rust(feat): adding hdf5 imports (#565)

Co-authored-by: Brandon Shippy <brandon.shippy@siftstack.com>

Author: Brandon-Shippy

Cargo.toml

@@ -59,6 +59,7 @@ reqwest = "0.13"
 serde = { version = "^1.0" }
 serde_json = { version = "^1.0" }
 tdms = "0.3.0"
+hdf5 = { package = "hdf5-metno", version = "0.12", features = ["static"] }
 tempdir = "0.3"
 tokio = { version = "1" }
 tokio-stream = "0.1"

rust/crates/sift_cli/Cargo.toml

@@ -32,6 +32,7 @@ pbjson-types = { workspace = true }
 reqwest = { workspace = true }
 sift_pbfs = { workspace = true }
 tdms = { workspace = true }
+hdf5 = { workspace = true }
 sift_rs = { workspace = true }
 tokio = { workspace = true, features = ["full", "net", "time"] }
 tokio-stream = { workspace = true }

rust/crates/sift_cli/src/cli/hdf5.rs

@@ -0,0 +1,20 @@
+use std::fmt::{self, Display};
+
+use clap::ValueEnum;
+
+#[derive(Debug, Copy, Clone, PartialEq, Eq, ValueEnum)]
+pub enum Hdf5Schema {
+    OneD,
+    TwoD,
+    Compound,
+}
+
+impl Display for Hdf5Schema {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::OneD => write!(f, "one-d"),
+            Self::TwoD => write!(f, "two-d"),
+            Self::Compound => write!(f, "compound"),
+        }
+    }
+}

rust/crates/sift_cli/src/cli/mod.rs

@@ -1,7 +1,9 @@
 use clap::{Parser, Subcommand, crate_description, crate_version};
 use clap_complete::Shell;
 use parquet::ComplexTypesMode;
+pub mod hdf5;
 pub mod tdms;
+use hdf5::Hdf5Schema;
 use std::path::PathBuf;
 use tdms::TdmsFallbackMethod;
 
@@ -161,6 +163,11 @@ pub enum ImportCmd {
     /// Import a Tdms file into Sift.
     Tdms(ImportTdmsArgs),
 
+    /// Import an HDF5 file into Sift. Supported channel types: bool,
+    /// int8/16/32/64, uint8/16/32/64, float32, float64. Datasets with other
+    /// types will produce a client-side error.
+    Hdf5(ImportHdf5Args),
+
     /// Import backup files generated by sift_stream into Sift.
     #[command(name = "backups")]
     Backup(BackupArgs),
@@ -420,3 +427,31 @@ pub struct ImportTdmsArgs {
     #[arg(long)]
     pub import_file_properties: bool,
 }
+
+#[derive(clap::Args)]
+pub struct ImportHdf5Args {
+    #[command(flatten)]
+    pub common: CommonImportArgs,
+
+    /// Schema type for the HDF5 file. Supported channel types across all
+    /// schemas: bool, int8/16/32/64, uint8/16/32/64, float32, float64.
+    #[arg(long)]
+    pub schema: Hdf5Schema,
+
+    /// Time format used in the time dataset/column
+    #[arg(long)]
+    pub time_format: Option<TimeFormat>,
+
+    /// Start time (RFC3339) if the time format is relative
+    #[arg(short = 's', long)]
+    pub relative_start_time: Option<String>,
+
+    /// (two-d / compound) Index of the time column or field. Defaults to 0.
+    /// Mutually exclusive with --time-field.
+    #[arg(long, conflicts_with = "time_field")]
+    pub time_index: Option<u64>,
+
+    /// (compound) Name of the time field. Mutually exclusive with --time-index.
+    #[arg(long)]
+    pub time_field: Option<String>,
+}