svn.loader: Mark visit status as not_found when relevant

When failing to communicate with the svn server at initialization time, a visit status
ends up in "not_found" status.

Other failures to communicate ends up in a visit status marked as "failed".

Related to T3030

swh/loader/svn/loader.py

-# Copyright (C) 2015-2020  The Software Heritage developers
+# Copyright (C) 2015-2021  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
@@ -16,9 +16,13 @@ from subprocess import Popen
 import tempfile
 from typing import Any, Dict, Iterator, List, Optional, Tuple
 
+from subvertpy import SubversionException
+
 from swh.core.config import merge_configs
 from swh.loader.core.loader import BaseLoader
 from swh.loader.core.utils import clean_dangling_folders
+from swh.loader.exception import NotFound
+from swh.loader.svn.svn import SvnRepo
 from swh.model import from_disk, hashutil
 from swh.model.model import (
     Content,
@@ -32,7 +36,7 @@ from swh.model.model import (
 )
 from swh.storage.algos.snapshot import snapshot_get_latest
 
-from . import converters, svn
+from . import converters
 from .exception import SvnLoaderHistoryAltered, SvnLoaderUneventful
 from .utils import (
     OutputStream,
@@ -398,9 +402,20 @@ Local repository not cleaned up for investigation: %s"""
                 dir=self.temp_directory,
             )
 
-        self.svnrepo = svn.SvnRepo(
-            self.svn_url, self.origin_url, local_dirname, self.max_content_length
-        )
+        try:
+            self.svnrepo = SvnRepo(
+                self.svn_url, self.origin_url, local_dirname, self.max_content_length
+            )
+        except SubversionException as e:
+            error_msgs = [
+                "Unable to connect to a repository at URL",
+                "Unknown URL type",
+            ]
+            for msg in error_msgs:
+                if msg in e.args[0]:
+                    self._load_status = "uneventful"
+                    raise NotFound(e)
+            raise
 
         try:
             revision_start, revision_end, revision_parents = self.start_from(

swh/loader/svn/tests/test_loader.py

@@ -5,6 +5,9 @@
 
 import os
 
+import pytest
+from subvertpy import SubversionException
+
 from swh.loader.svn.loader import SvnLoader, SvnLoaderFromRemoteDump
 from swh.loader.tests import (
     assert_last_visit_matches,
@@ -36,6 +39,59 @@ GOURMET_UPDATES_SNAPSHOT = Snapshot(
 )
 
 
+def test_loader_svn_not_found_no_mock(swh_config, tmp_path):
+    """Given an unknown repository, the loader visit ends up in status not_found"""
+    unknown_repo_url = "unknown-repository"
+    loader = SvnLoader(unknown_repo_url, destination_path=tmp_path)
+
+    assert loader.load() == {"status": "uneventful"}
+
+    assert_last_visit_matches(
+        loader.storage, unknown_repo_url, status="not_found", type="svn",
+    )
+
+
+@pytest.mark.parametrize(
+    "exception_msg", ["Unable to connect to a repository at URL", "Unknown URL type",]
+)
+def test_loader_svn_not_found(swh_config, tmp_path, exception_msg, mocker):
+    """Given unknown repository issues, the loader visit ends up in status not_found"""
+    mock = mocker.patch("swh.loader.svn.loader.SvnRepo")
+    mock.side_effect = SubversionException(exception_msg, 0)
+
+    unknown_repo_url = "unknown-repository"
+    loader = SvnLoader(unknown_repo_url, destination_path=tmp_path)
+
+    assert loader.load() == {"status": "uneventful"}
+
+    assert_last_visit_matches(
+        loader.storage, unknown_repo_url, status="not_found", type="svn",
+    )
+
+
+@pytest.mark.parametrize(
+    "exception",
+    [
+        SubversionException("Irrelevant message, considered a failure", 10),
+        SubversionException("Present but fails to read, considered a failure", 20),
+        ValueError("considered a failure"),
+    ],
+)
+def test_loader_svn_failures(swh_config, tmp_path, exception, mocker):
+    """Given any errors raised, the loader visit ends up in status failed"""
+    mock = mocker.patch("swh.loader.svn.loader.SvnRepo")
+    mock.side_effect = exception
+
+    existing_repo_url = "existing-repo-url"
+    loader = SvnLoader(existing_repo_url, destination_path=tmp_path)
+
+    assert loader.load() == {"status": "failed"}
+
+    assert_last_visit_matches(
+        loader.storage, existing_repo_url, status="failed", type="svn",
+    )
+
+
 def test_loader_svn_new_visit(swh_config, datadir, tmp_path):
     """Eventful visit should yield 1 snapshot"""
     archive_name = "pkg-gourmet"

[Phabricator Migration user]

Some references in the commit message have been migrated:

• T3030 is now swh-loader-core#3030