[sebxt]

Summary

_from_iso_format_to_datetime appended +00:00 to any string without a + character, so ISO 8601 timestamps with a negative offset (2021-07-14T03:00:17-05:00) were turned into invalid strings and raised ValueError. A Z suffix had the same problem on Python 3.11+ (on 3.7 to 3.10 it only parsed by accident). This affects Installation, Bot and anything else that goes through _timestamp_to_type with a string value.

The function now swaps a trailing Z or z for +00:00 (only the last character, per RFC 3339 §5.6), parses the string, and only sets UTC when the result has no tzinfo, the same approach #1970 took for naive datetime values. Strings without an offset are still treated as UTC, so existing behavior is unchanged for them.

Testing

./scripts/run_tests.sh tests/slack_sdk/oauth/installation_store/test_internals.py

New cases cover a positive offset, a negative offset, and the Z/z UTC designator, directly and through _timestamp_to_type. The negative offset cases fail on main and pass with this change; the UTC designator cases pass on Python 3.10 and 3.13.

Category

• slack_sdk.web.WebClient (sync/async) (Web API client)
• slack_sdk.webhook.WebhookClient (sync/async) (Incoming Webhook, response_url sender)
• slack_sdk.socket_mode (Socket Mode client)
• slack_sdk.signature (Request Signature Verifier)
• slack_sdk.oauth (OAuth Flow Utilities)
• slack_sdk.models (UI component builders)
• slack_sdk.scim (SCIM API client)
• slack_sdk.audit_logs (Audit Logs API client)
• slack_sdk.rtm_v2 (RTM client)
• /docs (Documents)
• /tutorial (PythOnBoardingBot tutorial)
• tests/integration_tests (Automated tests for this library)

Requirements

• I've read and understood the Contributing Guidelines and have done my best effort to follow them.
• I've read and agree to the Code of Conduct.
• I've run python3 -m venv .venv && source .venv/bin/activate && ./scripts/run_validation.sh after making the changes.

Fixes #1989

[salesforce-cla]

Thanks for the contribution! Before we can merge this, we need @auraseb to sign the Salesforce Inc. Contributor License Agreement.

[WilliamBergamin]

Hello 👋 Please sign the CLA

[WilliamBergamin]

Hello and thanks for your interest in our project 💯 The negative offset fix is real: I checked that the new tests fail on main and pass on this branch

There's one regression to fix before merging. On CPython 3.7-3.10, a timestamp ending in Z (e.g. "2021-07-14T08:00:17Z") parses on main and raises ValueError with this change. main only handles it by accident: it appends +00:00, and the C fromisoformat on those versions lets the stray Z through. With the string passed to fromisoformat as-is, nothing before 3.11 accepts Z (gh-80010). Since the SDK supports 3.7+ and both #1989 and the description say Z is fixed, I think we should handle it explicitly. Suggestions inline.

Also, the CLA check needs signing before we can run CI.

[WilliamBergamin]

Swap a trailing Z for +00:00 before parsing so Z works on every supported version of python. Only the last character, not str.replace: RFC 3339 §5.6 only allows Z as the final offset, which is also how CPython's own parser treats it. Lowercase z is allowed by the RFC and already parses on main under CPython, so keeping it avoids a second regression.

Suggested change

	def _from_iso_format_to_datetime(iso_datetime_str: str) -> datetime:
	if "+" not in iso_datetime_str:
	iso_datetime_str += "+00:00"
	return datetime.fromisoformat(iso_datetime_str)
	dt = datetime.fromisoformat(iso_datetime_str)
	if dt.tzinfo is None:
	dt = dt.replace(tzinfo=timezone.utc)
	return dt
	def _replace_utc_designator(iso_datetime_str: str) -> str:
	if iso_datetime_str.endswith(("Z", "z")):
	return iso_datetime_str[:-1] + "+00:00"
	return iso_datetime_str
	def _from_iso_format_to_datetime(iso_datetime_str: str) -> datetime:
	dt = datetime.fromisoformat(_replace_utc_designator(iso_datetime_str))
	if dt.tzinfo is None:
	dt = dt.replace(tzinfo=timezone.utc)
	return dt

[WilliamBergamin]

Could we add coverage for the UTC designator? No test exercises Z right now, which is why the 3.7-3.10 regression doesn't show up in CI.

    def test_iso_format_with_utc_designator(self):
        for iso_datetime_str in ["2021-07-14T08:00:17Z", "2021-07-14T08:00:17z", "2021-07-14T08:00:17.000Z"]:
            dt = _from_iso_format_to_datetime(iso_datetime_str)
            self.assertEqual(dt.timestamp(), 1626249617.0)

and a ("2023-11-28T22:11:19Z", int, 1701209479) case in the test_timestamp_to_type parametrize list.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.36%. Comparing base (ab4e498) to head (6bd1361).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1990   +/-   ##
=======================================
  Coverage   84.35%   84.36%           
=======================================
  Files         118      118           
  Lines       13793    13798    +5     
=======================================
+ Hits        11635    11640    +5     
  Misses       2158     2158           

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

[sebxt]

Thanks for catching the 3.7 to 3.10 regression. Applied your suggestion as written, added test_iso_format_with_utc_designator and the trailing Z case in test_timestamp_to_type. test_internals.py passes 16 of 16 on 3.10 and 3.13.

[hello-ashleyintech]

hi @auraseb! Thanks so much for your contribution, could you sign the Salesforce Inc. Contributor License Agreement?