fix: do not crash on unparseable cursor state value during concurrent cursor construction

[devin-ai-integration]

Summary

Fixes a hard source startup crash for oncall issue airbytehq/oncall#13084 (HubSpot), where the source dies before any stream is read with ValueError: No format in [...] matching True. Because the source dies at startup, no new state is ever committed — which is why manual state edits appeared to "revert" after each failed sync.

Root cause

HubSpot's associations_* streams (e.g. associations_contacts_companies) complete without a cursor value, so they persist the no-cursor sentinel state {"__ab_no_cursor_state_message": true}. On the next run the parent has incremental_dependency: true, so the CDK reconstructs parent state from the child state in ModelToComponentFactory._instantiate_parent_stream_state_manager. When no parent/global state is extracted, a fallback re-keyed the single child state value as the parent cursor:

cursor_values = child_state.values()          # -> [True]  (the sentinel value)
if cursor_values and len(cursor_values) == 1:
    ...
    stream_state=AirbyteStateBlob({cursor_field: list(cursor_values)[0]})  # -> {"lastmodifieddate": True}

That boolean then reached the datetime cursor parser during ConcurrentCursor construction and crashed the whole source.

Change (root fix)

In _instantiate_parent_stream_state_manager, do not treat the __ab_no_cursor_state_message sentinel (via the exported NO_CURSOR_STATE_KEY constant) as a cursor value. When the child state is only that sentinel, skip the re-keying and yield empty parent state so the parent starts cleanly from its configured start date:

if set(child_state.keys()) == {NO_CURSOR_STATE_KEY}:
    return ConnectorStateManager([])

Defense-in-depth (kept)

The earlier guard in DateTimeStreamStateConverter._get_sync_start is retained as a backstop: a boolean saved cursor value is never a valid datetime, so it is ignored (with a warning) and falls back to the start date rather than crashing. This guard is intentionally scoped to bool — a genuinely malformed datetime string must still raise, because StateDelegatingStream's api_retention_period cursor-age validation (#890) deliberately fails fast on unparseable cursor strings (test_cursor_age_validation_raises_error_for_unparseable_cursor).

No public interface, schema, spec, or state-format changes; non-breaking.

Declarative-First Evaluation

This is a CDK library fix in the shared declarative machinery (ModelToComponentFactory), not a connector-level change, so declarative-component alternatives do not apply.

Test Coverage

• Factory-level regression (test_instantiate_parent_stream_state_manager_ignores_no_cursor_sentinel): a parent ParentStreamConfig with incremental_dependency: true and a child state of {"__ab_no_cursor_state_message": True} now produces empty parent state and no longer raises ValueError: No format ... matching True.
• Converter-level (unit_tests/sources/streams/concurrent/test_datetime_state_converter.py): parametrized _get_sync_start boolean cases plus test_get_sync_start_with_boolean_state_value_does_not_raise and test_convert_from_sequential_state_with_boolean_state_value_does_not_raise.

All new tests fail without the corresponding fix and pass with it. test_model_to_component_factory.py (143 tests incl. test_state_delegating_stream.py) and the converter suite pass. ruff check, ruff format, and mypy are clean on the changed files.

Related to https://github.com/airbytehq/oncall/issues/13084:

• https://github.com/airbytehq/oncall/issues/13084

Note: this addresses the hard startup crash. The customer's original, softer complaint (slow syncs / repeated *_property_history re-emissions) is a separate performance concern tracked in the same issue.

Link to Devin session: https://app.devin.ai/sessions/ee9a064393c44676a341378e32c2a746

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring

[github-actions]

👋 Greetings, Airbyte Team Member!

Here are some helpful tips and reminders for your convenience.

💡 Show Tips and Tricks

Testing This CDK Version

You can test this version of the CDK using the following:

# Run the CLI from this branch:
uvx 'git+https://github.com/airbytehq/airbyte-python-cdk.git@devin/1783684033-fix-datetime-state-converter-bad-cursor-value#egg=airbyte-python-cdk[dev]' --help

# Update a connector to use the CDK from this branch ref:
cd airbyte-integrations/connectors/source-example
poe use-cdk-branch devin/1783684033-fix-datetime-state-converter-bad-cursor-value

PR Slash Commands

Airbyte Maintainers can execute the following slash commands on your PR:

• /autofix - Fixes most formatting and linting issues
• /poetry-lock - Updates poetry.lock file
• /test - Runs connector tests with the updated CDK
• /prerelease - Triggers a prerelease publish with default arguments
• /poe build - Regenerate git-committed build artifacts, such as the pydantic models which are generated from the manifest JSON schema in YAML.
• /poe <command> - Runs any poe command in the CDK environment

📚 Show Repo Guidance

Helpful Resources

• CDK API Reference

📝 Edit this welcome message.

[github-actions]

PyTest Results (Fast)

4 132 tests  +5   4 120 ✅ +5   7m 39s ⏱️ -2s
    1 suites ±0      12 💤 ±0 
    1 files   ±0       0 ❌ ±0 

Results for commit e6ff882. ± Comparison against base commit dd9558c.

♻️ This comment has been updated with latest results.

[github-actions]

PyTest Results (Full)

4 135 tests  +5   4 123 ✅ +5   11m 54s ⏱️ + 4m 39s
    1 suites ±0      12 💤 ±0 
    1 files   ±0       0 ❌ ±0 

Results for commit e6ff882. ± Comparison against base commit dd9558c.

♻️ This comment has been updated with latest results.