_origins
+    }

swh/lister/dlang/tests/test_tasks.py

+# Copyright (C) 2023  The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+from swh.lister.pattern import ListerStats
+
+
+def test_dlang_ping(swh_scheduler_celery_app, swh_scheduler_celery_worker):
+    res = swh_scheduler_celery_app.send_task("swh.lister.dlang.tasks.ping")
+    assert res
+    res.wait()
+    assert res.successful()
+    assert res.result == "OK"
+
+
+def test_dlang_lister(swh_scheduler_celery_app, swh_scheduler_celery_worker, mocker):
+    # setup the mocked DlangLister
+    lister = mocker.patch("swh.lister.dlang.tasks.DlangLister")
+    lister.from_configfile.return_value = lister
+    stats = ListerStats(pages=42, origins=42)
+    lister.run.return_value = stats
+
+    res = swh_scheduler_celery_app.send_task("swh.lister.dlang.tasks.DlangListerTask")
+    assert res
+    res.wait()
+    assert res.successful()
+    assert res.result == stats.dict()
+
+    lister.from_configfile.assert_called_once_with()
+    lister.run.assert_called_once_with()