Version 1.3 (#285)

* reinstate collection.delete_many with empty filter for truncate

* deprecation of delete_all (+related test utilities); removal of old core aliased methods

* Deprecation sine die of bulk_write

* changes

* vector/s/ize params deprecation, collection.py

* resiliency of tests (core+idiomatic) to vector not being returned under certain conditions

* improved vectorize_methods settings handling; partial support for hfded and oaiparams

* shorten model root name

* hf ded tweaks to vectorize coll models generation

* adapt vectorize tests to hf ded

* deprecation of vector/ize orderly brought to all Operation classes

* revert deprecation of bulk_insert

* adapt testing to modelName management in HD dedicated

* tests capture and check deprecations for all usages of vector/ize= standalone params

* added HCD and CASSANDRA to the Environment values

* more clarity in string repr of cursors

* use type alias for filter consistently

* callback mechanism to store the sort vector

* expose include_sort_vector option to find

* add prefetch iterator wrappers for the returnvector

* prefetching for include_sort_vector

* revert nullable-model-name changes and cope with disappeared .providerKey suffix in tests

* include_sort_vector and get_sort_vector, extensive test coverage

* add code examples for get_sort_vector in docstring for the two find methods

* add Content-Type header to all api requests

* changelog update

* token providers

* full transition to TokenProvider

* vectorize: support testing of openai project/org ids

* rename attribute token -> token_provider where needed

* docstring about token provider

* token inheritance tests; eq and copy methods of api commander

* changes file

* tests: modernize and lift upward fixtures and paradigmaticize object creation; add test on tokenless client

* astra/nonastra plug-your-db management

* astra/nonastra management fully covers vectorize

* nonAstra admin (i.e. namespace crud) tests

* rework fixture inheritance and naming for alternate keyspace (incl.astra/nonastra)

CHANGES

@@ -0,0 +1,105 @@
+v. 1.3.0
+========
+Integration testing covers Astra and nonAstra smoothly:
+    - idiomatic library
+    - vectorize_idiomatic
+    - nonAstra admin, i.e. namespace crud
+Add the TokenProvider abstract class => and StaticTokenProvider, UsernamePasswordTokenProvider
+Introduce CHANGES file.
+Add __eq__ and _copy methods to APICommander class
+Allow delete_many({}) with empty filter
+Implement include_sort_vector option to Collection.find and get_sort_vector to cursors
+Add Content-Type header to all API requests
+Added HCD and CASSANDRA Environment values (besides the other non-Astra DSE and OTHER)
+Clearer string repr of cursors ('retrieved' => 'yielded so far')
+Deprecation of collection delete_all method in favour of delete_many(filter={})
+    Introduction of a custom deprecation decorator for async method removal tests
+Deprecation of vector,vectors and vectorize params from collections and Operations
+Remove several long-deprecated methods from **core API** (i.e. internal changes):
+    AstraDBCollection.delete          => delete_one
+    AstraDBCollection.upsert          => upsert_one
+    AsyncAstraDBCollection.upsert     => upsert_one
+    AstraDB.truncate_collection       => AstraDBCollection.clear
+    AsyncAstraDB.truncate_collection  => AsyncAstraDBCollectionclear
+Add support for null tokens in the core library
+
+v. 1.2.1
+========
+Raise default chunk size for insert_many to 50
+Improvements in docstrings, testing, support for latest responses from vectorize.
+
+
+v. 1.2.0
+========
+Non-Astra environment awareness:
+    astrapy.constants.Environment enum for the "environment" parameter to client, etc
+    flexibility and adaptive defaults for data api url
+    environment knowledge trickles throughout all classes (client, admins, databases)
+    DataAPIDatabaseAdmin class (i.e. for namespace CRUD)
+(internal) astrapy.api_commander.APICommander
+(internal) astrapy.api_options.{BaseAPIOptions, CollectionAPIOptions}
+$vectorize support:
+    "service options" for creating/retrieving collections, covering $vectorize needs
+    embedding_api_key parameter to collection (for "header" usage)
+collection-level timeout parameter (overridable in single method calls)
+expand "projection" type to include slice projections
+client.get_database can accept an API endpoint directly
+insert_many and bulk_write default to ordered=False
+
+
+v. 1.1.0
+========
+Adds estimated_document_count method to Collection class.
+
+
+v. 1.0.0
+========
+(Introduction of the "idiomatic API", relegating "core" to become non-user-facing.)
+
+Split classes, modules, tests to keep the "idiomatic" layer well separate and not touch the "astrapy" layer
+DDL and some DML methods to m1
+Passing secondary keyspace to action workflows
+Fix bug in copy and to_[a]sync when set_caller is later used
+Full cross-namespace management in DDL
+More DDL methods and signature adjustments for Database
+Sl overridable copy methods
+Cursor/AsyncCursor, find and distinct
+Remove all 'unsupported' clutter; implement find_one (a/sync)
+DML for idiomatic + necessary changes around
+Bulk_write method
+More management methods
+Sl collateral commands
+Commandcursor (+async), used in list_collections
+Sl collection options + cap-aware count_documents
+Fix #245: Delete CHANGES.md
+Collection.drop + database/collection info and related methods/properties (metadata)
+Adapt to latest choices in API semantics
+More syntax changes as discussed + all docstrings
+Full support for dotted key names in distinct
+Refactor to feature 'idiomatic' first, with full back-compat with 'astrapy'
+Add sorting in hashing for distinct and factor it away
+Exception management with a hierarchy of Exception classes
+Full timeout support
+Sl api refinements
+ObjectIDs and UUIDs handled throughout (+ tests)
+Admin interfaces and classes
+Docstrings for all the admin/client parts + minor improvements to docstrings around
+Readme overhaul
+Exporting logger for back-compatibility
+Sl adjustments
+Custom payload serialization for httpx to block NaNs
+Improved docstrings (minor stuff)
+Methods repr/str to all objects for graceful display
+Full test suite on client/admin classes
+Abstract DatabaseAdmin, admin standard utility conversion/methods + tests thereof
+Collection options is a dataclass and not a dict anymore
+Pdoc annotations to control auto-docs
+Logging, create_database signature
+Added async support for admin, as alternate methods on original classes
+Use MultiCallTimeoutManager in create_collection methods
+V1.0.0 ("pm convergenge m1") gets to master
+
+
+(prior to 1.0.0)
+================
+What is now dubbed the "core API" and is not supposed to be used directly at all.

README.md

@@ -62,7 +62,7 @@ my_collection.update_one(
 
 cursor = my_collection.find(
     {},
-    vector=[0, 0.2, 0.4],
+    sort={"$vector": [0, 0.2, 0.4]},
     limit=2,
     include_similarity=True,
 )
@@ -414,6 +414,8 @@ That being said, there are no known breakings of backward compatibility:
 **legacy code would run with a newest astrapy version just as well.**
 Here is a recap of the minor changes that came _to the old API_ with 1.0.0:
 
+- Added support for null tokens (with the effect of no authentication/token header in requests)
+- Added Content-Type header to all HTTP requests to the API
 - Added methods to `[Async]AstraDBCollection`: `delete_one_filter`, 
 - Paginated find methods (sync/async) type change from Iterable to Generator
 - Bugfix: handling of the mutable caller identity in copy and convert (sync/async) methods

astrapy/__init__.py

@@ -89,6 +89,7 @@ def get_version() -> str:
     "api_options": False,
     "core": False,
     "db": False,
+    "meta": False,
     "ops": False,
     "ids": False,
 }