curt_hagenlocher

Apache Arrow is a cross-language development platform for in-memory data. It specifies a standardized language-independent columnar memory format for flat and hierarchical data, organized for efficient analytic operations on modern hardware.

Package Description

Package Description

Apache Arrow scalar types, including Parquet-compatible variants. Provides readers, writers, and object model for the Parquet Variant binary format for semi-structured data.

Package Description

Provides decompression support for the Arrow IPC format

Package Description

Package Description

Package Description

Package Description

Package Description

Clast republish of Google.Apis.Core: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

FSST (Fast Static Symbol Table) string compression for .NET. Supports the FSST8 (1-byte codes), FSST12 (12-bit codes) and FSST16 (2-byte codes) variants.

Clast republish of Google.Apis: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Google.Apis.Auth: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Adaptive Lossless floating-Point (ALP) compression for arrays of double values. Designed for decimal-style data such as prices, sensor readings, and scientific measurements.

Split Block Bloom Filter (Parquet SBBF) implementation with a generic-key facade. AVX2 and ARM NEON accelerated with a scalar fallback. Suitable both for reading and writing Parquet-format bloom filters and for in-process use over arbitrary key types via IHash64<T>. Targets netstandard2.0, net8.0, and net10.0.

Format-specific codecs for the Apache Arrow Variant binary format, including JSON reader/writer.

Clast republish of Google.Api.Gax: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of the generated Google.Apis.Storage.v1 REST client: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Google.Api.Gax.Rest: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Google.Cloud.Storage.V1: the internal UrlSigner JSON moves to System.Text.Json (Utf8JsonWriter), Newtonsoft.Json is removed, the library is trimming/AOT-compatible, and a net10.0 target is added. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Fixed-point decimal arithmetic for database engines: 32/64/128/256-bit mantissa tiers with SQL Server / Substrait precision-and-scale promotion rules, scalar and span (columnar) kernels for add/subtract/multiply/divide/modulus, and UTF-8 / UTF-16 parse and format. Targets netstandard2.0, net8.0, and net10.0.

FastLanes bit-packing codec: wire-compatible with the spiraldb/fastlanes crate used by Lance.

Pco numeric compression codec: wire-compatible with the pcodec/pcodec Rust reference implementation.

Clast republish of the generated Google.Apis.Bigquery.v2 REST client: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Grpc.Auth: re-pointed to the Newtonsoft-free, source-generated System.Text.Json Clast.Google.Apis.Auth so the gRPC credential path is trimming/AOT-compatible, with a net10.0 target added. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official gRPC or Google package.

Clast republish of Google.Api.Gax.Grpc: Newtonsoft.Json replaced with source-generated System.Text.Json, trimming/AOT-compatible, adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Google.Cloud.BigQuery.Storage.V1 (the gRPC BigQuery Storage Read/Write API): built on the trimming/AOT-compatible Clast gRPC stack (System.Text.Json, no Newtonsoft.Json) and adds a net10.0 target. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Clast republish of Google.Cloud.BigQuery.V2: the internal row/result JSON moves to System.Text.Json (JsonElement), Newtonsoft.Json is removed, the library is trimming/AOT-compatible, and a net10.0 target is added. Namespaces and public type names are unchanged. Independent republish from the clast-project fork — not an official Google package.

Streaming sketches with binary compatibility with the Apache DataSketches library: Theta (distinct counting, set operations) and HLL. Reads and writes the same serialized forms as datasketches-java and datasketches-cpp, including the compact Theta sketch stored in Apache Iceberg Puffin files. Targets netstandard2.0, net8.0, and net10.0.