[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 80.07414% with 215 lines in your changes missing coverage. Please review.
✅ Project coverage is 50.75%. Comparing base (da993f7) to head (1954640).
⚠️ Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
openml/_config.py	76.65%	67 Missing ⚠️
openml/_api/clients/http.py	79.46%	46 Missing ⚠️
openml/_api/resources/flow.py	86.02%	13 Missing ⚠️
openml/_api/resources/base/versions.py	85.71%	11 Missing ⚠️
openml/_api/setup/backend.py	80.70%	11 Missing ⚠️
openml/cli.py	0.00%	11 Missing ⚠️
openml/_api/setup/builder.py	81.08%	7 Missing ⚠️
openml/_api/resources/base/fallback.py	84.21%	6 Missing ⚠️
openml/flows/flow.py	57.14%	6 Missing ⚠️
openml/flows/functions.py	40.00%	6 Missing ⚠️
... and 12 more

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1609      +/-   ##
==========================================
- Coverage   53.09%   50.75%   -2.34%     
==========================================
  Files          37       61      +24     
  Lines        4362     5079     +717     
==========================================
+ Hits         2316     2578     +262     
- Misses       2046     2501     +455

☔ 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.

geetu040

please sync with base PR and update with these comments #1576 (comment)

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

geetu040 · 2026-02-26T02:41:34Z

tests/test_flows/test_flow.py

            9,
        ]:
-            flow_xml = _perform_api_call("flow/%d" % flow_id, request_method="get")
+            flow_xml = openml.config.get_backend().http_client.get(f"flow/{flow_id}").text


I have updated the PR to make this work, please sync

Suggested change

flow_xml = openml.config.get_backend().http_client.get(f"flow/{flow_id}").text

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

…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.

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

geetu040 added 17 commits February 24, 2026 15:47

majore config refactor

0fc917c

Merge branch 'pr-1577' into migration

3d86b18

update _config.py

aba3d3e

update test_openml_cache_dir_env_var

d99d54d

fix mutable SERVERS_REGISTRY

dc22e3a

update set_api_version for fallback

7318573

minor fix

29ef187

fixes for test_config

cf94c89

fixes in conftest urls

298fbda

update test_http.py

9870502

undo changes with test_openml_cache_dir_env_var

33065c2

fix server mode in test_config.py

76b92bb

move _HEADERS to confing

419edcb

add fixtures for migration tests

cb6d937

update test_http.py with fixtures

8544c8a

update test_versions.py

d4c413b

update test_versions.py

fab1a15

geetu040 suggested changes Feb 25, 2026

View reviewed changes

Omswastik-11 and others added 6 commits February 25, 2026 19:14

merge

4e75b92

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

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

	flow_xml = openml.config.get_backend().http_client.get(f"flow/{flow_id}").text
	flow_xml = openml._backend.http_client.get(f"flow/{flow_id}").text

	"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 []

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!

geetu040 left a comment

Choose a reason for hiding this comment

Uh oh!

geetu040 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 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!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Omswastik-11 commented Jan 8, 2026 •

edited

Loading

codecov-commenter commented Jan 12, 2026 •

edited

Loading