select.rs

static USAGE: &str = r#"
Select columns from CSV data efficiently.

This command lets you manipulate the columns in CSV data. You can re-order
them, duplicate them or drop them. Columns can be referenced by index or by
name if there is a header row (duplicate column names can be disambiguated with
more indexing). Column ranges can also be specified. Finally, columns can be
selected using regular expressions.

  Select the first and fourth columns:
  $ qsv select 1,4

  Select the first 4 columns (by index and by name):
  $ qsv select 1-4
  $ qsv select Header1-Header4

  Ignore the first 2 columns (by range and by omission):
  $ qsv select 3-
  $ qsv select '!1-2'

  Select the third column named 'Foo':
  $ qsv select 'Foo[2]'

  Select the first and last columns, _ is a special character for the last column:
  $ qsv select 1,_

  Reverse the order of columns:
  $ qsv select _-1

  Sort the columns lexicographically (i.e. by their byte values)
  $ qsv select 1- --sort

  Select some columns and then sort them:
  $ qsv select 1,4,5-7 --sort

  Randomly shuffle the columns:
  $ qsv select 1- --random
  # with a seed
  $ qsv select 1- --random --seed 42

  Select some columns and then shuffle them with a seed:
  $ qsv select 1,4,5-7 --random --seed 42

  Select columns using a regex using '/<regex>/':
  $ qsv select /^a/
  $ qsv select '/^.*\d.*$/'
  # remove SSN, account_no and password columns
  $ qsv select '!/SSN|account_no|password/'

  Re-order and duplicate columns arbitrarily:
  $ qsv select 3-1,Header3-Header1,Header1,Foo[2],Header1

  Quote column names that conflict with selector syntax:
  $ qsv select '\"Date - Opening\",\"Date - Actual Closing\"'

For more examples, see https://github.com/jqnatividad/qsv/blob/master/tests/test_select.rs.

Usage:
    qsv select [options] [--] <selection> [<input>]
    qsv select --help

select options:
These options only apply to the `select` command, not the `--select` option in other commands.

    -R, --random           Randomly shuffle the columns in the selection.
    --seed <number>        Seed for the random number generator.

    -S, --sort             Sort the selected columns lexicographically,
                           i.e. by their byte values.

Common options:
    -h, --help             Display this message
    -o, --output <file>    Write output to <file> instead of stdout.
    -n, --no-headers       When set, the first row will not be interpreted
                           as headers. (i.e., They are not searched, analyzed,
                           sliced, etc.)
    -d, --delimiter <arg>  The field delimiter for reading CSV data.
                           Must be a single character. (default: ,)
"#;

use rand::{seq::SliceRandom, SeedableRng};
use serde::Deserialize;

use crate::{
    config::{Config, Delimiter},
    select::SelectColumns,
    util, CliResult,
};

#[derive(Deserialize)]
struct Args {
    arg_input:       Option<String>,
    arg_selection:   SelectColumns,
    flag_random:     bool,
    flag_seed:       Option<u64>,
    flag_sort:       bool,
    flag_output:     Option<String>,
    flag_no_headers: bool,
    flag_delimiter:  Option<Delimiter>,
}

pub fn run(argv: &[&str]) -> CliResult<()> {
    let args: Args = util::get_args(USAGE, argv)?;

    if args.flag_sort && args.flag_random {
        return fail_clierror!("Cannot use both --random and --sort options.");
    }

    let rconfig = Config::new(&args.arg_input)
        .delimiter(args.flag_delimiter)
        .no_headers(args.flag_no_headers)
        .select(args.arg_selection);

    let mut rdr = rconfig.reader()?;
    let mut wtr = Config::new(&args.flag_output).writer()?;

    let headers = rdr.byte_headers()?.clone();
    let sel = if args.flag_random {
        // Use seed if it is provided when initializing the random number generator.
        let mut rng = if let Some(seed) = args.flag_seed {
            // we add the DevSkim ignore comment here because we don't need to worry about
            // cryptographic security in this context.
            rand::rngs::StdRng::seed_from_u64(seed) // DevSkim: ignore DS148264
        } else {
            rand::rngs::StdRng::from_entropy()
        };

        let initial_selection = rconfig.selection(&headers)?;

        // make a vector of the column indices (1-indexed).
        let mut shuffled_selection_vec: Vec<usize> =
            initial_selection.iter().map(|&i| i + 1).collect();

        shuffled_selection_vec.shuffle(&mut rng);

        // Convert the shuffled indices into a comma-separated string.
        let shuffled_selection_string = shuffled_selection_vec
            .into_iter()
            .map(|i| i.to_string())
            .collect::<Vec<String>>()
            .join(",");

        // Parse the shuffled string into a SelectColumns object.
        let shuffled_selection = SelectColumns::parse(&shuffled_selection_string)?;

        rconfig
            .clone()
            .select(shuffled_selection)
            .selection(&headers)?
    } else if args.flag_sort {
        // get the headers from the initial selection
        let initial_selection = rconfig.selection(&headers)?;
        let mut initial_headers_vec = initial_selection
            .iter()
            .map(|&i| &headers[i])
            .collect::<Vec<&[u8]>>();

        // sort the headers lexicographically
        initial_headers_vec.sort_unstable();

        // make a comma-separated string of the sorted, quoted headers
        let sorted_selection_string = initial_headers_vec
            .iter()
            .map(|h| format!("\"{}\"", String::from_utf8_lossy(h)))
            .collect::<Vec<String>>()
            .join(",");

        // Parse the sorted selection string into a SelectColumns object.
        let sorted_selection = SelectColumns::parse(&sorted_selection_string)?;

        rconfig
            .clone()
            .select(sorted_selection)
            .selection(&headers)?
    } else {
        rconfig.selection(&headers)?
    };

    if !rconfig.no_headers {
        wtr.write_record(sel.iter().map(|&i| &headers[i]))?;
    }
    let mut record = csv::ByteRecord::new();
    while rdr.read_byte_record(&mut record)? {
        wtr.write_record(sel.iter().map(|&i| &record[i]))?;
    }
    wtr.flush()?;
    Ok(())
}