Fix tests producing artifacts during testing.

test/functional/conftest.py

@@ -51,12 +51,14 @@ def redfish_conn_mock():
 def run_cli_command():
     """Fixture that returns a function to run CLI commands."""
 
-    def _run_command(args: List[str], check: bool = False):
+    def _run_command(args: List[str], check: bool = False, cwd=None):
         """Run a node-scraper CLI command.
 
         Args:
             args: List of command-line arguments
             check: If True, raise CalledProcessError on non-zero exit
+            cwd: Working directory for the command (default: the caller's). Use a tmp dir for
+                plugins whose probes write files to the cwd, e.g. wget saving index.html
 
         Returns:
             subprocess.CompletedProcess instance
@@ -68,6 +70,7 @@ def _run_command(args: List[str], check: bool = False):
             capture_output=True,
             text=True,
             check=check,
+            cwd=cwd,
         )
 
     return _run_command

test/functional/test_cli_no_console_log.py

@@ -28,7 +28,7 @@
 import sys
 
 
-def test_no_console_log_with_log_path_none_still_parses():
+def test_no_console_log_with_log_path_none_still_parses(tmp_path):
     """--log-path None + --no-console-log defaults to ./scraper_logs_* (no argparse error)."""
     result = subprocess.run(
         [
@@ -44,6 +44,7 @@ def test_no_console_log_with_log_path_none_still_parses():
         ],
         capture_output=True,
         text=True,
+        cwd=tmp_path,
     )
     assert result.returncode == 0, result.stderr
     combined = (result.stdout or "") + (result.stderr or "")

test/functional/test_network_plugin.py

@@ -158,6 +158,7 @@ def test_network_plugin_with_netprobe_wget(run_cli_command, tmp_path):
             "wget",
         ],
         check=False,
+        cwd=tmp_path,  # wget saves the fetched page (index.html, index.html.1, ...) into the cwd
     )
 
     assert result.returncode in [0, 1, 2]

test/functional/test_plugin_configs.py

@@ -181,25 +181,34 @@ def test_plugin_config_with_multiple_configs(run_cli_command, plugin_config_file
 def test_plugin_config_with_nonexistent_file(run_cli_command, tmp_path):
     """Test that a nonexistent config file path fails gracefully."""
     nonexistent_path = str(tmp_path / "nonexistent_config.json")
-    result = run_cli_command([f"--plugin-configs={nonexistent_path}"], check=False)
+    result = run_cli_command(
+        ["--log-path", str(tmp_path / "logs"), f"--plugin-configs={nonexistent_path}"],
+        check=False,
+    )
 
     assert result.returncode != 0
     output = (result.stdout + result.stderr).lower()
     assert "error" in output or "no plugin config found" in output
 
 
-def test_plugin_config_with_invalid_builtin_name(run_cli_command):
+def test_plugin_config_with_invalid_builtin_name(run_cli_command, tmp_path):
     """Test that an invalid built-in config name fails gracefully."""
-    result = run_cli_command(["--plugin-configs=NonExistentConfig"], check=False)
+    result = run_cli_command(
+        ["--log-path", str(tmp_path / "logs"), "--plugin-configs=NonExistentConfig"],
+        check=False,
+    )
 
     assert result.returncode != 0
     output = (result.stdout + result.stderr).lower()
     assert "error" in output or "no plugin config found" in output
 
 
-def test_plugin_config_with_invalid_json(run_cli_command, invalid_plugin_config):
+def test_plugin_config_with_invalid_json(run_cli_command, invalid_plugin_config, tmp_path):
     """Test that an invalid JSON file fails gracefully."""
-    result = run_cli_command([f"--plugin-configs={invalid_plugin_config}"], check=False)
+    result = run_cli_command(
+        ["--log-path", str(tmp_path / "logs"), f"--plugin-configs={invalid_plugin_config}"],
+        check=False,
+    )
 
     assert result.returncode != 0
     output = (result.stdout + result.stderr).lower()

test/functional/test_run_plugins.py

@@ -93,9 +93,12 @@ def test_run_all_plugins_together(run_cli_command, all_plugins, tmp_path):
     assert len(output) > 0
 
 
-def test_run_plugin_with_invalid_name(run_cli_command):
+def test_run_plugin_with_invalid_name(run_cli_command, tmp_path):
     """Test that running a non-existent plugin logs a warning and falls back to default config."""
-    result = run_cli_command(["run-plugins", "NonExistentPlugin"], check=False)
+    result = run_cli_command(
+        ["--log-path", str(tmp_path / "logs"), "run-plugins", "NonExistentPlugin"],
+        check=False,
+    )
 
     # Invalid plugin is ignored and default config runs instead
     # Exit code depends on whether default config plugins succeed

test/unit/framework/test_compare_runs.py

@@ -126,32 +126,38 @@ def test_run_compare_runs_no_data(caplog, tmp_path):
     (tmp_path / "run2").mkdir()
     logger = logging.getLogger()
     plugin_reg = PluginRegistry()
-    run_compare_runs(str(tmp_path / "run1"), str(tmp_path / "run2"), plugin_reg, logger)
+    run_compare_runs(
+        str(tmp_path / "run1"),
+        str(tmp_path / "run2"),
+        plugin_reg,
+        logger,
+        artifact_dir=str(tmp_path),
+    )
     assert "No plugin data found" in caplog.text
 
 
-def test_run_compare_runs_with_fixture_dirs(caplog, framework_fixtures_path):
+def test_run_compare_runs_with_fixture_dirs(caplog, framework_fixtures_path, tmp_path):
     """With two log dirs using fixture (same content), table is produced and no diffs."""
     caplog.set_level(logging.INFO)
     logger = logging.getLogger()
     plugin_reg = PluginRegistry()
     base = framework_fixtures_path / "log_dir"
-    run_compare_runs(str(base), str(base), plugin_reg, logger)
+    run_compare_runs(str(base), str(base), plugin_reg, logger, artifact_dir=str(tmp_path))
     assert "Loading run 1" in caplog.text
     assert "Loading run 2" in caplog.text
     assert "Plugin" in caplog.text
     assert "No differences" in caplog.text or "difference" in caplog.text.lower()
 
 
-def test_run_compare_runs_sysctl_fixture(caplog, framework_fixtures_path):
+def test_run_compare_runs_sysctl_fixture(caplog, framework_fixtures_path, tmp_path):
     """Compare runs using compare_runs_sysctl fixtures; run1 and run2 differ (SysctlPlugin)."""
     caplog.set_level(logging.INFO)
     logger = logging.getLogger()
     base = framework_fixtures_path / "compare_runs_sysctl"
     run1 = base / "run1"
     run2 = base / "run2"
     plugin_reg = PluginRegistry()
-    run_compare_runs(str(run1), str(run2), plugin_reg, logger)
+    run_compare_runs(str(run1), str(run2), plugin_reg, logger, artifact_dir=str(tmp_path))
     assert "Loading run 1" in caplog.text
     assert "Loading run 2" in caplog.text
     assert "Sysctl" in caplog.text or "sysctl" in caplog.text
@@ -167,7 +173,7 @@ def test_run_compare_runs_one_run_missing_plugin(caplog, framework_fixtures_path
     run2 = tmp_path / "run2"
     run2.mkdir()
     # run2 has no collector dirs
-    run_compare_runs(str(run1), str(run2), plugin_reg, logger)
+    run_compare_runs(str(run1), str(run2), plugin_reg, logger, artifact_dir=str(tmp_path))
     assert "Loading run 1" in caplog.text
     assert "Loading run 2" in caplog.text
     assert "not found in run 2" in caplog.text or "NOT_RAN" in caplog.text