Expression support for polars str-slice (#13783)

# Description
Provides expression support for `polars str-slice`:
<img width="893" alt="Screenshot 2024-09-04 at 18 03 05"
src="https://github.com/user-attachments/assets/d8a8a2a7-53cf-4c3a-ae7a-dfdaf48a05ee">

# User-Facing Changes
- `polars str-slice` can now be used as an expression
This commit is contained in:
Jack Wright 2024-09-05 09:06:37 -07:00 committed by GitHub
parent abd230e12e
commit f611196373
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}; use super::super::super::values::{Column, NuDataFrame};
@ -8,7 +13,7 @@ use nu_protocol::{
Value, Value,
}; };
use polars::{ use polars::{
prelude::{IntoSeries, NamedFrom, StringNameSpaceImpl}, prelude::{lit, Expr, IntoSeries, NamedFrom, Null, StringNameSpaceImpl},
series::Series, series::Series,
}; };
@ -30,15 +35,40 @@ impl PluginCommand for StrSlice {
Signature::build(self.name()) Signature::build(self.name())
.required("start", SyntaxShape::Int, "start of slice") .required("start", SyntaxShape::Int, "start of slice")
.named("length", SyntaxShape::Int, "optional length", Some('l')) .named("length", SyntaxShape::Int, "optional length", Some('l'))
.input_output_type( .input_output_types(vec![
Type::Custom("dataframe".into()), (
Type::Custom("dataframe".into()), Type::Custom("expression".into()),
) Type::Custom("expression".into()),
),
(
Type::Custom("dataframe".into()),
Type::Custom("dataframe".into()),
),
])
.category(Category::Custom("dataframe".into())) .category(Category::Custom("dataframe".into()))
} }
fn examples(&self) -> Vec<Example> { fn examples(&self) -> Vec<Example> {
vec![ vec![
Example {
description: "Creates slices from the strings in a specified column",
example: "[[a]; [abcded] [abc321] [abc123]] | polars into-df | polars select (polars col a | polars str-slice 1 --length 2) | polars collect",
result: Some(
NuDataFrame::try_from_columns(
vec![Column::new(
"a".to_string(),
vec![
Value::test_string("bc"),
Value::test_string("bc"),
Value::test_string("bc"),
],
)],
None,
)
.expect("simple df for test should not fail")
.into_value(Span::test_data()),
),
},
Example { Example {
description: "Creates slices from the strings", description: "Creates slices from the strings",
example: "[abcded abc321 abc123] | polars into-df | polars str-slice 1 --length 2", example: "[abcded abc321 abc123] | polars into-df | polars str-slice 1 --length 2",
@ -87,15 +117,47 @@ impl PluginCommand for StrSlice {
call: &EvaluatedCall, call: &EvaluatedCall,
input: PipelineData, input: PipelineData,
) -> Result<PipelineData, LabeledError> { ) -> 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) => {
command_df(plugin, engine, call, lazy.collect(call.head)?)
}
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, plugin: &PolarsPlugin,
engine: &EngineInterface, engine: &EngineInterface,
call: &EvaluatedCall, call: &EvaluatedCall,
input: PipelineData, expr: NuExpression,
) -> Result<PipelineData, ShellError> {
let start = lit(call.req::<i64>(0)?);
let length: Expr = call
.get_flag::<i64>("length")?
.map(lit)
.unwrap_or(lit(Null {}));
let res: NuExpression = expr.into_polars().str().slice(start, length).into();
res.to_pipeline_data(plugin, engine, call.head)
}
fn command_df(
plugin: &PolarsPlugin,
engine: &EngineInterface,
call: &EvaluatedCall,
df: NuDataFrame,
) -> Result<PipelineData, ShellError> { ) -> Result<PipelineData, ShellError> {
let start: i64 = call.req(0)?; let start: i64 = call.req(0)?;
let start = Series::new("", &[start]); let start = Series::new("", &[start]);
@ -106,7 +168,6 @@ fn command(
None => Series::new_null("", 1), None => Series::new_null("", 1),
}; };
let df = NuDataFrame::try_from_pipeline_coerce(plugin, input, call.head)?;
let series = df.as_series(call.head)?; let series = df.as_series(call.head)?;
let chunked = series.str().map_err(|e| ShellError::GenericError { let chunked = series.str().map_err(|e| ShellError::GenericError {