nushell/crates/nu-cli/src/commands/group_by.rs

221 lines
8.8 KiB
Rust
Raw Normal View History

2019-10-21 01:42:07 +02:00
use crate::commands::WholeStreamCommand;
use crate::prelude::*;
use indexmap::indexmap;
use nu_errors::ShellError;
use nu_protocol::{ReturnSuccess, Signature, SyntaxShape, UntaggedValue, Value};
use nu_source::Tagged;
2019-10-21 01:42:07 +02:00
pub struct GroupBy;
#[derive(Deserialize)]
pub struct GroupByArgs {
column_name: Option<Tagged<String>>,
2019-10-21 01:42:07 +02:00
}
2020-05-29 10:22:52 +02:00
#[async_trait]
2019-10-21 01:42:07 +02:00
impl WholeStreamCommand for GroupBy {
fn name(&self) -> &str {
"group-by"
}
fn signature(&self) -> Signature {
Signature::build("group-by").optional(
"column_name",
SyntaxShape::String,
"the name of the column to group by",
)
2019-10-21 01:42:07 +02:00
}
fn usage(&self) -> &str {
"Creates a new table with the data from the table rows grouped by the column given."
}
2020-05-29 10:22:52 +02:00
async fn run(
2019-10-21 01:42:07 +02:00
&self,
args: CommandArgs,
registry: &CommandRegistry,
) -> Result<OutputStream, ShellError> {
group_by(args, registry).await
2019-10-21 01:42:07 +02:00
}
fn examples(&self) -> Vec<Example> {
vec![
Example {
description: "Group items by type",
example: r#"ls | group-by type"#,
result: None,
},
Example {
description: "Group items by their value",
example: "echo [1 3 1 3 2 1 1] | group-by",
result: Some(vec![UntaggedValue::row(indexmap! {
"1".to_string() => UntaggedValue::Table(vec![
UntaggedValue::int(1).into(),
UntaggedValue::int(1).into(),
UntaggedValue::int(1).into(),
UntaggedValue::int(1).into(),
]).into(),
"3".to_string() => UntaggedValue::Table(vec![
UntaggedValue::int(3).into(),
UntaggedValue::int(3).into(),
]).into(),
"2".to_string() => UntaggedValue::Table(vec![
UntaggedValue::int(2).into(),
]).into(),
})
.into()]),
},
]
}
2019-10-21 01:42:07 +02:00
}
pub async fn group_by(
args: CommandArgs,
registry: &CommandRegistry,
) -> Result<OutputStream, ShellError> {
let registry = registry.clone();
let name = args.call_info.name_tag.clone();
let (GroupByArgs { column_name }, input) = args.process(&registry).await?;
let values: Vec<Value> = input.collect().await;
if values.is_empty() {
Err(ShellError::labeled_error(
"Expected table from pipeline",
"requires a table input",
name,
))
} else {
match crate::utils::data::group(column_name, &values, None, &name) {
Ok(grouped) => Ok(OutputStream::one(ReturnSuccess::value(grouped))),
Err(err) => Err(err),
2019-10-21 01:42:07 +02:00
}
}
}
pub fn group(
column_name: &Tagged<String>,
values: Vec<Value>,
tag: impl Into<Tag>,
) -> Result<Value, ShellError> {
crate::utils::data::group(Some(column_name.clone()), &values, None, tag)
}
#[cfg(test)]
mod tests {
use crate::commands::group_by::group;
use indexmap::IndexMap;
use nu_errors::ShellError;
use nu_protocol::{UntaggedValue, Value};
use nu_source::*;
fn string(input: impl Into<String>) -> Value {
UntaggedValue::string(input.into()).into_untagged_value()
}
fn row(entries: IndexMap<String, Value>) -> Value {
UntaggedValue::row(entries).into_untagged_value()
}
fn table(list: &[Value]) -> Value {
UntaggedValue::table(list).into_untagged_value()
}
fn nu_releases_commiters() -> Vec<Value> {
2019-11-12 08:07:43 +01:00
vec![
row(
indexmap! {"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("August 23-2019")},
),
row(
indexmap! {"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("August 23-2019")},
),
row(
indexmap! {"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("October 10-2019")},
),
row(
indexmap! {"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("Sept 24-2019")},
),
row(
indexmap! {"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("October 10-2019")},
),
row(
indexmap! {"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("Sept 24-2019")},
),
row(
indexmap! {"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("October 10-2019")},
),
row(
indexmap! {"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("Sept 24-2019")},
),
row(
indexmap! {"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("August 23-2019")},
),
2019-11-12 08:07:43 +01:00
]
}
#[test]
fn groups_table_by_date_column() -> Result<(), ShellError> {
2019-11-12 08:07:43 +01:00
let for_key = String::from("date").tagged_unknown();
assert_eq!(
group(&for_key, nu_releases_commiters(), Tag::unknown())?,
row(indexmap! {
"August 23-2019".into() => table(&[
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("August 23-2019")}),
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("August 23-2019")}),
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("August 23-2019")})
]),
"October 10-2019".into() => table(&[
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("October 10-2019")}),
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("October 10-2019")}),
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("October 10-2019")})
]),
"Sept 24-2019".into() => table(&[
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("Sept 24-2019")}),
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("Sept 24-2019")}),
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("Sept 24-2019")})
]),
})
);
Ok(())
}
2019-11-12 08:07:43 +01:00
#[test]
fn groups_table_by_country_column() -> Result<(), ShellError> {
2019-11-12 08:07:43 +01:00
let for_key = String::from("country").tagged_unknown();
assert_eq!(
group(&for_key, nu_releases_commiters(), Tag::unknown())?,
2019-11-12 08:07:43 +01:00
row(indexmap! {
"EC".into() => table(&[
2019-11-12 08:07:43 +01:00
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("August 23-2019")}),
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("Sept 24-2019")}),
row(indexmap!{"name".into() => string("AR"), "country".into() => string("EC"), "date".into() => string("October 10-2019")})
]),
"NZ".into() => table(&[
2019-11-12 08:07:43 +01:00
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("August 23-2019")}),
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("October 10-2019")}),
row(indexmap!{"name".into() => string("JT"), "country".into() => string("NZ"), "date".into() => string("Sept 24-2019")})
]),
"US".into() => table(&[
2019-11-12 08:07:43 +01:00
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("October 10-2019")}),
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("Sept 24-2019")}),
row(indexmap!{"name".into() => string("YK"), "country".into() => string("US"), "date".into() => string("August 23-2019")}),
]),
})
);
Ok(())
2019-11-12 08:07:43 +01:00
}
#[test]
fn examples_work_as_expected() {
use super::GroupBy;
use crate::examples::test as test_examples;
test_examples(GroupBy {})
}
2019-10-21 01:42:07 +02:00
}