fix: improve type safety and error handling in tests #3326

mkoura · 2026-02-10T15:48:03Z

Use enum.StrEnum instead of str, enum.Enum for TableCondition and ColumnCondition for better type safety.
Use getattr for subtest scenario names to avoid AttributeError if name is missing.
Add type: ignore comments to suppress mypy/pyright warnings where function signatures are not fully compatible.
Use tp.cast for type casting in test_ledger_state and test_staking_rewards for clarity and type safety.
Simplify conditional expressions using or instead of ternary operators.
Add missing psycopg2.extensions import for type checking.
Add type: ignore for possibly missing LockNotAvailable attribute in dbsync_queries.

- Use enum.StrEnum instead of str, enum.Enum for TableCondition and ColumnCondition for better type safety. - Use getattr for subtest scenario names to avoid AttributeError if __name__ is missing. - Add type: ignore comments to suppress mypy/pyright warnings where function signatures are not fully compatible. - Use tp.cast for type casting in test_ledger_state and test_staking_rewards for clarity and type safety. - Simplify conditional expressions using `or` instead of ternary operators. - Add missing psycopg2.extensions import for type checking. - Add type: ignore for possibly missing LockNotAvailable attribute in dbsync_queries.

Copilot

Pull request overview

This PR focuses on improving type safety and making test utilities more robust to edge-cases in naming/introspection, while also simplifying a set of conditional expressions.

Changes:

Replaced multiple x if x else y patterns with x or y across utils/tests.
Improved robustness of subtest scenario naming using getattr(..., "__name__", ...).
Added/adjusted typing aids (casts, imports for type checking, and type-ignore comments).

Reviewed changes

Copilot reviewed 21 out of 21 changed files in this pull request and generated 13 comments.

Show a summary per file

File	Description
cardano_node_tests/utils/governance_utils.py	Simplifies list defaulting logic when building governance action data.
cardano_node_tests/utils/dbsync_queries.py	Adds psycopg2 typing import and suppresses type-checker warning on a psycopg2 error attribute.
cardano_node_tests/utils/dbsync_conn.py	Adds psycopg2 typing import to support type checking.
cardano_node_tests/utils/clusterlib_utils.py	Adjusts type-ignore comment for `record_cli_coverage` call.
cardano_node_tests/utils/cluster_nodes.py	Simplifies defaulting of `service_names` argument.
cardano_node_tests/tests/tests_plutus_v2/test_mint_raw.py	Simplifies defaulting of datum hash file argument.
cardano_node_tests/tests/tests_plutus_v2/test_mint_build.py	Simplifies defaulting of inline datum file argument.
cardano_node_tests/tests/tests_plutus_v2/mint_raw.py	Simplifies defaulting of datum hash file argument.
cardano_node_tests/tests/tests_plutus_v2/mint_build.py	Simplifies defaulting of inline datum file argument.
cardano_node_tests/tests/tests_plutus/test_spend_build.py	Simplifies multiple optional file/value arguments to empty-string defaults.
cardano_node_tests/tests/tests_plutus/spend_raw.py	Simplifies multiple optional file/value arguments to empty-string defaults.
cardano_node_tests/tests/tests_plutus/spend_build.py	Simplifies multiple optional file/value arguments to empty-string defaults.
cardano_node_tests/tests/tests_conway/test_guardrails.py	Simplifies limit fallback logic and makes subtest naming resilient to missing `__name__`.
cardano_node_tests/tests/test_staking_rewards.py	Adds explicit cast for snapshot record typing.
cardano_node_tests/tests/test_pools.py	Makes subtest naming resilient to missing `__name__`.
cardano_node_tests/tests/test_mnemonic.py	Adds type-ignore comments for Hypothesis composite-strategy call patterns.
cardano_node_tests/tests/test_ledger_state.py	Adds explicit cast for snapshot record typing.
cardano_node_tests/tests/test_dbsync_config.py	Uses `enum.StrEnum` for stronger string-enum typing; makes subtest naming resilient.
cardano_node_tests/tests/test_dbsync.py	Simplifies tx_count summation defaulting to 0.
cardano_node_tests/tests/conftest.py	Changes type-check suppression style for xdist NodeManager EXIT_TIMEOUT override.
agent_docs/subtests.md	Updates docs to reflect safer subtest scenario naming.

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

cardano_node_tests/tests/tests_conway/test_guardrails.py

Copilot · 2026-02-10T15:53:44Z

cardano_node_tests/tests/tests_conway/test_guardrails.py

+    type_upper_limit = param.param_upper_limit or get_upper_limit_according_to_type(
+        param_predicates["type"]
    )


Using or here changes semantics when the limit is a valid falsy value (e.g., 0). If param.param_lower_limit or param.param_upper_limit can be 0, this code will incorrectly fall back to the default. Prefer an explicit is not None check (or use a sentinel) so that 0 is preserved.

Suggested change

type_upper_limit = param.param_upper_limit or get_upper_limit_according_to_type(

param_predicates["type"]

)

if param.param_upper_limit is not None:

type_upper_limit = param.param_upper_limit

else:

type_upper_limit = get_upper_limit_according_to_type(param_predicates["type"])

Copilot · 2026-02-10T15:53:44Z

cardano_node_tests/utils/clusterlib_utils.py

        custom_clusterlib.record_cli_coverage(
            cli_args=cardano_cli_args,
-            coverage_dict=cluster_obj.cli_coverage,  # pyright: ignore [reportAttributeAccessIssue]
+            coverage_dict=cluster_obj.cli_coverage,  # type: ignore[invalid-argument-type]


The ignore code invalid-argument-type is not a mypy error code (mypy typically uses arg-type for this case), and pyright won’t recognize it as a targeted suppression either. Use a correct mypy code (e.g., # type: ignore[arg-type]) and/or keep a # pyright: ignore[...] directive that matches the actual pyright diagnostic to ensure this suppression works with the intended type checker(s).

Suggested change

coverage_dict=cluster_obj.cli_coverage, # type: ignore[invalid-argument-type]

coverage_dict=cluster_obj.cli_coverage, # type: ignore[arg-type]

Copilot · 2026-02-10T15:53:45Z

cardano_node_tests/utils/dbsync_queries.py

        try:
            result = cur.fetchone()
-        except psycopg2.errors.LockNotAvailable as e:
+        except psycopg2.errors.LockNotAvailable as e:  # type: ignore[possibly-missing-attribute]


The ignore code possibly-missing-attribute is unlikely to be recognized by mypy (common mypy codes here are attr-defined/union-attr) and won’t map to a pyright rule name either. If the goal is to silence type-checking about LockNotAvailable not existing in some psycopg2 stubs, prefer a supported suppression (e.g., a mypy-compatible # type: ignore[attr-defined], or a # pyright: ignore[reportAttributeAccessIssue]) so CI type checks don’t regress.

Suggested change

except psycopg2.errors.LockNotAvailable as e: # type: ignore[possibly-missing-attribute]

except psycopg2.errors.LockNotAvailable as e: # type: ignore[attr-defined]

Copilot · 2026-02-10T15:53:45Z

cardano_node_tests/tests/test_mnemonic.py

+    tokens = draw(word_sequence(count=count, ensure_invalid=False))  # type: ignore[missing-argument]
+    line = draw(join_with_weirdness(tokens=tokens))  # type: ignore[missing-argument]


The targeted ignore code missing-argument is not a standard mypy error code (mypy typically reports these as call-arg). If these functions are Hypothesis @st.composite strategies, the runtime calling pattern is fine, but the suppression should use a code that your type checker actually understands (or fall back to plain # type: ignore / an appropriate # pyright: ignore[...]). Otherwise, the warnings you intended to suppress may still fail CI.

Copilot · 2026-02-10T15:53:46Z

cardano_node_tests/tests/test_mnemonic.py

+    wrong_count_files(),  # type: ignore[missing-argument]
+    unknown_word_files(),  # type: ignore[missing-argument]
+    separator_mess_files(),  # type: ignore[missing-argument]
+    encoding_garbage_files(),  # type: ignore[missing-argument]


The targeted ignore code missing-argument is not a standard mypy error code (mypy typically reports these as call-arg). If these functions are Hypothesis @st.composite strategies, the runtime calling pattern is fine, but the suppression should use a code that your type checker actually understands (or fall back to plain # type: ignore / an appropriate # pyright: ignore[...]). Otherwise, the warnings you intended to suppress may still fail CI.

Suggested change

wrong_count_files(), # type: ignore[missing-argument]

unknown_word_files(), # type: ignore[missing-argument]

separator_mess_files(), # type: ignore[missing-argument]

encoding_garbage_files(), # type: ignore[missing-argument]

wrong_count_files(), # type: ignore[call-arg]

unknown_word_files(), # type: ignore[call-arg]

separator_mess_files(), # type: ignore[call-arg]

encoding_garbage_files(), # type: ignore[call-arg]

Copilot · 2026-02-10T15:53:46Z

cardano_node_tests/tests/tests_conway/test_guardrails.py


        for subt in get_subtests():
-            with subtests.test(scenario=subt.__name__):
+            with subtests.test(scenario=getattr(subt, "__name__", "")):


Falling back to an empty string makes subtest output harder to interpret when __name__ is missing. Consider using a more informative fallback (e.g., repr(subt) or subt.__class__.__name__) so failures still include a meaningful scenario label. The same pattern appears in other tests updated in this PR.

Suggested change

with subtests.test(scenario=getattr(subt, "__name__", "")):

with subtests.test(scenario=getattr(subt, "__name__", repr(subt))):

Copilot · 2026-02-10T15:53:46Z

cardano_node_tests/tests/test_dbsync_config.py



-class TableCondition(str, enum.Enum):
+class TableCondition(enum.StrEnum):


enum.StrEnum is only available on Python 3.11+. If this project (or any of its supported environments) still runs on Python 3.10 or earlier, this will fail at import time. If older versions are supported, keep class X(str, enum.Enum) or add a compatibility shim/backport strategy.

Copilot · 2026-02-10T15:53:47Z

cardano_node_tests/tests/test_dbsync_config.py



-class ColumnCondition(str, enum.Enum):
+class ColumnCondition(enum.StrEnum):


enum.StrEnum is only available on Python 3.11+. If this project (or any of its supported environments) still runs on Python 3.10 or earlier, this will fail at import time. If older versions are supported, keep class X(str, enum.Enum) or add a compatibility shim/backport strategy.

cardano_node_tests/tests/conftest.py

mkoura requested review from ArturWieczorek and saratomaz as code owners February 10, 2026 15:48

mkoura requested review from Copilot and removed request for ArturWieczorek and saratomaz February 10, 2026 15:50

Copilot AI reviewed Feb 10, 2026

View reviewed changes

mkoura merged commit 6b98b33 into master Feb 10, 2026
8 of 9 checks passed

mkoura deleted the update_typechecking branch February 10, 2026 16:11

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix: improve type safety and error handling in tests #3326

fix: improve type safety and error handling in tests #3326

mkoura commented Feb 10, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Copilot AI Feb 10, 2026

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

	coverage_dict=cluster_obj.cli_coverage, # type: ignore[invalid-argument-type]
	coverage_dict=cluster_obj.cli_coverage, # type: ignore[arg-type]

	except psycopg2.errors.LockNotAvailable as e: # type: ignore[possibly-missing-attribute]
	except psycopg2.errors.LockNotAvailable as e: # type: ignore[attr-defined]

		tokens = draw(word_sequence(count=count, ensure_invalid=False)) # type: ignore[missing-argument]
		line = draw(join_with_weirdness(tokens=tokens)) # type: ignore[missing-argument]

	with subtests.test(scenario=getattr(subt, "__name__", "")):
	with subtests.test(scenario=getattr(subt, "__name__", repr(subt))):



		class TableCondition(str, enum.Enum):
		class TableCondition(enum.StrEnum):



		class ColumnCondition(str, enum.Enum):
		class ColumnCondition(enum.StrEnum):

fix: improve type safety and error handling in tests #3326

fix: improve type safety and error handling in tests #3326

Conversation

mkoura commented Feb 10, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 10, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant