foxhunt/data/examples/inspect_parquet_schema.rs

//! Inspect Parquet file schema for debugging

use parquet::arrow::arrow_reader::ParquetRecordBatchReaderBuilder;
use std::fs::File;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let file_path =
        "/home/jgrusewski/Work/foxhunt/test_data/real/parquet/BTC-USD_30day_2024-09.parquet";
    let file = File::open(file_path)?;
    let builder = ParquetRecordBatchReaderBuilder::try_new(file)?;
    let schema = builder.schema().clone();

    println!("Parquet Schema:");
    println!("================");
    for (i, field) in schema.fields().iter().enumerate() {
        println!(
            "Column {}: {} -> {:?} (nullable: {})",
            i,
            field.name(),
            field.data_type(),
            field.is_nullable()
        );
    }

    // Read first few rows to see data
    let mut reader = builder.build()?;
    if let Some(Ok(batch)) = reader.next() {
        println!("\nFirst batch rows: {}", batch.num_rows());
        println!("First batch columns: {}", batch.num_columns());

        // Print first 3 rows
        for col_idx in 0..batch.num_columns() {
            println!("\nColumn {}: {}", col_idx, schema.field(col_idx).name());
            println!(
                "{:?}",
                batch.column(col_idx).slice(0, 3.min(batch.num_rows()))
            );
        }
    }

    Ok(())
}