Module embed_struct 

Functions

basic_embedded_struct

Tests that embedded structs are registered in the app schema but don’t create their own database tables (they’re inlined into parent models).

create_and_query_embedded

Tests basic CRUD operations with embedded fields across all ID types. Validates create, read, update (both instance and query-based), and delete.

crud_nested_embedded

Tests CRUD operations with 2-level nested embedded structs. Validates that creating, reading, updating (instance and query-based), and deleting records with nested embedded structs works end-to-end.

deeply_nested_embedded_schema

Tests deeply nested embedded types (3+ levels) to verify schema building handles arbitrary nesting depth correctly. Validates:

embedded_struct_fields_codegen

Tests code generation for embedded struct field accessors:

embedded_struct_with_jiff_fields

Tests that jiff temporal types inside embedded structs round-trip correctly. Covers Timestamp (epoch nanos), civil::Date, civil::Time, and civil::DateTime.

embedded_struct_with_uuid_field

Tests that UUID fields inside embedded structs round-trip correctly. UUID requires a type cast on databases that don’t support it natively (e.g., SQLite stores it as text). This exercises the table_to_model lifting path for embedded struct fields with non-trivial type mappings.

partial_update_embedded_fields

Tests partial updates of embedded struct fields via stmt::patch / stmt::apply. This validates that individual fields within an embedded struct can be updated without replacing the entire struct.

partial_update_nested_embedded

Tests partial updates of deeply nested embedded fields via nested stmt::patch calls. Validates that patching a leaf field inside an outer embedded struct updates only that leaf, leaving all other fields unchanged in the database.

patch_data_enum_inside_embedded_struct

A data-carrying enum inside an embedded struct is replaced through a patch on the struct: the record value is substituted into every column of the nested enum (discriminant and variant fields), and the struct’s other fields are untouched.

query_based_partial_update_embedded

Tests partial updates of embedded fields using the query/filter-based path. User::filter_by_id(id).update().address(stmt::patch(...)) follows a different code path than the instance-based user.update().address(stmt::patch(...)), so both need coverage.

query_embedded_fields_comparison_ops

Tests comparison operators (gt, lt, ge, le, ne) on embedded struct fields. SQL-only: DynamoDB doesn’t support range queries on non-key attributes. Validates that all comparison operators work correctly with embedded fields.

query_embedded_multiple_fields

Tests querying by multiple embedded fields in a single query (AND conditions). SQL-only: DynamoDB requires partition key in queries. Validates that complex filters with multiple embedded fields work correctly.

query_embedded_struct_fields

Tests querying by embedded struct fields with composite keys (DynamoDB compatible). Validates:

root_model_with_embedded_field

Tests the complete schema generation and mapping for embedded fields:

unit_enum_in_embedded_struct

Tests a unit enum embedded as a field inside an embedded struct (enum-in-struct nesting). The struct flattens to columns including the enum’s discriminant column.

update_with_embedded_field_filter

Tests UPDATE operations filtered by embedded struct fields. SQL-only: DynamoDB requires partition key in queries/updates. Validates that updates can target rows based on embedded field values.

whole_embedded_struct_eq_ne

Tests eq and ne on a whole multi-field embedded value: both decompose into per-column comparisons — equality into AND, inequality into OR — so ne matches rows differing in any column.