Merge branch 'main' into gpu-codecs

Author: TomAugspurger

.github/dependabot.yml

@@ -10,17 +10,6 @@ updates:
       actions:
         patterns:
           - "*"
-
-  # Updates for support/v2 branch
-  - package-ecosystem: "pip"
-    directory: "/"
-    target-branch: "support/v2"
-    schedule:
-      interval: "weekly"
-    groups:
-      requirements:
-        patterns:
-          - "*"
   - package-ecosystem: "github-actions"
     directory: "/"
     target-branch: "support/v2"

changes/3655.bugfix.md

@@ -0,0 +1 @@
+Fixed a bug in the sharding codec that prevented nested shard reads in certain cases.

changes/3657.bugfix.md

@@ -0,0 +1 @@
+Fix obstore _transform_list_dir implementation to correctly relativize paths (removing lstrip usage).

changes/3700.bugfix.md

@@ -0,0 +1 @@
+CacheStore, LoggingStore and LatencyStore now support with_read_only.

changes/3702.bugfix.md

@@ -0,0 +1 @@
+Skip chunk coordinate enumeration in resize when the array is only growing, avoiding unbounded memory usage for large arrays.

changes/3704.misc.md

@@ -0,0 +1 @@
+Remove an expensive `isinstance` check from the bytes codec decoding routine.

changes/3705.bugfix.md

@@ -0,0 +1 @@
+Fix a performance bug in morton curve generation.

changes/3706.misc.md

@@ -0,0 +1 @@
+Allow NumPy ints as input when declaring a shape.

changes/3708.misc.md

@@ -0,0 +1 @@
+Optimize Morton order computation with hypercube optimization, vectorized decoding, and singleton dimension removal, providing 10-45x speedup for typical chunk shapes.

changes/3712.misc.md

@@ -0,0 +1 @@
+Added benchmarks for Morton order computation in sharded arrays.