Files
foxhunt/data/examples/convert_dbn_to_parquet.rs
jgrusewski e8a68ee39f Download 360 DBN files (36.3 MB) using Rust databento client
- Created data/examples/download_ml_training_data.rs using reqwest + Databento HTTP API
- Downloaded 90 days × 4 symbols (ES.FUT, NQ.FUT, ZN.FUT, 6E.FUT)
- Files saved to test_data/real/databento/ml_training/
- Total: 360 files, 15 MB compressed DBN format
- Used existing Rust pattern from download_nq_fut.rs
- API key loaded from .env file
- 100% success rate (360/360 files)
- Ready for ML training benchmarks

Next: Create simplified training benchmark for RTX 3050 Ti GPU measurements
2025-10-13 13:30:02 +02:00

135 lines
4.2 KiB
Rust

//! CLI Tool: DBN to Parquet Converter
//!
//! Command-line tool for converting Databento binary format (DBN) files to Parquet format.
//!
//! ## Usage
//!
//! ```bash
//! # Convert a single DBN file
//! cargo run --example convert_dbn_to_parquet -- \
//! --input test_data/real/databento/ES.FUT_ohlcv-1m_2024-01-02.dbn \
//! --output market_data/converted
//!
//! # With custom configuration
//! cargo run --example convert_dbn_to_parquet -- \
//! --input data.dbn \
//! --output ./parquet \
//! --batch-size 5000 \
//! --compression snappy
//! ```
use anyhow::{Context, Result};
use clap::Parser;
use data::parquet_persistence::ParquetConfig;
use data::providers::databento::{ConversionReport, DbnToParquetConverter};
use std::path::PathBuf;
use tracing_subscriber::{layer::SubscriberExt, util::SubscriberInitExt};
#[derive(Parser, Debug)]
#[clap(name = "dbn-to-parquet")]
#[clap(about = "Convert Databento DBN files to Parquet format", long_about = None)]
struct Args {
/// Input DBN file path
#[clap(short, long)]
input: PathBuf,
/// Output directory for Parquet files
#[clap(short, long, default_value = "./market_data")]
output: PathBuf,
/// Batch size for processing (events per batch)
#[clap(short, long, default_value_t = 10000)]
batch_size: usize,
/// Compression algorithm (snappy, gzip, lz4, zstd, none)
#[clap(short, long, default_value = "snappy")]
compression: String,
/// Enable verbose logging
#[clap(short, long)]
verbose: bool,
}
#[tokio::main]
async fn main() -> Result<()> {
let args = Args::parse();
// Initialize tracing
let log_level = if args.verbose {
tracing::Level::DEBUG
} else {
tracing::Level::INFO
};
tracing_subscriber::registry()
.with(
tracing_subscriber::fmt::layer()
.with_target(false)
.with_level(true),
)
.with(
tracing_subscriber::filter::LevelFilter::from_level(log_level),
)
.init();
// Validate input file exists
if !args.input.exists() {
anyhow::bail!("Input file does not exist: {:?}", args.input);
}
// Create output directory if needed
if !args.output.exists() {
std::fs::create_dir_all(&args.output)
.with_context(|| format!("Failed to create output directory: {:?}", args.output))?;
}
// Parse compression algorithm
let compression = match args.compression.to_lowercase().as_str() {
"snappy" => parquet::basic::Compression::SNAPPY,
"gzip" => parquet::basic::Compression::GZIP(parquet::basic::GzipLevel::default()),
"lz4" => parquet::basic::Compression::LZ4,
"zstd" => parquet::basic::Compression::ZSTD(parquet::basic::ZstdLevel::default()),
"none" => parquet::basic::Compression::UNCOMPRESSED,
other => anyhow::bail!("Unknown compression algorithm: {}", other),
};
// Configure Parquet writer
let config = ParquetConfig {
base_path: args.output.to_string_lossy().to_string(),
batch_size: args.batch_size,
compression,
..Default::default()
};
// Create converter
tracing::info!("Initializing DBN to Parquet converter...");
let mut converter = DbnToParquetConverter::new(config).await?;
// Convert file
tracing::info!("Converting {:?}...", args.input);
let report = converter.convert_file(&args.input).await?;
// Print results
print_report(&report);
if !report.is_success() {
anyhow::bail!("Conversion completed with {} errors", report.events_failed);
}
tracing::info!("✓ Conversion successful!");
Ok(())
}
fn print_report(report: &ConversionReport) {
println!("\n{}", "=".repeat(60));
println!("Conversion Report");
println!("{}", "=".repeat(60));
println!("Events processed: {}", report.events_processed);
println!("Events skipped: {}", report.events_skipped);
println!("Events failed: {}", report.events_failed);
println!("Duration: {:?}", report.duration);
println!("Throughput: {} events/sec", report.throughput_events_per_sec);
println!("Success rate: {:.2}%", report.success_rate());
println!("{}", "=".repeat(60));
}