fix(util): guard _datetime_from_weaviate_str against empty string by devteamaegis · Pull Request #2053 · weaviate/weaviate-python-client

devteamaegis · 2026-06-01T19:07:23Z

What's broken

_datetime_from_weaviate_str in weaviate/util.py evaluates string[-1] on line 754 without checking whether string is empty. Python's negative indexing on an empty string raises IndexError. The protobuf wire-format default for a string field is "", so any gRPC query response containing an object with an unset date property crashes the client during deserialization in base_executor.py.

Traceback: IndexError: string index out of range at weaviate/util.py line 754.

Why it happens

The function assumes the input is a non-empty datetime string and immediately indexes from the end, which is invalid for the empty-string protobuf default.

Fix

Added a two-line guard at the top of _datetime_from_weaviate_str: if string is empty, return datetime.min. This matches the existing behaviour for year-zero dates and avoids any semantic change for valid inputs.

Test

Added ("", datetime.min) as a parametrized case in the existing test_datetime_from_weaviate_str test in test/test_util.py. All 40 tests pass.

Fixes #2052

Empty string is the protobuf wire-format default for an unset string field. Calling string[-1] on "" raises IndexError, crashing gRPC deserialization for any object with an unset date property. Return datetime.min for empty input, matching the existing behaviour for year-zero dates. Fixes weaviate#2052

orca-security-eu

Orca Security Scan Summary

Status	Check	Issues by priority
Passed	Infrastructure as Code	0 0 0 0	View in Orca
Passed	SAST	0 0 0 0	View in Orca
Passed	Secrets	0 0 0 0	View in Orca
Passed	Vulnerabilities	0 0 0 0	View in Orca

weaviate-git-bot · 2026-06-01T21:00:41Z

To avoid any confusion in the future about your contribution to Weaviate, we work with a Contributor License Agreement. If you agree, you can simply add a comment to this PR that you agree with the CLA so that we can merge.

beep boop - the Weaviate bot 👋🤖

PS:
Are you already a member of the Weaviate Forum?

dirkkul · 2026-06-02T06:37:04Z


 def _datetime_from_weaviate_str(string: str) -> datetime.datetime:
+    if not string:
+        return datetime.datetime.min


Wouldn't "None" be a better return here?

If we have properties I would expect None if it is unset in Weaviate

@dirkkul

Per review (@dirkkul): an unset date property should surface as None, not a datetime.min sentinel. Empty-string input now returns None; year-0 out-of-range handling is unchanged. Callers propagate the value as a property (typed Any / Optional[List[Any]]) so None flows through.

devteamaegis · 2026-07-26T16:10:08Z

Good call @dirkkul — changed it to return None for an unset/empty datetime instead of datetime.min. That matches the "unset property should be None" expectation, and it flows through safely: both callers in base_executor.py propagate the value as a property (Any / Optional[List[Any]]), neither does date math on it. Updated the return type to Optional[datetime.datetime] and the test case to expect None. Left the year-0 → datetime.min path untouched since that's separate existing behavior with its own warning.

Copilot

🟡 Changes recommended

The implementation/test behavior for empty strings (None) conflicts with the PR description’s stated behavior (datetime.min), so the intended semantics need to be made consistent.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

This PR hardens datetime deserialization in weaviate/util.py to prevent an IndexError when Weaviate returns an unset date property as the protobuf default empty string (""), and extends the existing unit test coverage for this edge case.

Changes:

Add an early guard in _datetime_from_weaviate_str for empty strings.
Adjust _datetime_from_weaviate_str’s return annotation and behavior to allow an “empty value” outcome.
Extend test_datetime_from_weaviate_str with an empty-string parameterized case and update expected type accordingly.

File summaries

File	Description
`weaviate/util.py`	Adds an empty-string guard and changes the helper’s return contract to allow a sentinel for unset date values.
`test/test_util.py`	Adds a test case for `""` and updates the expected type to match the new behavior.

Review details

Files reviewed: 2/2 changed files
Comments generated: 1
Review effort level: Lite

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

+def _datetime_from_weaviate_str(string: str) -> Optional[datetime.datetime]:
+    if not string:
+        return None


g-despot · 2026-09-07T11:47:51Z

Hi @devteamaegis! 😄
Could you please write a comment that you agree to our CLA so we can continue with this PR?

devteamaegis · 2026-09-07T17:54:26Z

Hi @g-despot! I have read the Weaviate Contributor License Agreement and I hereby agree to its terms. Also pushed a small ruff format fix so the linter check passes — should be good to continue now. Thanks!

…ently An unset date property arrives as null_value and never reaches this path, so an empty string can only be a malformed value from the server. Emit Con006 rather than dropping it quietly, and cover the scalar and array paths against a mock gRPC response.

g-despot · 2026-09-08T08:22:26Z

Thanks a lot for fixing the issue and kudos on your first Weaviate contribution! 😄

orca-security-eu Bot reviewed Jun 1, 2026

View reviewed changes

dirkkul reviewed Jun 2, 2026

View reviewed changes

devteamaegis force-pushed the fix/indexerror-empty-datetime-str branch from bcfe5e9 to 6987d13 Compare June 11, 2026 19:46

g-despot requested a lite review from Copilot September 7, 2026 11:42

Copilot started reviewing on behalf of g-despot September 7, 2026 11:43 View session

Copilot AI reviewed Sep 7, 2026

View reviewed changes

Comment thread weaviate/util.py

Comment on lines +753 to +755

def _datetime_from_weaviate_str(string: str) -> Optional[datetime.datetime]:

if not string:

return None

g-despot and others added 2 commits September 7, 2026 13:48

Merge branch 'main' into fix/indexerror-empty-datetime-str

c9526e9

Format test with ruff (line-length 100)

f167516

g-despot merged commit f336b6d into weaviate:main Sep 8, 2026
247 of 249 checks passed

Navigation Menu

Uh oh!

fix(util): guard _datetime_from_weaviate_str against empty string - #2053

fix(util): guard _datetime_from_weaviate_str against empty string#2053
g-despot merged 5 commits into
weaviate:mainfrom
devteamaegis:fix/indexerror-empty-datetime-str

devteamaegis commented Jun 1, 2026

Uh oh!

orca-security-eu Bot left a comment •

edited

Loading

Uh oh!

weaviate-git-bot commented Jun 1, 2026

Uh oh!

dirkkul Jun 2, 2026

Uh oh!

devteamaegis commented Jul 26, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

g-despot commented Sep 7, 2026

Uh oh!

devteamaegis commented Sep 7, 2026

Uh oh!

Uh oh!

g-despot commented Sep 8, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Uh oh!

Conversation

devteamaegis commented Jun 1, 2026

What's broken

Why it happens

Fix

Test

Uh oh!

orca-security-eu Bot left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Orca Security Scan Summary

Uh oh!

weaviate-git-bot commented Jun 1, 2026

Uh oh!

dirkkul Jun 2, 2026

Choose a reason for hiding this comment

Uh oh!

devteamaegis commented Jul 26, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Uh oh!

g-despot commented Sep 7, 2026

Uh oh!

devteamaegis commented Sep 7, 2026

Uh oh!

Uh oh!

g-despot commented Sep 8, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

orca-security-eu Bot left a comment •

edited

Loading