use clap::Parser;
use datafusion::error::Result;
use datafusion::execution::context::SessionConfig;
use datafusion::prelude::SessionContext;
use datafusion_cli::{
exec, print_format::PrintFormat, print_options::PrintOptions, DATAFUSION_CLI_VERSION,
};
use mimalloc::MiMalloc;
use std::env;
use std::path::Path;
#[global_allocator]
static GLOBAL: MiMalloc = MiMalloc;
#[derive(Debug, Parser, PartialEq)]
#[clap(author, version, about, long_about= None)]
struct Args {
#[clap(
short = 'p',
long,
help = "Path to your data, default to current directory",
validator(is_valid_data_dir)
)]
data_path: Option<String>,
#[clap(
short = 'c',
long,
help = "The batch size of each query, or use DataFusion default",
validator(is_valid_batch_size)
)]
batch_size: Option<usize>,
#[clap(
short,
long,
multiple_values = true,
help = "Execute commands from file(s), then exit",
validator(is_valid_file)
)]
file: Vec<String>,
#[clap(
short = 'r',
long,
multiple_values = true,
help = "Run the provided files on startup instead of ~/.datafusionrc",
validator(is_valid_file),
conflicts_with = "file"
)]
rc: Option<Vec<String>>,
#[clap(long, arg_enum, default_value_t = PrintFormat::Table)]
format: PrintFormat,
#[clap(
short,
long,
help = "Reduce printing other than the results and work quietly"
)]
quiet: bool,
}
#[tokio::main]
pub async fn main() -> Result<()> {
env_logger::init();
let args = Args::parse();
if !args.quiet {
println!("DataFusion CLI v{}", DATAFUSION_CLI_VERSION);
}
if let Some(ref path) = args.data_path {
let p = Path::new(path);
env::set_current_dir(&p).unwrap();
};
let mut session_config = SessionConfig::from_env().with_information_schema(true);
if let Some(batch_size) = args.batch_size {
session_config = session_config.with_batch_size(batch_size);
};
let mut ctx = SessionContext::with_config(session_config.clone());
let mut print_options = PrintOptions {
format: args.format,
quiet: args.quiet,
};
let files = args.file;
let rc = match args.rc {
Some(file) => file,
None => {
let mut files = Vec::new();
let home = dirs::home_dir();
if let Some(p) = home {
let home_rc = p.join(".datafusionrc");
if home_rc.exists() {
files.push(home_rc.into_os_string().into_string().unwrap());
}
}
files
}
};
if !files.is_empty() {
exec::exec_from_files(files, &mut ctx, &print_options).await
} else {
if !rc.is_empty() {
exec::exec_from_files(rc, &mut ctx, &print_options).await
}
exec::exec_from_repl(&mut ctx, &mut print_options).await;
}
Ok(())
}
fn is_valid_file(dir: &str) -> std::result::Result<(), String> {
if Path::new(dir).is_file() {
Ok(())
} else {
Err(format!("Invalid file '{}'", dir))
}
}
fn is_valid_data_dir(dir: &str) -> std::result::Result<(), String> {
if Path::new(dir).is_dir() {
Ok(())
} else {
Err(format!("Invalid data directory '{}'", dir))
}
}
fn is_valid_batch_size(size: &str) -> std::result::Result<(), String> {
match size.parse::<usize>() {
Ok(size) if size > 0 => Ok(()),
_ => Err(format!("Invalid batch size '{}'", size)),
}
}