v0.1.0

First public release on crates.io. dynamodb-facade is a typed facade over aws-sdk-dynamodb that replaces string-spliced expressions, hand-built key maps, pagination loops, and 25-item batch chunking with composable, compile-time-checked Rust.

Added

• Expression builders, not strings. Condition::eq, Condition::exists, Condition::begins_with, Condition::between, Condition::is_in, Condition::size_cmp, variadic Condition::and / Condition::or, combined with &, |, !. Update::set, Update::remove, Update::increment, Update::decrement, Update::init_increment, Update::init_decrement, Update::set_custom (with UpdateSetRhs), Update::list_append, Update::list_prepend, Update::add, Update::delete, Update::and, Update::combine, Update::try_combine. KeyCondition::pk(...) with .sk_eq, .sk_begins_with, .sk_between, .sk_lt/.sk_le/.sk_gt/ .sk_ge. Every #name / :value placeholder is managed internally.
• Zero-sized schema types. attribute_definitions!, table_definitions! and index_definitions! declarative macros generate marker types encoding the table / index key shape. Simple-key and composite-key tables and indexes are both supported. Using a sort-key method on a simple-key schema is a compile error.
• Typed items via dynamodb_item!. Wires a user struct to a table with constant or variable PK / SK, a type discriminator attribute, marker-only attributes (e.g. GSI partition keys that serde already serialises), and delegation of key construction to another type. has_attributes! macro for hand-rolled DynamoDBItem implementations.
• Typestate operation builders. get, put, delete, update, query, scan and their _by_id / instance variants. Three orthogonal typestate dimensions: output format (Typed ↔ Raw), return value (ReturnNothing ↔ Return<Old> ↔ Return<New>), and expression-set state (NoCondition / NoFilter / NoProjection ↔ AlreadyHas*). Calling .condition(), .filter() or .project() twice is a compile error. .exists() / .not_exists() shorthands derived from the item's PK. .raw() escape hatch to an untyped Item<TD>.
• Single-table friendly. First-class support for the mono-table pattern (PK + SK + type discriminator). Item<TD>::attribute::<ItemType>() for runtime dispatch on scan/query results, T::from_item / T::try_from_item for typed reconstruction, Item::into_key_only / Item::minimal_from / Item::with_attributes / Item::extract_key / Item::from_key_and_attributes for key-level manipulation.
• Automatic pagination. .all() collects all pages into a Vec, .stream() yields an impl Stream<Item = Result<T>> (or Item<TD> in Raw mode). No ExclusiveStartKey bookkeeping. Works on both query and scan. query_all / query_all_index for types with a constant partition key.
• Index queries. T::query_index::<Idx>(client, key_cond) and T::index_key_condition::<Idx>(id) for typed index queries, QueryRequest::new_index::<Idx>(...) for raw-level index queries when the entity type does not match the index's primary attribute.
• Automatic batch chunking and retry. dynamodb_batch_write splits Vec<WriteRequest> into 25-item batches, runs them in parallel, and retries UnprocessedItems with backoff (up to 5 attempts). T::batch_put, T::batch_delete, T::batch_delete_by_id, and the free batch_delete helper produce the individual requests. Mixed put + delete batches supported.
• Typed transactions. T::transact_put, T::transact_delete, T::transact_update / transact_update_by_id, T::transact_condition and T::transact_delete_by_id build a TransactWriteItem with the same condition DSL as stand-alone operations, then .build() plugs into the SDK's transact_write_items() fluent builder.
• Flexible serialisation. Items round-trip through serde_dynamo by default; DynamoDBItem<TD> can be hand-implemented when serde is not a good fit (e.g. enum stored as a single attribute). The TD table parameter is generic, so a single struct may implement DynamoDBItem for multiple tables (useful for migrations).
• Typed values. IntoAttributeValue trait for domain newtypes, with implementations for the standard scalar types, AsSet<T> for DynamoDB set types (SS/NS/BS), AsNumber<T> for numeric-flavoured attributes.
• Unified error type. Error enum with DynamoDB, Serde, FailedBatchWrite, Other, Custom variants and a Result<T> alias. Error::as_dynamodb_error() downcasts for matching specific SDK errors (e.g. ConditionalCheckFailedException). Error::custom / Error::other constructors.
• Escape hatch preserved. Every builder exposes .into_inner() returning the underlying aws_sdk_dynamodb fluent builder. The crate re-exports aws_sdk_dynamodb::{Client, Error as DynamoDBError, types::AttributeValue} so downstream code does not need to pin the SDK version separately.
• MSRV 1.85.0, edition 2024. tokio / futures-based async, identical runtime assumptions to aws-sdk-dynamodb itself.
• Feature flags. test-fixtures exposes the shared PlatformTable / User / Enrollment / TypeIndex / EmailIndex domain types used in doc examples. integration gates the Docker-backed end-to-end test suite against DynamoDB Local.
• Documentation. Full API documentation on docs.rs, a README.md with a side-by-side raw-SDK-vs-facade comparison, and EXAMPLES.md — a 13-section tour across a single-table domain covering schema design, every CRUD variant, index queries, scans with dispatch, the full condition and update DSL, batch writes, and transactions.