Allow dynamic changes to multiselect options when `key` is provided by lukasmasuch · Pull Request #13448 · streamlit/streamlit

lukasmasuch · 2025-12-23T11:18:27Z

Describe your changes

Allow dynamically changing the options for st.multiselect without triggering an identity change / state reset. If the current selected options isn't in the list of available option, it will be reset to the default value.

GitHub Issue Link (if applicable)

Closes st.multiselect returns empty on dynamically changing options. #7855
Related issue Persist widget identity when key is set, even if the parameters change #11277

Testing Plan

Added e2e and unit tests.

Contribution License Agreement

By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license.

snyk-io · 2025-12-23T11:18:38Z

✅ Snyk checks have passed. No issues have been found so far.

Status	Scanner	Critical	High	Medium	Low	Total (0)
✅	Open Source Security	0	0	0	0	0 issues
✅	Licenses	0	0	0	0	0 issues

💻 Catch issues earlier using the plugins for VS Code, JetBrains IDEs, Visual Studio, and Eclipse.

github-actions · 2025-12-23T11:18:43Z

✅ PR preview is ready!

Name	Link
📦 Wheel file	https://core-previews.s3-us-west-2.amazonaws.com/pr-13448/streamlit-1.52.2-py3-none-any.whl
📦 `@streamlit/component-v2-lib`	Download from artifacts
🕹️ Preview app	pr-13448.streamlit.app (☁️ Deploy here if not accessible)

lib/tests/streamlit/elements/multiselect_test.py

…multiselect-options

Copilot

Pull request overview

This pull request enables dynamic changes to st.multiselect options when a key is provided, maintaining widget identity and preserving valid selections while filtering out invalid ones when the options list changes.

Key changes:

Removed options and format_func from the identity computation when a key is provided, allowing these to change without resetting widget state
Added validation logic to filter multiselect values against current options, preserving valid selections and removing invalid ones
Refactored selectbox validation logic into reusable utility functions

Reviewed changes

Copilot reviewed 7 out of 10 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
`lib/streamlit/elements/widgets/multiselect.py`	Modified identity computation to exclude `options` and `format_func` from whitelisted params; added validation logic to filter invalid selections when options change
`lib/streamlit/elements/widgets/selectbox.py`	Refactored validation logic to use new `validate_value_against_options` utility function
`lib/streamlit/elements/lib/options_selector_utils.py`	Added two new validation functions: `validate_value_against_options` for selectbox and `validate_multiselect_value_against_options` for multiselect
`lib/tests/streamlit/elements/multiselect_test.py`	Updated stable ID tests to reflect new whitelisted params; added three new AppTest tests covering option expansion, filtering, and complete removal scenarios
`lib/tests/streamlit/elements/lib/options_selector_utils_test.py`	Added comprehensive parameterized tests for both new validation utility functions
`e2e_playwright/st_multiselect_test.py`	Enhanced `test_dynamic_multiselect_props` to test both filtering and preservation behaviors with improved documentation
`e2e_playwright/st_multiselect.py`	Updated test app to change both options and format_func between initial and updated states; changed `accept_new_options` to `False` consistently to avoid identity changes

lib/streamlit/elements/widgets/multiselect.py

lib/tests/streamlit/elements/multiselect_test.py

lukasmasuch · 2026-01-05T19:31:40Z

@cursor review

cursor

✅ Bugbot reviewed your changes and found no bugs!

…multiselect-options

github-actions · 2026-01-05T23:22:46Z

Summary

This PR enables dynamic changes to st.multiselect options when a key is provided, without resetting the widget state entirely. When options change dynamically:

Valid selections are preserved: Values that exist in both old and new option sets are kept
Invalid selections are filtered out: Values no longer in the new options are removed
The widget identity remains stable (no re-mount/state loss)

The implementation introduces two new validation utility functions that are shared between selectbox and multiselect widgets, improving code reusability.

Key Changes:

Added validate_value_against_options() and validate_multiselect_value_against_options() in options_selector_utils.py
Updated multiselect.py to use the new validation when accept_new_options=False
Refactored selectbox.py to use the shared validation function
Removed options and format_func from the key_as_main_identity set for multiselect

Code Quality

Strengths:

Clean separation of concerns with validation logic extracted to reusable utility functions
Well-documented functions with numpydoc-style docstrings
Proper type annotations throughout
Code follows existing patterns in the codebase
Good use of existing index_() function for value comparison (handles float epsilon)
Appropriate use of reset_state_value() to avoid the "cannot be modified after widget instantiated" error

Minor Observations:

options_selector_utils.py:367 - The # noqa: PERF203 comment acknowledges a potential performance concern with exceptions in loops. This is acceptable for typical use cases with small option sets, but worth noting for users with large selections.
multiselect.py:552-565 - The logic correctly skips validation when accept_new_options=True, preserving user-entered values that aren't in the original options list.
selectbox.py:599-609 - Good refactoring to use the shared validation function, reducing code duplication.

Test Coverage

Unit Tests:

✅ Comprehensive parameterized tests for validate_value_against_options() covering:
- Value in options (no reset)
- None value handling
- Value not in options (reset to default)
- Custom default index
- None default index
- Empty options
- Numeric and float values
✅ Comprehensive parameterized tests for validate_multiselect_value_against_options() covering:
- All values in options
- Empty values list
- Some values not in options (partial filtering)
- All values not in options (complete reset to empty)
- Numeric value types
✅ Three new AppTest integration tests:
- test_multiselect_preserves_selection_when_options_expand
- test_multiselect_filters_invalid_selections
- test_multiselect_resets_when_all_selections_removed

E2E Tests:

✅ Updated test_dynamic_multiselect_props with clear documentation of test scenarios
✅ Tests selection reset when values are removed from options
✅ Tests selection preservation when values exist in both option sets
✅ Updated snapshots for the new behavior

Test Quality Notes:

E2E tests follow best practices: use expect for auto-wait, label-based locators, and helper utilities
Tests have descriptive docstrings explaining the test scenarios
Good use of select_for_multiselect and expect_prefixed_markdown utilities

Backwards Compatibility

✅ This change is backwards compatible.

Existing apps using st.multiselect with a key will experience the new filtering behavior, which is an improvement (previously, dynamic option changes would cause state issues)
Apps without a key continue to work as before (widget identity is content-based)
The accept_new_options=True path is unchanged - user-entered values are still preserved
The validation only kicks in when options change AND the widget has an existing value from session state

Migration Note: Users relying on the previous behavior where changing options would reset state might see different behavior now - selections that remain valid will be preserved. This is the intended fix for issues #7855 and #11277.

Security & Risk

✅ No security concerns identified.

No new external inputs or user data handling
The validation functions only compare values against provided options
Session state updates use the existing reset_state_value() API designed for this purpose

Risk Assessment:

Low risk: The changes are localized to option validation logic
The implementation gracefully handles edge cases (empty options, None values, empty selections)
The existing index_() function handles float comparison correctly

Recommendations

The PR is well-implemented and ready for merge. A few minor suggestions for consideration:

Consider adding a note about format_func in the documentation: The E2E test comments explain that changing format_func may cause selections to be filtered if the formatted label no longer matches. This behavior might be worth documenting for users.
Test coverage for session state key update: While the logic updates session state when key is not None, the unit tests pass None for the key parameter. Consider adding a test that verifies session state is actually updated when a key is provided.
Consider edge case: When format_func changes between renders, the validation compares raw values (not formatted), which is correct. However, if users have complex objects as options, they should ensure their objects implement proper equality. This is existing behavior but worth noting.

Verdict

APPROVED: This PR provides a well-tested, clean implementation that fixes a real user pain point (#7855, #11277). The code follows existing patterns, has comprehensive test coverage, and is backwards compatible. The validation logic is correctly shared between selectbox and multiselect widgets.

This is an automated AI review. Please verify the feedback and use your judgment.

lib/tests/streamlit/elements/multiselect_test.py

sfc-gh-bnisco · 2026-01-07T19:47:34Z

lib/streamlit/elements/lib/options_selector_utils.py

    )
+
+
+def validate_value_against_options(


suggestion (non-blocking): Because these functions validate and also have a possible side-effect of changing state, consider either:

Giving them either a name that makes the potential for side-effects clearer.

Refactoring this into 2 different functions and have the widget layer compose the functions so that we can keep validation logic pure.

Renamed the function to make it a bit more explicit

github-actions · 2026-01-09T14:09:55Z

📉 Frontend coverage change detected

The frontend unit test (vitest) coverage has decreased by 0.0400%

Current PR: 86.5200% (12812 lines, 1727 missed)
Latest develop: 86.5600% (12812 lines, 1721 missed)

💡 Consider adding more unit tests to maintain or improve coverage.

📊 View detailed coverage comparison

…#13646) When using st.multiselect() with custom class objects and a format_func, selections would be cleared after each script rerun in v1.53.0. This was a regression introduced by PR #13448. The issue occurred because register_widget() deepcopies widget values, and the validation logic used == comparison which falls back to identity checks for objects without __eq__. Deepcopied instances fail identity comparison, causing valid selections to be filtered out. The fix changes the validation to use format_func() for comparisons instead, comparing by formatted string representation. This is more robust and works correctly with custom objects regardless of __eq__ implementation. The solution also includes error handling for edge cases where format_func fails on incompatible value types. - Modified validate_and_sync_multiselect_value_with_options() to accept and use format_func parameter - Added comprehensive unit tests for custom objects with and without __eq__ - Added edge case test for format_func failures - Added E2E test to verify selections persist across script reruns Co-Authored-By: Claude <[email protected]>

…13648) ## Describe your changes Fixes issue #13646 where `st.multiselect()` and `st.selectbox` clears user selections after each script rerun when using custom class objects without `__eq__` implementation and a `format_func` parameter. This was a regression introduced in PR #13448. **Root Cause**: The validation logic compared multiselect values using `==`, which falls back to identity comparison for objects without `__eq__`. Since `register_widget()` deepcopies widget values, the new instances fail identity comparison with original options, causing valid selections to be filtered out. **Solution**: Changed validation to compare values using `format_func()` by their formatted string representation instead of using `==`. This is more robust and works correctly regardless of whether custom classes implement `__eq__`. ## Testing Plan - Unit Tests: Added 4 comprehensive test cases covering custom objects without `__eq__`, partial matches, objects with `__str__`, and edge cases where `format_func` fails on incompatible types - E2E Tests: Added `test_multiselect_custom_objects_without_eq` to verify selections persist across script reruns with custom class objects - Backward Compatibility: The `format_func` parameter has a default value of `str`, so existing code continues to work without changes --- **Contribution License Agreement** By submitting this pull request you agree that all contributions to this project are made under the Apache 2.0 license. --------- Co-authored-by: Claude <[email protected]>

Allow dynamic changes to multiselect options when key is provided

27e1732

sfc-gh-lmasuch added security-assessment-completed change:feature PR contains new feature or enhancement implementation impact:users PR changes affect end users labels Dec 23, 2025

github-code-quality bot found potential problems Dec 23, 2025

View reviewed changes

lib/tests/streamlit/elements/multiselect_test.py Fixed Show fixed Hide fixed

lib/tests/streamlit/elements/multiselect_test.py Fixed Show fixed Hide fixed

lib/tests/streamlit/elements/multiselect_test.py Fixed Show fixed Hide fixed

Merge remote-tracking branch 'upstream/develop' into feature/dynamic-…

a8fa2c8

…multiselect-options

lukasmasuch changed the title ~~Allow dynamic changes to multiselect options when key is provided~~ Allow dynamic changes to multiselect options when key is provided Jan 5, 2026

lukasmasuch marked this pull request as ready for review January 5, 2026 19:17

Copilot AI review requested due to automatic review settings January 5, 2026 19:17

Copilot started reviewing on behalf of lukasmasuch January 5, 2026 19:18 View session

Update snapshots

39fec89

Copilot AI reviewed Jan 5, 2026

View reviewed changes

Update

a400cd9

cursor bot reviewed Jan 5, 2026

View reviewed changes

lukasmasuch added 5 commits January 5, 2026 22:49

Update

c10f424

Update

371525f

Update

4f44de0

Fix

1a7a167

Merge remote-tracking branch 'upstream/develop' into feature/dynamic-…

1cf4f0e

…multiselect-options

lukasmasuch added the ai-review If applied to PR or issue will run AI review workflow label Jan 5, 2026

github-actions bot removed the ai-review If applied to PR or issue will run AI review workflow label Jan 5, 2026

Fix e2e test

10ff621

github-code-quality bot found potential problems Jan 5, 2026

View reviewed changes

lib/tests/streamlit/elements/multiselect_test.py Dismissed Show dismissed Hide dismissed

lib/tests/streamlit/elements/multiselect_test.py Dismissed Show dismissed Hide dismissed

lib/tests/streamlit/elements/multiselect_test.py Dismissed Show dismissed Hide dismissed

Add test

f5f105a

github-code-quality bot found potential problems Jan 5, 2026

View reviewed changes

lib/tests/streamlit/elements/multiselect_test.py Dismissed Show dismissed Hide dismissed

sfc-gh-bnisco approved these changes Jan 7, 2026

View reviewed changes

Update

54ae829

lukasmasuch enabled auto-merge (squash) January 9, 2026 14:00

Merge branch 'develop' into feature/dynamic-multiselect-options

7647932

lukasmasuch merged commit 4eaf69f into develop Jan 9, 2026
44 checks passed

lukasmasuch deleted the feature/dynamic-multiselect-options branch January 9, 2026 15:03

github-actions bot mentioned this pull request Jan 20, 2026

Multiselect with format func getting cleared right after selecting an element since 1.53.0 #13646

Closed

4 tasks

lukasmasuch mentioned this pull request Jan 20, 2026

Fix selectbox and multiselect clearing selections for custom objects #13648

Merged

		)


		def validate_value_against_options(

Conversation

lukasmasuch commented Dec 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Describe your changes

GitHub Issue Link (if applicable)

Testing Plan

Uh oh!

snyk-io bot commented Dec 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ Snyk checks have passed. No issues have been found so far.

Uh oh!

github-actions bot commented Dec 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ PR preview is ready!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

lukasmasuch commented Jan 5, 2026

Uh oh!

cursor bot left a comment

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Jan 5, 2026

Summary

Code Quality

Test Coverage

Backwards Compatibility

Security & Risk

Recommendations

Verdict

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

sfc-gh-bnisco Jan 7, 2026

Choose a reason for hiding this comment

Uh oh!

lukasmasuch Jan 9, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Jan 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📉 Frontend coverage change detected

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

lukasmasuch commented Dec 23, 2025 •

edited

Loading

snyk-io bot commented Dec 23, 2025 •

edited

Loading

github-actions bot commented Dec 23, 2025 •

edited

Loading

github-actions bot commented Jan 9, 2026 •

edited

Loading