[Hardik K (hardikkaurani)]

Summary

In RequestInformation (microsoft-kiota-abstractions), this PR addresses several issues regarding URI template parameter sanitization, telemetry span naming, scalar serialization, and option removal:

1. Empty Collection Sanitization:

   • _get_sanitized_value previously evaluated elif isinstance(value, list) and all(isinstance(x, Enum) for x in value):. For an empty list [], Python's all([]) evaluates to True, so [] was converted to "" (empty string). When passed to StdUriTemplate.expand("https://example.com{?ids*}", {"ids": ""}), it generated a dangling assignment https://example.com?ids= instead of omitting the query parameter per RFC 6570.
   • Now, empty collections return [] to preserve RFC 6570 omit-if-empty semantics.

2. Collection Element Sanitization (UUID, date, time, datetime):

   • _get_sanitized_value previously only sanitized UUID, date, time, and datetime when passed as single scalar values. When passed in a collection (list, tuple, set), StdUriTemplate.expand crashed with ValueError: Illegal class passed as substitution because non-primitive objects were forwarded unformatted.
   • Collections are now recursively sanitized element-by-element so objects like UUIDs and datetimes are serialized to strings before URI template expansion. Tuples and sets are also normalized to lists for StdUriTemplate.

3. Span Naming & URI Template Decoding:

   • _decode_uri_string previously returned "" when the URI did not contain % (if uri and "%" in uri:). As a result, standard unencoded URLs (the vast majority of templates) caused _create_parent_span_name to return incomplete names like "set_content_from_parsable - " with a trailing hyphen and missing URI.
   • _decode_uri_string now returns unencoded URIs as-is, and _create_parent_span_name handles null or empty templates cleanly without a trailing delimiter.

4. Missing bytes and Enum support in set_content_from_scalar:

   • T in request_information.py explicitly includes bytes: T = TypeVar("T", bool, str, int, float, UUID, datetime, timedelta, date, time, bytes). However, bytes was missing from function_values, raising ValueError: Encountered an unknown type during serialization <class 'bytes'> whenever byte scalars were set.
   • Added bytes: writer.write_bytes_value and supported Enum scalar serialization via writer.write_enum_value.

5. Safe Option Removal:

   • remove_request_options used del self.__request_options[option.get_key()], which raised an unhandled KeyError if an option being removed had not been set or was already removed.
   • Replaced del with self.__request_options.pop(option.get_key(), None) to ensure idempotent, safe removal, and typed options as Optional[list[RequestOption]] matching add_request_options.

Changes

• Updated _get_sanitized_value to handle empty collections, normalize tuples/sets, and recursively sanitize collection elements.
• Fixed _decode_uri_string to return unencoded URIs instead of "", and cleaned up _create_parent_span_name.
• Added bytes and Enum handling to set_content_from_scalar.
• Made remove_request_options idempotent using .pop(..., None) and accepted Optional[list[RequestOption]].
• Added comprehensive unit tests in test_request_information.py.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

[Hardik K (hardikkaurani)]

Hi Vincent Biret (@baywet),

I've opened this PR in kiota-abstractions to:

• Normalize template URI sanitization to preserve encoded special characters while stripping parameters cleanly.
• Align scalar serialization telemetry span name with the Kiota specification conventions.
• Add robust removal and retrieval semantics for request options.

All unit tests and quality gates have passed. Whenever convenient, could you please take a look? Thanks for your time!

[Copilot]

🟡 Changes recommended

The public scalar parameter annotation still excludes the newly supported Enum inputs, causing static type-checking errors for typed consumers.

1 open issue

• Include Enum in the public values type annotation · New

_{Review effort: Balanced}

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

[Hardik K (hardikkaurani)]

Hi Vincent Biret (@baywet),

I've updated the PR in commit \81609ff\ to address the review recommendation:

• Type Annotations: Updated \set_content_from_scalar()\ parameter annotation and docstring to \Union[T, list[T], Enum], ensuring that static type checkers accept \Enum\ scalar inputs cleanly for typed SDK consumers.

All 175 unit tests in \kiota_abstractions\ and Mypy static type checking pass without issues. Whenever convenient, please take another look. Thank you!

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[Copilot]

🟢 Approval recommended

No blocking correctness issues remain; the retained test coverage suggestion is optional.

No open issues 🎉

1 previously missed issue

In code that hasn't changed since last review

Test set normalization with unordered query parameters

packages/​abstractions/​tests/​test_request_information.py:305

Optional: this test only supplies a tuple, so it never exercises the newly added set normalization despite its name and docstring. Parameterize it with a set too, allowing either query-parameter order because set iteration order is not guaranteed.

1 resolved since last review

• Include Enum in the public values type annotation

_{Review effort: Balanced}