[ENH] V1 -> V2 Migration - Flows (module) by Omswastik-11 · Pull Request #1609 · openml/openml-python

Omswastik-11 · 2026-01-08T12:58:31Z

added a Create method in FlowAPI for publishing flow but not refactored with old publish . (Needs discussion on this)
Added tests using fake_methods so that we can test without local V2 server . I have tested the FlowsV2 methods (get and exists ) and delete and list were not implemented in V2 server so I skipped them .

…into issue1564

…into pr/1577

codecov-commenter · 2026-01-12T13:26:15Z

Codecov Report

❌ Patch coverage is 64.04078% with 388 lines in your changes missing coverage. Please review.
✅ Project coverage is 54.34%. Comparing base (da993f7) to head (ffe596d).
⚠️ Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
openml/_config.py	68.66%	89 Missing ⚠️
openml/_api/clients/http.py	61.16%	87 Missing ⚠️
openml/_api/resources/base/versions.py	31.16%	53 Missing ⚠️
openml/_api/resources/flow.py	46.23%	50 Missing ⚠️
openml/_api/resources/base/fallback.py	26.31%	28 Missing ⚠️
openml/cli.py	0.00%	14 Missing ⚠️
openml/_api/resources/base/base.py	62.85%	13 Missing ⚠️
openml/_api/setup/backend.py	80.70%	11 Missing ⚠️
openml/flows/flow.py	21.42%	11 Missing ⚠️
openml/_api/setup/builder.py	81.08%	7 Missing ⚠️
... and 10 more

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1609      +/-   ##
==========================================
+ Coverage   53.09%   54.34%   +1.25%     
==========================================
  Files          37       61      +24     
  Lines        4362     5073     +711     
==========================================
+ Hits         2316     2757     +441     
- Misses       2046     2316     +270

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

…-11/openml-python into flow-migration-stacked

tests/test_flows/test_flow.py

…into flow-migration-stacked

Copilot

Pull request overview

Copilot reviewed 53 out of 54 changed files in this pull request and generated 11 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-26T10:57:05Z

tests/test_openml/test_config.py

 from typing import Any, Iterator
 from pathlib import Path
 import platform
+from urllib.parse import urlparse


urlparse is imported but never used in this test module. Please remove the unused import to avoid lint failures.

Copilot · 2026-02-26T10:57:06Z

tests/test_evaluations/test_evaluations_example.py

        # Example script which will appear in the upcoming OpenML-Python paper
        # This test ensures that the example will keep running!
-        with overwrite_config_context(
+        with openml.config.overwrite_config_context(  # noqa: F823


overwrite_config_context is referenced via openml.config and should be resolvable here, so the # noqa: F823 suppression looks incorrect/unnecessary. Please remove it (or use the correct code if there is an actual linter error to suppress).

Copilot · 2026-02-26T10:57:06Z

tests/test_flows/test_flow_functions.py

+@mock.patch.object(requests.Session, "request")
+def test_delete_flow_not_owned(mock_request, test_files_directory, test_api_key):
+    openml.config.start_using_configuration_for_example()
    content_file = test_files_directory / "mock_responses" / "flows" / "flow_delete_not_owned.xml"


start_using_configuration_for_example() mutates global configuration state; these tests never call stop_using_configuration_for_example(), which can leak state into later tests and cause order-dependent failures. Please wrap this in a context/fixture that guarantees stop_... runs (e.g., try/finally or a dedicated pytest fixture).

Copilot · 2026-02-26T10:57:07Z

openml/_config.py

+            "apikey": "normaluser",
+        },
+        APIVersion.V2: {
+            "server": "http://localhost:8002/api/v1/xml/",


SERVERS_REGISTRY['local'][APIVersion.V2]['server'] points to a v1 XML base path (/api/v1/xml/). This looks inconsistent with the v2 clients (which call endpoints like flows/.../ under an /api/v2/ base). Please correct the local v2 base URL (and/or add a test to validate v2 server base URLs).

Suggested change

"server": "http://localhost:8002/api/v1/xml/",

"server": "http://localhost:8002/api/v2/",

Copilot · 2026-02-26T10:57:07Z

openml/_api/resources/base/versions.py

+            if isinstance(id_value, (str, int)):
+                return int(id_value)
+
+        # Fallback: check all values for numeric/string IDs
+        for v in root_value.values():
+            if isinstance(v, (str, int)):
+                return int(v)


In _extract_id_from_upload, the fallback loop attempts int(v) for the first string value it finds. This can raise ValueError on non-numeric strings like the XML namespace (e.g. "http://openml.org/openml"). Please restrict the fallback to numeric strings (or explicitly look for known keys) before casting to int.

Suggested change

if isinstance(id_value, (str, int)):

return int(id_value)

# Fallback: check all values for numeric/string IDs

for v in root_value.values():

if isinstance(v, (str, int)):

return int(v)

if isinstance(id_value, int):

return id_value

if isinstance(id_value, str):

id_str = id_value.strip()

if id_str.isdigit():

return int(id_str)

# Fallback: check all values for numeric/string IDs

for v in root_value.values():

if isinstance(v, int):

return v

if isinstance(v, str):

v_str = v.strip()

if v_str.isdigit():

return int(v_str)

Copilot · 2026-02-26T10:57:08Z

openml/_config.py

+        self._config = replace(
+            self._config,
+            servers=config["servers"],
+            api_version=config["api_version"],
+            fallback_api_version=config["fallback_api_version"],
+            show_progress=config["show_progress"],
+            avoid_duplicate_runs=config["avoid_duplicate_runs"],
+            retry_policy=config["retry_policy"],
+            connection_n_retries=int(config["connection_n_retries"]),
+        )


_setup() assigns api_version and fallback_api_version directly from the parsed config dict. If these values come from a config file/CLI they will be strings (e.g. "v2"), which will break later lookups like servers[self.api_version] (servers keys are APIVersion). Please coerce string values to APIVersion (and validate) when loading config, and consider similarly validating/normalizing servers.

Copilot · 2026-02-26T10:57:08Z

openml/_api/resources/flow.py

+        return OpenMLFlow._from_dict(xmltodict.parse(flow_xml))
+


FlowV1API.get() does not detect v1-style error payloads (<oml:error>...) that are returned with HTTP 200. HTTPClient only validates by status code, so this method can end up passing an error dict into OpenMLFlow._from_dict() and failing with a confusing parsing error. Please add an <oml:error> check similar to exists() / list() and raise OpenMLServerException with the server-provided code/message.

Suggested change

return OpenMLFlow._from_dict(xmltodict.parse(flow_xml))

result_dict = xmltodict.parse(flow_xml)

# Detect v1-style error payloads and raise a clear exception

if "oml:error" in result_dict:

err = result_dict["oml:error"]

code = int(err.get("oml:code", 0)) if "oml:code" in err else None

message = err.get("oml:message", "Server returned an error")

raise OpenMLServerException(message=message, code=code)

return OpenMLFlow._from_dict(result_dict)

Copilot · 2026-02-26T10:57:08Z

openml/_api/clients/http.py

+        path_parts = parsed_url.path.strip("/").split("/")
+
+        filtered_params = {k: v for k, v in params.items() if k != "api_key"}
+        params_part = [urlencode(filtered_params)] if filtered_params else []


HTTPCache.get_key() uses urlencode(filtered_params) over the raw dict, which makes the cache key depend on the insertion order of params. This can cause avoidable cache misses for semantically identical requests. Please sort parameters (e.g., by key) before encoding to make cache keys stable.

Suggested change

params_part = [urlencode(filtered_params)] if filtered_params else []

sorted_params = sorted(filtered_params.items())

params_part = [urlencode(sorted_params)] if sorted_params else []

Copilot · 2026-02-26T10:57:09Z

openml/flows/flow.py

+    def push_tag(self, tag: str) -> None:
+        """Annotates this flow with a tag on the server.
+
+        Parameters
+        ----------
+        tag : str
+            Tag to attach to the flow.
+        """
+        if self.flow_id is None:
+            raise ValueError("Flow does not have an ID. Please publish the flow before tagging.")
+        openml._backend.flow.tag(self.flow_id, tag)
+
+    def remove_tag(self, tag: str) -> None:
+        """Removes a tag from this flow on the server.
+
+        Parameters
+        ----------
+        tag : str
+            Tag to remove from the flow.
+        """
+        if self.flow_id is None:
+            raise ValueError("Flow does not have an ID. Please publish the flow before untagging.")
+        openml._backend.flow.untag(self.flow_id, tag)


OpenMLFlow already inherits push_tag / remove_tag from OpenMLBase. Re-defining them here creates duplicated API paths and potentially inconsistent behavior across resource types (some entities tag via openml.utils._tag_openml_base, flows via openml._backend). Consider removing these overrides and updating the shared implementation in OpenMLBase to use the backend for all resources instead.

Copilot · 2026-02-26T10:57:09Z

tests/test_api/test_versions.py

+def dummy_task_v2(http_client_v2, minio_client) -> DummyTaskV1API:
+    return DummyTaskV2API(http=http_client_v2, minio=minio_client)
+
+
+@pytest.fixture
+def dummy_task_fallback(dummy_task_v1, dummy_task_v2) -> DummyTaskV1API:
+    return FallbackProxy(dummy_task_v2, dummy_task_v1)


The fixture return type annotations in this file look incorrect: dummy_task_v2 is annotated as DummyTaskV1API but returns DummyTaskV2API, and dummy_task_fallback is annotated as DummyTaskV1API but returns FallbackProxy. Please fix the annotations to match the actual returned objects to avoid type-checking confusion.

Co-authored-by: Armaghan Shakir <raoarmaghanshakir040@gmail.com>

Copilot

Pull request overview

Copilot reviewed 57 out of 58 changed files in this pull request and generated 8 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-27T13:08:28Z

tests/test_api/test_versions.py

+@pytest.fixture
+def dummy_task_v2(http_client_v2, minio_client) -> DummyTaskV1API:
+    return DummyTaskV2API(http=http_client_v2, minio=minio_client)
+
+
+@pytest.fixture
+def dummy_task_fallback(dummy_task_v1, dummy_task_v2) -> DummyTaskV1API:
+    return FallbackProxy(dummy_task_v2, dummy_task_v1)
+


Fixture return type annotations are inconsistent: dummy_task_v2 is annotated as returning DummyTaskV1API but returns DummyTaskV2API, and dummy_task_fallback is annotated as DummyTaskV1API but returns a FallbackProxy. Fix the annotations to match the actual return types to avoid confusing readers and type checkers.

Copilot · 2026-02-27T13:08:29Z

tests/test_flows/test_flow.py

+import requests
+from openml.testing import SimpleImputer, TestBase, create_request_response



create_request_response is imported but unused in this file. Either remove the import or use it to build the mocked Response objects for consistency with other tests.

Copilot · 2026-02-27T13:08:29Z

openml/_api/clients/http.py

+        OpenMLHashException
+            If checksum verification fails.
+        """
+        url = urljoin(self.server, path)


urljoin(self.server, path) will generate incorrect URLs if self.server does not end with a trailing / (e.g. base .../api/v1/xml + task/1 becomes .../api/v1/task/1). Existing user config files and tests still set servers without the trailing slash, so requests can silently hit the wrong endpoint. Normalize server to always end with / (e.g. in config setup/server setter) or avoid urljoin here and manually ensure exactly one / separator.

Suggested change

url = urljoin(self.server, path)

if path.startswith(("https://", "https://")):

url = path

else:

base = self.server.rstrip("/")

url = f"{base}/{path.lstrip('/')}"

Copilot · 2026-02-27T13:08:29Z

tests/test_flows/test_flow.py

+            flow_xml = openml._backend.http_client.get(f"flow/{flow_id}").text
            flow_dict = xmltodict.parse(flow_xml)


openml.config.get_backend() does not exist (config is now an OpenMLConfigManager instance). This will raise an AttributeError and break the test. Use the new backend entrypoint (openml._backend.http_client.get(...) or openml._backend.flow.get(...)) instead of calling a non-existent config method.

Copilot · 2026-02-27T13:08:29Z

tests/test_openml/test_config.py

+    openml.config.server = "temp-server1"
+    openml.config.apikey = "temp-apikey1"
+    openml.config.get_servers(mode)["server"] = 'temp-server2'
+    openml.config.get_servers(mode)["apikey"] = 'temp-server2'
+


get_servers() returns a dict keyed by APIVersion, but this test indexes it with string keys (["server"]/["apikey"]), which will raise KeyError. To test deepcopy/immutability, mutate openml.config.get_servers(mode)[api_version]["server"] (and apikey) instead.

Copilot · 2026-02-27T13:08:29Z

tests/test_openml/test_config.py

    @pytest.mark.production_server()
    def test_switch_to_example_configuration(self):
        """Verifies the test configuration is loaded properly."""
-        # Below is the default test key which would be used anyway, but just for clarity:
-        openml.config.apikey = "any-api-key"
-        openml.config.server = self.production_server
+        openml.config.set_servers("production")

        openml.config.start_using_configuration_for_example()

-        assert openml.config.apikey == TestBase.user_key
-        assert openml.config.server == self.test_server
+        openml.config.servers = openml.config.get_servers("test")

    @pytest.mark.production_server()
    def test_switch_from_example_configuration(self):
        """Verifies the previous configuration is loaded after stopping."""
        # Below is the default test key which would be used anyway, but just for clarity:
-        openml.config.apikey = TestBase.user_key
-        openml.config.server = self.production_server
+        openml.config.set_servers("production")

        openml.config.start_using_configuration_for_example()
        openml.config.stop_using_configuration_for_example()
-
-        assert openml.config.apikey == TestBase.user_key
-        assert openml.config.server == self.production_server
+        openml.config.servers = openml.config.get_servers("production")



The example-configuration tests no longer assert that start_using_configuration_for_example() actually switched to the test config (or that stopping restores the previous config). They currently just overwrite openml.config.servers manually, which defeats the purpose of the test and could hide regressions. Add assertions comparing openml.config.servers (and/or server/apikey) before/after start/stop instead of mutating the config inside the test.

Copilot · 2026-02-27T13:08:30Z

openml/__init__.py

@@ -33,6 +35,7 @@
    utils,
 )
 from .__version__ import __version__
+from ._api import _backend
 from .datasets import OpenMLDataFeature, OpenMLDataset
 from .evaluations import OpenMLEvaluation
 from .flows import OpenMLFlow
@@ -49,6 +52,11 @@
    OpenMLTask,
 )

+if TYPE_CHECKING:
+    from ._config import OpenMLConfigManager
+
+config: OpenMLConfigManager = _config_module.__config
+


The package-level change makes openml.config an instance attribute and removes the openml/config.py module. This is a breaking change for users who do import openml.config or from openml.config import .... If backward compatibility is desired, consider reintroducing a thin openml/config.py shim that re-exports the config manager (and legacy symbols) so existing imports keep working.

Copilot · 2026-02-27T13:08:30Z

tests/test_openml/test_config.py

+from urllib.parse import urlparse



from urllib.parse import urlparse is imported but never used in this test module, which will fail linting in configurations that enforce unused-import checks. Remove the import or use it in the tests.

…-11/openml-python into flow-migration-stacked

Copilot

Pull request overview

Copilot reviewed 57 out of 58 changed files in this pull request and generated 9 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-27T13:52:57Z

openml/_api/resources/flow.py

+    def publish(self, path: str | None = None, files: Mapping[str, Any] | None = None) -> int:  # type: ignore[override]  # noqa: ARG002
+        self._not_supported(method="publish")


The publish method signature in FlowV2API has a # type: ignore[override] comment, but this is because the return type doesn't match the parent class ResourceV2API.publish. However, the parent ResourceV2API.publish raises OpenMLNotSupportedError, so it never actually returns. The return type annotation -> int is misleading since the method always raises an exception. Consider using -> NoReturn as the return type or removing the return type annotation entirely since the method only raises.

Copilot · 2026-02-27T13:52:58Z

openml/_api/resources/flow.py

+    def list(
+        self,
+        limit: int | None = None,  # noqa: ARG002
+        offset: int | None = None,  # noqa: ARG002
+        tag: str | None = None,  # noqa: ARG002
+        uploader: str | None = None,  # noqa: ARG002
+    ) -> pd.DataFrame:
+        self._not_supported(method="list")


Similar to the publish method, the list method in FlowV2API should have -> NoReturn as its return type instead of -> pd.DataFrame since it only calls self._not_supported() which never returns.

Copilot · 2026-02-27T13:52:58Z

openml/_api/resources/flow.py

+    def publish(self, path: str | None = None, files: Mapping[str, Any] | None = None) -> int:
+        """Publish a flow on the OpenML server.
+
+        Parameters
+        ----------
+        files : Mapping[str, Any] | None
+            Files to upload (including description).
+
+        Returns
+        -------
+        int
+            The server-assigned flow id.
+        """
+        path = "flow"
+        return super().publish(path, files)


The path parameter in the publish method is marked as optional (str | None), but on line 159 it's immediately overwritten with the hardcoded value "flow". This makes the parameter pointless. Either remove the path parameter from the method signature entirely, or remove the line that overwrites it if the parameter should actually be used.

Copilot · 2026-02-27T13:52:58Z

openml/flows/functions.py

+        Whether to ignore the cache. If ``true`` this will download and overwrite the flow xml
+        even if the requested flow is already cached.


The docstring states "If true this will download..." but the parameter is ignore_cache, not true. It should say "If True, this will download..." or "If set to True, this will download..." for consistency with Python boolean conventions.

Copilot · 2026-02-27T13:52:59Z

openml/_api/resources/base/versions.py

+    def publish(self, path: str, files: Mapping[str, Any] | None) -> int:  # noqa: ARG002
+        self._not_supported(method="publish")
+
+    def delete(self, resource_id: int) -> bool:  # noqa: ARG002
+        self._not_supported(method="delete")
+
+    def tag(self, resource_id: int, tag: str) -> list[str]:  # noqa: ARG002
+        self._not_supported(method="tag")
+
+    def untag(self, resource_id: int, tag: str) -> list[str]:  # noqa: ARG002
+        self._not_supported(method="untag")


The return type annotation for these methods in ResourceV2API is incorrect. Since these methods call self._not_supported() which has a return type of NoReturn, these methods should also have -> NoReturn as their return type instead of -> int, -> bool, or -> list[str]. The # noqa: ARG002 comment suggests awareness that the arguments are unused, but the return type should also reflect that these methods never return.

Copilot · 2026-02-27T13:52:59Z

openml/_config.py

+            "openml_logger",
+            "_examples",
+            "OPENML_CACHE_DIR_ENV_VAR",
+            "OPENML_SKIP_PARQUET_ENV_VAR",


The attribute OPENML_TEST_SERVER_ADMIN_KEY_ENV_VAR is missing from the allowed set in __setattr__ on line 166-177. This could cause issues if code tries to set this attribute. It should be added to the set that includes OPENML_CACHE_DIR_ENV_VAR and OPENML_SKIP_PARQUET_ENV_VAR.

Suggested change

"OPENML_SKIP_PARQUET_ENV_VAR",

"OPENML_SKIP_PARQUET_ENV_VAR",

"OPENML_TEST_SERVER_ADMIN_KEY_ENV_VAR",

Copilot · 2026-02-27T13:52:59Z

openml/_config.py

+        self._HEADERS: dict[str, str] = {"user-agent": f"openml-python/{__version__}"}
+
+        self._config: OpenMLConfig = OpenMLConfig()
+        # for legacy test `test_non_writable_home`


In line 147, there's a comment "# for legacy test test_non_writable_home" but the _defaults attribute is also used in other methods like set_field_in_config_file (line 461). The comment is misleading as it suggests this is only for one specific test when it may have broader usage. Consider updating the comment or verifying if _defaults is truly only needed for that one test.

Suggested change

# for legacy test `test_non_writable_home`

# snapshot of default config (used for resets, e.g. in legacy tests like `test_non_writable_home`)

Copilot · 2026-02-27T13:53:00Z

tests/test_utils/test_utils.py


 def _mocked_perform_api_call(call, request_method):
-    url = openml.config.server + "/" + call
+    url = openml.config.server  + call


The URL construction has a spacing issue. There's an extra space before the + operator which results in inconsistent concatenation. The line should be url = openml.config.server + call without the extra space before the + operator.

Copilot · 2026-02-27T13:53:00Z

tests/test_flows/test_flow.py

        TestBase._mark_entity_for_removal("flow", flow.flow_id, flow.name)
        TestBase.logger.info(f"collected from {__file__.split('/')[-1]}: {flow.flow_id}")

+


There's a trailing whitespace at the end of line 301. Please remove the whitespace after the comment to maintain code cleanliness.

satvshr and others added 24 commits December 30, 2025 02:07

changes made

bf6e000

set up folder structure and base code

0159f47

Merge branch 'issue1564' of https://github.com/satvshr/openml-python …

8b6af81

…into issue1564

bug fixing

834782c

test failures fix

38ae9be

Update flow_id_tutorial.py

93ab9c2

_defaults bug fixing

aa25dd6

Merge branch 'main' into migration

58e9175

removed __setattr__ given it is not supported

a98b6b1

Merge branch 'main' into issue1564

7c82054

Merge branch 'main' into migration

bdd65ff

Merge branch 'main' into issue1564

f8fbe1e

Merge branch 'main' into issue1564

4fdcb64

Merge branch 'main' into issue1564

b3513f0

fix pre-commit

52ef379

Update all files

146dd21

Update introduction_tutorial.py

7a67bf0

refactor

5dfcbce

implement cache_dir

2acbe99

refactor

af99880

Merge branch 'main' into pr/1577

b111905

Merge branch 'issue1564' of https://github.com/satvshr/openml-python …

83f36c2

…into pr/1577

Merge branch 'main' into pr/1576

74ab366

migrate flow module

561b204

Omswastik-11 changed the title ~~[ENH] V1 -> V2 Migration~~ [ENH] V1 -> V2 Migration - Flows (module) Jan 8, 2026

implement FlowsV2.exists() and get() with JSON parsing

860b1b6

Omswastik-11 marked this pull request as ready for review January 8, 2026 15:09

geetu040 mentioned this pull request Jan 9, 2026

[ENH] V1 → V2 API Migration #1575

Open

25 tasks

skip delete flows tests

36c22aa

Omswastik-11 and others added 5 commits February 25, 2026 19:39

modify the tests

49c6d59

Merge branch 'flow-migration-stacked' of https://github.com/Omswastik…

af9f5e6

…-11/openml-python into flow-migration-stacked

fix error message in HTTPClient.server

276324a

fixes in test_versions.py: use DummyTaskAPI instead of TaskAPI

73f7594

add clients in openml._backend

2ee7fa3

geetu040 suggested changes Feb 26, 2026

View reviewed changes

tests/test_flows/test_flow.py Outdated Show resolved Hide resolved

geetu040 mentioned this pull request Feb 26, 2026

[ENH] V1 → V2 API Migration - core structure #1576

Open

Merge branch 'migration' of https://github.com/geetu040/openml-python …

1954640

…into flow-migration-stacked

Copilot AI review requested due to automatic review settings February 26, 2026 10:48

Copilot started reviewing on behalf of Omswastik-11 February 26, 2026 10:49 View session

Copilot AI reviewed Feb 26, 2026

View reviewed changes

geetu040 and others added 9 commits February 26, 2026 22:57

fixes with openml.config.[server|apikey] leakage

4be5bbd

remove unused fixtures: use_api_[v1|v2]

9027c01

add more config tests

e5461a9

make SERVERS_REGISTRY private

7d899a9

fix marker: uses_test_server->test_server

8587414

fix UserWarning

23a3450

update fixture: with_server

ac28f82

merge

097e27c

add a fallback in v2

c28c808

Copilot AI review requested due to automatic review settings February 27, 2026 13:00

Copilot started reviewing on behalf of Omswastik-11 February 27, 2026 13:01 View session

Update tests/test_flows/test_flow.py

2d698ed

Co-authored-by: Armaghan Shakir <raoarmaghanshakir040@gmail.com>

Copilot AI reviewed Feb 27, 2026

View reviewed changes

Omswastik-11 added 3 commits February 27, 2026 18:49

fix the test_api_key name in test

6559219

Merge branch 'flow-migration-stacked' of https://github.com/Omswastik…

9f3ca32

…-11/openml-python into flow-migration-stacked

remove authentication require ness for list

ffe596d

Copilot AI review requested due to automatic review settings February 27, 2026 13:48

Copilot started reviewing on behalf of Omswastik-11 February 27, 2026 13:48 View session

Copilot AI reviewed Feb 27, 2026

View reviewed changes

	"server": "http://localhost:8002/api/v1/xml/",
	"server": "http://localhost:8002/api/v2/",

-            if isinstance(id_value, (str, int)):
-                return int(id_value)
-        # Fallback: check all values for numeric/string IDs
-        for v in root_value.values():
-            if isinstance(v, (str, int)):
-                return int(v)
+            if isinstance(id_value, int):
+                return id_value
+            if isinstance(id_value, str):
+                id_str = id_value.strip()
+                if id_str.isdigit():
+                    return int(id_str)
+        # Fallback: check all values for numeric/string IDs
+        for v in root_value.values():
+            if isinstance(v, int):
+                return v
+            if isinstance(v, str):
+                v_str = v.strip()
+                if v_str.isdigit():
+                    return int(v_str)

-        return OpenMLFlow._from_dict(xmltodict.parse(flow_xml))
+        result_dict = xmltodict.parse(flow_xml)
+        # Detect v1-style error payloads and raise a clear exception
+        if "oml:error" in result_dict:
+            err = result_dict["oml:error"]
+            code = int(err.get("oml:code", 0)) if "oml:code" in err else None
+            message = err.get("oml:message", "Server returned an error")
+            raise OpenMLServerException(message=message, code=code)
+        return OpenMLFlow._from_dict(result_dict)

	params_part = [urlencode(filtered_params)] if filtered_params else []
	sorted_params = sorted(filtered_params.items())
	params_part = [urlencode(sorted_params)] if sorted_params else []

		import requests
		from openml.testing import SimpleImputer, TestBase, create_request_response

-        url = urljoin(self.server, path)
+        if path.startswith(("https://", "https://")):
+            url = path
+        else:
+            base = self.server.rstrip("/")
+            url = f"{base}/{path.lstrip('/')}"

		flow_xml = openml._backend.http_client.get(f"flow/{flow_id}").text
		flow_dict = xmltodict.parse(flow_xml)

		def publish(self, path: str \| None = None, files: Mapping[str, Any] \| None = None) -> int: # type: ignore[override] # noqa: ARG002
		self._not_supported(method="publish")

		Whether to ignore the cache. If ``true`` this will download and overwrite the flow xml
		even if the requested flow is already cached.

	"OPENML_SKIP_PARQUET_ENV_VAR",
	"OPENML_SKIP_PARQUET_ENV_VAR",
	"OPENML_TEST_SERVER_ADMIN_KEY_ENV_VAR",

	# for legacy test `test_non_writable_home`
	# snapshot of default config (used for resets, e.g. in legacy tests like `test_non_writable_home`)

		TestBase._mark_entity_for_removal("flow", flow.flow_id, flow.name)
		TestBase.logger.info(f"collected from {__file__.split('/')[-1]}: {flow.flow_id}")

Uh oh!

Conversation

Omswastik-11 commented Jan 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov-commenter commented Jan 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 27, 2026

Omswastik-11 commented Jan 8, 2026 •

edited

Loading

codecov-commenter commented Jan 12, 2026 •

edited

Loading