daredis 0.1.0

A Redis client for Dart with connection pooling, cluster support, dedicated Pub/Sub and transaction sessions, and typed helper APIs.

0.1.0

Breaking changes

• Removed DaredisCluster.clientPoolSize. Cluster clients now use one slot-aware router with per-node pools, so nodePoolSize is the concurrency knob to configure.
• Changed pipeline execution to run on a single connection. On Redis Cluster, all keyed commands in one pipeline must now route to the same node.
• Made RedisPubSub.close() terminal. Closing a pub/sub session now completes its message stream and the same session cannot be reopened.
• Made RedisTransaction and RedisClusterTransaction single-use after close(). Open a fresh session with openTransaction() for the next transaction.

Improvements

• Tightened low-level connection lifecycle handling around idempotent connect(), manual disconnect, and auth error propagation.

Docs

• Refreshed README, examples, and package-level docs to match the updated client and session semantics.

0.0.5

• Added slot-routed Redis Cluster transactions via openTransaction(String routingKey).
• Extracted cluster command policy and redirect helpers into dedicated internal modules.
• Clarified native Redis Cluster multi-key rules in documentation and tests.

0.0.4

• Added slot-routed transactions for Redis Cluster via openTransaction(String routingKey).
• Enforced single-slot validation for cluster transaction sessions.
• Fixed cluster test cleanup to avoid cross-slot multi-key deletes.

0.0.3

• Added a publish-friendly example/example.dart entry example.
• Expanded dartdoc coverage for public client, session, connection, and utility APIs.
• Aligned the README quick start with the published package example.

0.0.2

• Refined command capability modeling around concrete client and session types.
• Moved command helpers to mixin-based composition for standalone, cluster, and transaction clients.
• Improved integration tests and cluster test configuration coverage.

0.0.1

• Initial release with single-node/cluster clients and connection pooling.