hyperstreamdb 0.3.3__tar.gz → 0.4.0__tar.gz

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/Cargo.lock

@@ -14,12 +14,6 @@ version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "aae1277d39aeec15cb388266ecc24b11c80469deae6067e17a1a7aa9e5c1f234"
 
-[[package]]
-name = "ahash"
-version = "0.4.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0453232ace82dee0dd0b4c87a59bd90f7b53b314f3e0f61fe2ee7c8a16482289"
-
 [[package]]
 name = "ahash"
 version = "0.8.12"
@@ -160,15 +154,6 @@ dependencies = [
  "uuid",
 ]
 
-[[package]]
-name = "approx"
-version = "0.5.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "cab112f0a86d568ea0e627cc1d6be74a1e9cd55214684db5561995f6dad897c6"
-dependencies = [
- "num-traits",
-]
-
 [[package]]
 name = "ar_archive_writer"
 version = "0.5.1"
@@ -232,7 +217,7 @@ version = "57.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4c8955af33b25f3b175ee10af580577280b4bd01f7e823d94c7cdef7cf8c9aef"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow-buffer",
  "arrow-data",
  "arrow-schema",
@@ -402,7 +387,7 @@ version = "57.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "68bf3e3efbd1278f770d67e5dc410257300b161b93baedb3aae836144edcaf4b"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow-array",
  "arrow-buffer",
  "arrow-data",
@@ -1330,17 +1315,6 @@ dependencies = [
  "half",
 ]
 
-[[package]]
-name = "cl3"
-version = "0.9.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b823f24e72fa0c68aa14a250ae1c0848e68d4ae188b71c3972343e45b46f8644"
-dependencies = [
- "libc",
- "opencl-sys",
- "thiserror 1.0.69",
-]
-
 [[package]]
 name = "clap"
 version = "4.6.0"
@@ -1893,7 +1867,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "205dc1e20441973f470e6b7ef87626a3b9187970e5106058fef1b713047f770c"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "arrow-ipc",
  "chrono",
@@ -2156,7 +2130,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4e42c227d9e55a6c8041785d4a8a117e4de531033d480aae10984247ac62e27e"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "datafusion-common",
  "datafusion-doc",
@@ -2177,7 +2151,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cead3cfed825b0b688700f4338d281cd7857e4907775a5b9554c083edd5f3f95"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "datafusion-common",
  "datafusion-expr-common",
@@ -2288,7 +2262,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "987c5e29e96186589301b42e25aa7d11bbe319a73eb02ef8d755edc55b5b89fc"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "datafusion-common",
  "datafusion-expr",
@@ -2327,7 +2301,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "602d1970c0fe87f1c3a36665d131fbfe1c4379d35f8fc5ec43a362229ad2954d"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "chrono",
  "datafusion-common",
@@ -2363,7 +2337,7 @@ version = "52.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c21d94141ea5043e98793f170798e9c1887095813b8291c5260599341e383a38"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow",
  "arrow-ord",
  "arrow-schema",
@@ -3047,15 +3021,6 @@ dependencies = [
  "zerocopy",
 ]
 
-[[package]]
-name = "hashbrown"
-version = "0.9.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d7afe4a420e3fe79967a00898cc1f4db7c8a49a9333a29f8a4bd76a253d5cd04"
-dependencies = [
- "ahash 0.4.8",
-]
-
 [[package]]
 name = "hashbrown"
 version = "0.12.3"
@@ -3067,6 +3032,10 @@ name = "hashbrown"
 version = "0.14.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e5274423e17b7c9fc20b6e7e208532f9b19825d82dfd615708b70edd83df41f1"
+dependencies = [
+ "ahash",
+ "allocator-api2",
+]
 
 [[package]]
 name = "hashbrown"
@@ -3364,20 +3333,11 @@ dependencies = [
  "tracing",
 ]
 
-[[package]]
-name = "hyperloglogplus"
-version = "0.4.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "621debdf94dcac33e50475fdd76d34d5ea9c0362a834b9db08c3024696c1fbe3"
-dependencies = [
- "serde",
-]
-
 [[package]]
 name = "hyperstreamdb"
-version = "0.3.3"
+version = "0.4.0"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "anyhow",
  "apache-avro",
  "arrow",
@@ -3407,27 +3367,23 @@ dependencies = [
  "env_logger",
  "faststr",
  "futures",
- "hashbrown 0.9.1",
+ "hashbrown 0.14.5",
  "hive_metastore",
- "hyperloglogplus",
  "jni",
  "lazy_static",
  "log",
  "metal 0.27.0",
  "moka",
  "murmur3",
- "num-traits",
  "num_cpus",
  "numpy",
  "object_store",
  "once_cell",
- "opencl3",
  "opentelemetry",
  "opentelemetry-otlp",
  "opentelemetry_sdk",
  "parking_lot",
  "parquet",
- "pilota",
  "pollster",
  "prometheus",
  "proptest",
@@ -3444,7 +3400,6 @@ dependencies = [
  "serde_json",
  "sha2",
  "skiplist",
- "smartcore",
  "sqlx",
  "tempfile",
  "thiserror 1.0.69",
@@ -4105,7 +4060,7 @@ version = "0.7.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cdc67a1d6ef0340a7f5152b9fd34ad7477b4d518920f7557267f1fc6e5a62641"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "faststr",
  "paste",
  "rustc-hash 2.1.1",
@@ -4332,20 +4287,6 @@ dependencies = [
  "windows-sys 0.61.2",
 ]
 
-[[package]]
-name = "num"
-version = "0.4.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "35bd024e8b2ff75562e5f34e7f4905839deb4b22955ef5e73d2fea1b9813cb23"
-dependencies = [
- "num-bigint",
- "num-complex",
- "num-integer",
- "num-iter",
- "num-rational",
- "num-traits",
-]
-
 [[package]]
 name = "num-bigint"
 version = "0.4.6"
@@ -4408,17 +4349,6 @@ dependencies = [
  "num-traits",
 ]
 
-[[package]]
-name = "num-rational"
-version = "0.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f83d14da390562dca69fc84082e73e548e1ad308d24accdedd2720017cb37824"
-dependencies = [
- "num-bigint",
- "num-integer",
- "num-traits",
-]
-
 [[package]]
 name = "num-traits"
 version = "0.2.19"
@@ -4561,25 +4491,6 @@ version = "11.1.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d6790f58c7ff633d8771f42965289203411a5e5c68388703c06e14f24770b41e"
 
-[[package]]
-name = "opencl-sys"
-version = "0.2.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "de15dd01496ae90c5799f5266184ab020082b4065800ff0b732f489371d0e5cf"
-dependencies = [
- "libc",
-]
-
-[[package]]
-name = "opencl3"
-version = "0.9.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "26ab4a90cb496f787d3934deb0c54fa9d65e7bed710c10071234aab0196fba04"
-dependencies = [
- "cl3",
- "libc",
-]
-
 [[package]]
 name = "openssl-probe"
 version = "0.2.1"
@@ -4727,7 +4638,7 @@ version = "57.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6ee96b29972a257b855ff2341b37e61af5f12d6af1158b6dcdb5b31ea07bb3cb"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "arrow-array",
  "arrow-buffer",
  "arrow-cast",
@@ -4815,7 +4726,7 @@ version = "0.11.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "520e32b295867834617b8a352b6fd4837e578f328f4e48813db4904ebc0d70b2"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "anyhow",
  "async-recursion",
  "bytes",
@@ -6188,19 +6099,6 @@ dependencies = [
  "serde",
 ]
 
-[[package]]
-name = "smartcore"
-version = "0.3.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c42ca1fcd851ada8834d3dfcd088850dc8c703bde50c2baccd89181b74dc3ade"
-dependencies = [
- "approx",
- "cfg-if",
- "num",
- "num-traits",
- "rand 0.8.5",
-]
-
 [[package]]
 name = "snap"
 version = "1.1.1"
@@ -7373,7 +7271,7 @@ version = "0.10.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a4a41a936651c3b8206339898466fd4f39596cfedf789e5ee6a8d02a45feb45f"
 dependencies = [
- "ahash 0.8.12",
+ "ahash",
  "anyhow",
  "bytes",
  "chrono",

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "hyperstreamdb"
-version = "0.3.3"
+version = "0.4.0"
 edition = "2021"
 license = "MIT AND Apache-2.0"
 description = "HyperStreamDB - Serverless Index-Streaming Database with Overlay Indexing and Vector Search"
@@ -30,20 +30,16 @@ name = "hyperstreamdb"
 crate-type = ["cdylib", "rlib"]
 
 [features]
-default = []
+default = ["simd"]
 enterprise = []
-python = ["dep:pyo3", "pyo3/extension-module", "dep:numpy", "dep:pythonize"]
+python = ["dep:pyo3", "pyo3/extension-module", "dep:numpy", "dep:pythonize", "wgpu", "pollster"]
 java = ["dep:jni"]
-simdeez_f = []
 stdsimd = []
-intel_cpu = ["simdeez_f", "stdsimd"]
-simd = ["intel_cpu"]
+simd = []
 
 # GPU Acceleration Features
-cuda = ["dep:cudarc"]
-mps = []
-rocm = []
-intel = []
+wgpu = ["dep:wgpu"]
+pollster = ["dep:pollster"]
 
 [dependencies]
 # core columnar memory format
@@ -79,7 +75,6 @@ aws-sdk-glue = "1.0"
 
 # Thrift for Hive Metastore
 volo-thrift = "0.10"
-pilota = "0.11"
 
 # Hardware Acceleration (moved to bottom of file)
 # indexing
@@ -90,11 +85,9 @@ bincode = "1.3"
 cpu-time = "1.0"
 env_logger = "0.10"
 log = "0.4"
-hashbrown = "0.9"
-num-traits = "0.2"
+hashbrown = "0.14"
 num_cpus = "1.16"
 parking_lot = "0.12"
-skiplist = "0.3"
 # lance removed due to missing protoc
 
 # error handling
@@ -117,9 +110,7 @@ datafusion-functions = "52.0.0"
 datafusion-functions-aggregate = "52.0.0"
 datafusion-functions-aggregate-common = "52.0.0"
 async-stream = "0.3.6"
-smartcore = "0.3"  # For k-means clustering (IVF index)
-rand = "0.8"  # For k-means initialization
-hyperloglogplus = "0.4"  # For NDV (distinct count) estimation
+rand = "0.8"
 clap = { version = "4.5.54", features = ["derive"] }
 rustyline = "17.0.2"
 hive_metastore = "0.2.0"
@@ -132,28 +123,19 @@ opentelemetry-otlp = { version = "0.15", features = ["grpc-tonic"] }
 opentelemetry_sdk = { version = "0.22", features = ["rt-tokio"] }
 tracing-opentelemetry = "0.23"
 base64 = "0.22.1"
-sqlx = { version = "0.8", features = ["runtime-tokio-rustls", "sqlite", "postgres", "mysql", "chrono", "uuid", "any"] }
+sqlx = { version = "0.8", features = ["runtime-tokio-rustls", "sqlite", "chrono", "uuid", "any"] }
 murmur3 = "0.5.2"
 regex = "1.10"
 
 sha2 = "0.10"
-wgpu = "0.20.0"
-pollster = "0.4.0"
+wgpu = { version = "0.20.0", optional = true }
+pollster = { version = "0.4.0", optional = true }
 
 [target.'cfg(not(target_os = "macos"))'.dependencies]
-cudarc = { version = "0.13.0", features = ["driver", "nvrtc", "cuda-12000"], optional = true }
-opencl3 = { version = "0.9" } # Intel / OpenCL
+cudarc = { version = "0.13.0", features = ["driver", "nvrtc", "cuda-12000"] }
 
 [target.'cfg(target_os = "macos")'.dependencies]
-metal = { version = "0.27.0" } # MPS
-opencl3 = { version = "0.9" } # Intel / OpenCL
-
-[target.'cfg(target_os = "linux")'.dependencies]
-# rocm-smi = { version = "0.4.0", optional = true } # ROCm monitoring
-# Note: rocm-rs is not on crates.io, we will use git or a local path if needed.
-# For now, we will use opencl3 as a fallback/stub for ROCm or assume system libraries.
-# If rocm-rs is required, we can add:
-# rocm-rs = { git = "https://github.com/RadeonOpenCompute/rocm-rs", optional = true } 
+metal = { version = "0.27.0" } # MPS 
 
 [dependencies.jni]
 version = "0.21"
@@ -180,6 +162,7 @@ rand_distr = "0.4"
 # hnsw_rs removed and internalized
 tempfile = "3.10"
 proptest = "1.4"
+skiplist = "0.3"
 
 [[bin]]
 name = "generate_iceberg_manifests"
@@ -196,4 +179,3 @@ harness = false
 [[bench]]
 name = "bench_table"
 harness = false
-

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: hyperstreamdb
-Version: 0.3.3
+Version: 0.4.0
 Classifier: Development Status :: 3 - Alpha
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Rust
@@ -16,20 +16,12 @@ Classifier: Topic :: Database
 Classifier: Topic :: Scientific/Engineering
 Requires-Dist: pyarrow>=14.0.0
 Requires-Dist: numpy>=1.20.0
-Requires-Dist: mps ; extra == 'all-gpu'
-Requires-Dist: cuda ; extra == 'all-gpu'
-Requires-Dist: rocm ; extra == 'all-gpu'
-Requires-Dist: intel ; extra == 'all-gpu'
 Requires-Dist: pytest>=7.0 ; extra == 'dev'
 Requires-Dist: pytest-asyncio>=0.21 ; extra == 'dev'
 Requires-Dist: maturin>=1.7 ; extra == 'dev'
 Requires-Dist: maturin>=1.7 ; extra == 'src'
 Provides-Extra: all_gpu
-Provides-Extra: cuda
 Provides-Extra: dev
-Provides-Extra: intel
-Provides-Extra: mps
-Provides-Extra: rocm
 Provides-Extra: src
 License-File: LICENSE-APACHE
 License-File: LICENSE-MIT

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/build.rs

@@ -19,10 +19,10 @@ fn main() {
     
     let target_os = env::var("CARGO_CFG_TARGET_OS").unwrap_or_default();
     
-    // CUDA kernel compilation - only when the cuda feature is explicitly enabled
+    // CUDA kernel compilation - unconditionally on non-macos platforms
     // (MPS is handled by #[cfg(target_os = "macos")] in source code)
     // (Intel/ROCm are handled natively via WGPU)
-    if env::var("CARGO_FEATURE_CUDA").is_ok() && target_os != "macos" {
+    if target_os != "macos" {
         let has_nvcc = Command::new("nvcc").arg("--version").output().is_ok();
         
         if has_nvcc {
@@ -51,7 +51,7 @@ fn main() {
                 }
             }
         } else {
-            println!("cargo:warning=CUDA feature enabled but nvcc not found. Creating dummy PTX files to allow CI to pass Cargo Check/Docs.");
+            println!("cargo:warning=nvcc not found. Creating dummy PTX files to allow CI to pass Cargo Check/Docs.");
             let out_dir = env::var("OUT_DIR").unwrap();
             let kernels = vec![
                 "l2_distance",

hyperstreamdb-0.4.0/docs/monitoring.md

@@ -0,0 +1,57 @@
+# HyperStreamDB Monitoring & Operations
+
+HyperStreamDB is designed to be highly observable and easy to operate in serverless, ephemeral environments as well as traditional long-running daemon setups.
+
+## Telemetry & Tracing
+
+HyperStreamDB leverages a push-based OpenTelemetry (OTLP) pipeline to export distributed traces. Because it often runs in serverless functions (like AWS Lambda) where instances may be frozen between invocations, the `Table` API explicitly flushes traces on teardown (via the `Drop` trait) to ensure no data is lost.
+
+### Configuration
+
+Tracing is disabled by default. To enable tracing, set the `JAEGER_ENABLED` environment variable. When enabled, traces are automatically pushed to an OTLP-compatible endpoint.
+
+| Environment Variable | Description | Default |
+|----------------------|-------------|---------|
+| `JAEGER_ENABLED` | Set to `true` to enable OpenTelemetry exporting. | `false` |
+| `OTEL_EXPORTER_OTLP_ENDPOINT` | The destination URL for OTLP traces. | `http://localhost:4317` (OTLP/gRPC default) |
+| `RUST_LOG` | The log level filter (e.g., `info`, `hyperstreamdb=debug`). | `info` |
+
+### Core Instrumentation
+We instrument key paths to provide visibility into latency and bottlenecks:
+- **Write Path:** `write_async`, `commit_async`
+- **Read Path:** `read_async`, `stream_all`, `vector_search_index`, `vector_search_flat`
+- **Manifest Orchestration:** Optimistic concurrency loops, schema updates, and conflict resolution in the `ManifestManager`.
+
+## Stateless Operational CLI (`hdb`)
+
+The `hdb` binary is a standalone CLI tool that performs administrative actions directly against the object storage tier without requiring a long-running database server to be active.
+
+### Usage
+
+```bash
+# Start an interactive SQL REPL
+hdb repl
+
+# Execute a single SQL query
+hdb query --query "SELECT * FROM my_table LIMIT 10"
+
+# Register a table in the session
+hdb register --name my_table --uri s3://my-bucket/my-table
+```
+
+### Table Management
+
+You can perform routine maintenance using the `hdb table` subcommand:
+
+```bash
+# Inspect table metadata and statistics
+hdb table inspect --uri s3://my-bucket/my-table
+
+# Compact small data files to optimize read performance
+hdb table compact --uri s3://my-bucket/my-table
+
+# Vacuum (delete) data files that are no longer referenced and older than N days
+hdb table vacuum --uri s3://my-bucket/my-table --older-than-days 7
+```
+
+*Note: The CLI is stateless; it interacts directly with the storage URI.*

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/pyproject.toml

@@ -41,11 +41,7 @@ dev = [
     "maturin>=1.7",
 ]
 # Hardware acceleration extras
-mps = []   # Apple Silicon (Metal)
-cuda = []  # NVIDIA (CUDA)
-rocm = []  # AMD (via OpenCL/ROCm)
-intel = [] # Intel (OpenCL)
-all_gpu = ["mps", "cuda", "rocm", "intel"]
+all_gpu = []
 src = ["maturin>=1.7"]
 
 [project.urls]
@@ -54,9 +50,7 @@ Repository = "https://github.com/rla3rd/hyperstreamdb"
 
 [tool.maturin]
 features = ["python"]
-# Map pip extras to Cargo features for source builds
-# e.g. pip install hyperstreamdb[cuda] --no-binary :all:
-extra-features = { cuda = ["python", "cuda"], mps = ["python", "mps"], rocm = ["python", "rocm"], intel = ["python", "intel"] }
+# Build both the Rust library and Python bindings (multi-version compatible)
 module-name = "hyperstreamdb.hyperstreamdb"
 python-source = "python"
 # Build both the Rust library and Python bindings (multi-version compatible)

{hyperstreamdb-0.3.3 → hyperstreamdb-0.4.0}/python/hyperstreamdb/__init__.py

@@ -1,3 +1,5 @@
+from typing import List, Optional, Union, Dict, Any
+import os
 from .hyperstreamdb import Device as _Device
 from .hyperstreamdb import Table as _RustTable
 from .hyperstreamdb import Session as _RustSession
@@ -15,7 +17,7 @@ class Device:
     HyperStreamDB Compute Device (CPU, CUDA, MPS, ROCm, Intel).
     - **Torch Alignment** - Automatically aliases `cuda` to `rocm` on AMD hardware if `torch.version.hip` is detected.
     """
-    def __new__(cls, device: str = "cpu", index: int = 0):
+    def __new__(cls, device: str = "cpu", index: Optional[int] = None):
         device = device.lower()
         # 1. Handle Torch-style alignment
         if device.startswith("cuda"):
@@ -93,8 +95,6 @@ try:
     import polars as pl
 except ImportError:
     pl = None
-from typing import List, Optional, Union, Dict, Any
-import os
 
 class IndexType:
     """
@@ -110,6 +110,10 @@ def _resolve_uri(uri: str) -> str:
     if not uri.startswith(("s3://", "file://", "az://", "gs://", "http://", "https://")):
         return f"file://{os.abspath(uri)}" if hasattr(os, "abspath") else uri
     return uri
+
+def open_table(uri: str, **kwargs) -> Table:
+    """Open an existing HyperStreamDB table."""
+    return Table(uri, **kwargs)
 class Query:
     """
     Fluent Query interface for HyperStreamDB.
@@ -619,10 +623,19 @@ class Table:
         """
         return self._inner.set_index_columns(config)
 
-    def add_index(self, column: str, algorithm: Union[str, Dict[str, Any]] = "hnsw"):
+    def add_index(self, column: str, algorithm: Union[str, Dict[str, Any]] = "hnsw", **kwargs):
         """
         Add an indexing strategy to a column.
         """
+        if isinstance(algorithm, str):
+            algorithm = {"type": algorithm}
+        
+        if kwargs:
+            # Map 'device' to 'build_device' for consistency with set_index_config
+            if 'device' in kwargs:
+                kwargs['build_device'] = kwargs.pop('device')
+            algorithm.update(kwargs)
+            
         return self._inner.add_index(column, algorithm)
 
     def drop_index(self, column: str):

hyperstreamdb-0.4.0/scripts/update_cache.py

@@ -0,0 +1,46 @@
+import re
+import os
+
+files_to_update = [
+    "src/core/index/hnsw_ivf.rs",
+    "src/core/manifest.rs",
+    "src/core/reader.rs"
+]
+
+cache_names = {
+    "HNSW_IVF_CACHE": "hnsw_ivf",
+    "LATEST_VERSION_CACHE": "latest_version",
+    "MANIFEST_CACHE": "manifest",
+    "MANIFEST_LIST_CACHE": "manifest_list",
+    "BLOOM_FILTER_CACHE": "bloom_filter",
+    "INVERTED_INDEX_CACHE": "inverted_index",
+    "BYTE_CACHE": "byte",
+    "INDEX_CACHE": "index",
+    "PARQUET_META_CACHE": "parquet_meta",
+    "BLOCK_CACHE": "block",
+    "HNSW_CACHE": "hnsw"
+}
+
+for file_path in files_to_update:
+    with open(file_path, "r") as f:
+        content = f.read()
+
+    # We need to add `use crate::core::cache::CacheExt;` at the top of the files if not already there
+    if "use crate::core::cache::CacheExt;" not in content:
+        # Find the first use statement and add it
+        content = re.sub(r'^(use [^;]+;)', r'use crate::core::cache::CacheExt;\n\1', content, count=1, flags=re.MULTILINE)
+
+    for cache_var, cache_name in cache_names.items():
+        # Match something like `CACHE_VAR.get(&key).await`
+        # or `crate::core::cache::CACHE_VAR.get(&key).await`
+        pattern = r'(' + cache_var + r')\.get\(([^)]+)\)'
+        
+        def replacer(match):
+            return f'{match.group(1)}.get_with_metrics({match.group(2)}, "{cache_name}")'
+            
+        content = re.sub(pattern, replacer, content)
+
+    with open(file_path, "w") as f:
+        f.write(content)
+
+print("Updated files.")