Document new features in README and changelog

Author: laurentpellegrino

CHANGELOG.md

@@ -6,10 +6,31 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ## [Unreleased]
+### Added
+- Add an asyncio-based `AsyncIpregistryClient` with the same feature set, available with the `async`
+  extra: `pip install ipregistry[async]`.
+- Add configurable retries through `IpregistryConfig`: `retry_max_attempts`, `retry_interval`,
+  `retry_on_server_error` and `retry_on_too_many_requests` (honoring the `Retry-After` header).
+  Transient network errors are now retried too.
+- Add a typed `ErrorCode` enum exposed as `ApiError.error_code`.
+- Add automatic splitting of batch lookups larger than the API limit of 1024 items into concurrent
+  chunks, tunable with the `max_batch_size` and `batch_concurrency` client options.
+- Add `close()` and context manager support to `IpregistryClient`, plus an optional `session`
+  option to provide a custom `requests.Session`.
+- Add a `user_agent` configuration option to customize the User-Agent header.
+- Add a release workflow publishing to PyPI through trusted publishing, a ruff lint workflow, and
+  a weekly scheduled workflow running tests against the live API.
 ### Changed
 - Require Python 3.10+.
-- Update all dependencies to their latest versions.
+- Update all dependencies to their latest versions and remove the dependency on tenacity.
+- Reuse pooled HTTP connections through a `requests.Session` instead of opening a new connection
+  per request.
+- Align the default request timeout with other Ipregistry client libraries: 15 seconds instead of 5.
+- Origin lookups are no longer cached: cached requester data would be stale or wrong when the
+  network context changes.
+- Cache keys are now deterministic regardless of the order lookup options are given in.
 - Remove the redundant retry on `origin_lookup_ip` that could triple retry attempts.
+- Detect the Yahoo Slurp crawler in `UserAgents.is_bot`.
 ### Fixed
 - Fix nested model fields (`Connection`, `CurrencyFormat`, `UserAgent` device/engine/os, `RequesterIpInfo` user_agent)
   raising a validation error when excluded from a response using the `fields` parameter.

README.md

@@ -9,7 +9,7 @@ This is the official Python client library for the [Ipregistry](https://ipregist
 allowing you to lookup your own IP address or specified ones. Responses return multiple data points including carrier, 
 company, currency, location, timezone, threat information, and more.
 
-Starting version 3 of the library, support for Python 2 has been dropped and the library requires Python 3.6+.
+Starting version 3 of the library, support for Python 2 has been dropped. Version 5 and above require Python 3.10+.
 
 ## Getting Started
 
@@ -85,9 +85,51 @@ response = client.parse_user_agent('Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/
 print(response.data)
 ```
 
+#### Asynchronous Client
+
+An asyncio-based client with the same feature set is available when the library is installed
+with the `async` extra (`pip install ipregistry[async]`):
+
+```python
+from ipregistry import AsyncIpregistryClient
+
+async def main():
+    async with AsyncIpregistryClient("YOUR_API_KEY") as client:
+        response = await client.lookup_ip("54.85.132.205")
+        print(response.data.location.country.code)
+```
+
 More advanced examples are available in the [samples](https://github.com/ipregistry/ipregistry-python/tree/master/samples) 
 folder.
 
+### Configuration
+
+Timeouts, retries and the User-Agent header are configurable through `IpregistryConfig`:
+
+```python
+from ipregistry import IpregistryClient, IpregistryConfig
+
+config = IpregistryConfig(
+    "YOUR_API_KEY",
+    timeout=15,                        # seconds; a (connect, read) tuple is also accepted
+    retry_max_attempts=3,              # attempts per request, including the initial one
+    retry_interval=1,                  # base delay in seconds, doubled on each retry
+    retry_on_server_error=True,        # retry 5xx responses
+    retry_on_too_many_requests=False,  # retry 429 responses, honoring Retry-After
+    user_agent=None                    # custom User-Agent header value
+)
+client = IpregistryClient(config)
+```
+
+Transient network errors are always retried up to `retry_max_attempts`.
+
+The client reuses pooled HTTP connections through a `requests.Session`. You may pass your own
+session with `IpregistryClient("YOUR_API_KEY", session=my_session)` for proxy or TLS control,
+and release resources with `client.close()` or by using the client as a context manager.
+
+Batch lookups larger than the API limit of 1024 items are automatically split into concurrent
+chunks. Tune this with `IpregistryClient("YOUR_API_KEY", max_batch_size=1024, batch_concurrency=4)`.
+
 ### Caching
 
 This Ipregistry client library has built-in support for in-memory caching. By default caching is disabled. 
@@ -130,7 +172,7 @@ All Ipregistry exceptions inherit `IpregistryError` class.
 
 Main subtypes are `ApiError` and `ClientError`.
 
-Errors of type _ApiError_ include a code field that maps to the one described in the [Ipregistry documentation](https://ipregistry.co/docs/errors).
+Errors of type _ApiError_ include a `code` field that maps to the one described in the [Ipregistry documentation](https://ipregistry.co/docs/errors), along with a typed `error_code` enum value (`ErrorCode`) that is `None` for unrecognized codes.
 
 ### Filtering bots