Added expression support for polars contains (#13769)

# Description
Adds the ability to use `polars contains` as an expression:
<img width="785" alt="Screenshot 2024-09-03 at 14 39 03"
src="https://github.com/user-attachments/assets/35c0d4e5-6bef-4974-a31f-463c7203bd03">


# User-Facing Changes
- `polars contains` can now be used with expressions
This commit is contained in:
Jack Wright 2024-09-04 03:19:45 -07:00 committed by GitHub
parent eb0de25d19
commit 63b94dbd28
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -1,4 +1,9 @@
use crate::{values::CustomValueSupport, PolarsPlugin};
use crate::{
values::{
cant_convert_err, CustomValueSupport, NuExpression, PolarsPluginObject, PolarsPluginType,
},
PolarsPlugin,
};
use super::super::super::values::{Column, NuDataFrame};
@ -7,7 +12,7 @@ use nu_protocol::{
Category, Example, LabeledError, PipelineData, ShellError, Signature, Span, SyntaxShape, Type,
Value,
};
use polars::prelude::{IntoSeries, StringNameSpaceImpl};
use polars::prelude::{lit, IntoSeries, StringNameSpaceImpl};
#[derive(Clone)]
pub struct Contains;
@ -30,33 +35,62 @@ impl PluginCommand for Contains {
SyntaxShape::String,
"Regex pattern to be searched",
)
.input_output_type(
Type::Custom("dataframe".into()),
Type::Custom("dataframe".into()),
)
.input_output_types(vec![
(
Type::Custom("expression".into()),
Type::Custom("expression".into()),
),
(
Type::Custom("dataframe".into()),
Type::Custom("dataframe".into()),
),
])
.category(Category::Custom("dataframe".into()))
}
fn examples(&self) -> Vec<Example> {
vec![Example {
description: "Returns boolean indicating if pattern was found",
example: "[abc acb acb] | polars into-df | polars contains ab",
result: Some(
NuDataFrame::try_from_columns(
vec![Column::new(
"0".to_string(),
vec![
Value::test_bool(true),
Value::test_bool(false),
Value::test_bool(false),
],
)],
None,
)
.expect("simple df for test should not fail")
.into_value(Span::test_data()),
),
}]
vec![
Example {
description: "Returns boolean indicating if pattern was found in a column",
example: "let df = [[a]; [abc] [acb] [acb]] | polars into-df;
let df2 = $df | polars with-column [(polars col a | polars contains ab | polars as b)] | polars collect;
$df2.b",
result: Some(
NuDataFrame::try_from_columns(
vec![Column::new(
"b".to_string(),
vec![
Value::test_bool(true),
Value::test_bool(false),
Value::test_bool(false),
],
)],
None,
)
.expect("simple df for test should not fail")
.into_value(Span::test_data()),
),
},
Example {
description: "Returns boolean indicating if pattern was found",
example: "[abc acb acb] | polars into-df | polars contains ab",
result: Some(
NuDataFrame::try_from_columns(
vec![Column::new(
"0".to_string(),
vec![
Value::test_bool(true),
Value::test_bool(false),
Value::test_bool(false),
],
)],
None,
)
.expect("simple df for test should not fail")
.into_value(Span::test_data()),
),
},
]
}
fn run(
@ -66,17 +100,48 @@ impl PluginCommand for Contains {
call: &EvaluatedCall,
input: PipelineData,
) -> Result<PipelineData, LabeledError> {
command(plugin, engine, call, input).map_err(LabeledError::from)
let value = input.into_value(call.head)?;
match PolarsPluginObject::try_from_value(plugin, &value)? {
PolarsPluginObject::NuDataFrame(df) => command_df(plugin, engine, call, df),
PolarsPluginObject::NuLazyFrame(lazy) => {
let df = lazy.collect(call.head)?;
command_df(plugin, engine, call, df)
}
PolarsPluginObject::NuExpression(expr) => command_expr(plugin, engine, call, expr),
_ => Err(cant_convert_err(
&value,
&[
PolarsPluginType::NuDataFrame,
PolarsPluginType::NuLazyFrame,
PolarsPluginType::NuExpression,
],
)),
}
.map_err(LabeledError::from)
}
}
fn command(
fn command_expr(
plugin: &PolarsPlugin,
engine: &EngineInterface,
call: &EvaluatedCall,
input: PipelineData,
expr: NuExpression,
) -> Result<PipelineData, ShellError> {
let pattern: String = call.req(0)?;
let res: NuExpression = expr
.into_polars()
.str()
.contains(lit(pattern), false)
.into();
res.to_pipeline_data(plugin, engine, call.head)
}
fn command_df(
plugin: &PolarsPlugin,
engine: &EngineInterface,
call: &EvaluatedCall,
df: NuDataFrame,
) -> Result<PipelineData, ShellError> {
let df = NuDataFrame::try_from_pipeline_coerce(plugin, input, call.head)?;
let pattern: String = call.req(0)?;
let series = df.as_series(call.head)?;