extern/nushell
1
0
Fork 1
mirror of https://github.com/nushell/nushell.git synced 2025-01-23 06:39:17 +01:00
Code Issues Packages Projects Releases Wiki Activity

Ensure that two columns named index don't exist when converting a Dataframe to a nu Value. (#12501)

Browse Source 
# Description
@maxim-uvarov discovered an issue with the current implementation. When
executing [[index a]; [1 1]] | polars into-df, a plugin_failed_to_decode
error occurs. This happens because a Record is created with two columns
named "index" as an index column is added during conversion. This pull
request addresses the problem by not adding an index column if there is
already a column named "index" in the dataframe.

---------

Co-authored-by: Jack Wright <jack.wright@disqo.com>
This commit is contained in:
Jack Wright 2024-04-13 04:33:29 -07:00 committed by GitHub
parent f975c9923a
commit 1bded8572c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 18 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
crates/nu_plugin_polars/src/dataframe/values/nu_dataframe/conversion.rs
View File

@ -154,12 +154,16 @@ pub fn create_column(
// Adds a separator to the vector of values using the column names from the
// dataframe to create the Values Row
pub fn add_separator(values: &mut Vec<Value>, df: &DataFrame, span: Span) {
// returns true if there is an index column contained in the dataframe
pub fn add_separator(values: &mut Vec<Value>, df: &DataFrame, has_index: bool, span: Span) {
let mut record = Record::new();
record.push("index", Value::string("...", span));
if !has_index {
record.push("index", Value::string("...", span));
}
for name in df.get_column_names() {
// there should only be one index field
record.push(name, Value::string("...", span))
}

16
crates/nu_plugin_polars/src/dataframe/values/nu_dataframe/mod.rs
View File

@ -296,6 +296,13 @@ impl NuDataFrame {
}
}
pub fn has_index(&self) -> bool {
self.columns(Span::unknown())
.unwrap_or_default() // just assume there isn't an index
.iter()
.any(|col| col.name() == "index")
}
// Print is made out a head and if the dataframe is too large, then a tail
pub fn print(&self, span: Span) -> Result<Vec<Value>, ShellError> {
let df = &self.df;
@ -304,7 +311,7 @@ impl NuDataFrame {
if df.height() > size {
let sample_size = size / 2;
let mut values = self.head(Some(sample_size), span)?;
conversion::add_separator(&mut values, df, span);
conversion::add_separator(&mut values, df, self.has_index(), span);
let remaining = df.height() - sample_size;
let tail_size = remaining.min(sample_size);
let mut tail_values = self.tail(Some(tail_size), span)?;
@ -323,7 +330,6 @@ impl NuDataFrame {
pub fn head(&self, rows: Option<usize>, span: Span) -> Result<Vec<Value>, ShellError> {
let to_row = rows.unwrap_or(5);
let values = self.to_rows(0, to_row, span)?;
Ok(values)
}
@ -334,7 +340,6 @@ impl NuDataFrame {
let from_row = to_row.saturating_sub(size);
let values = self.to_rows(from_row, to_row, span)?;
Ok(values)
}
@ -368,11 +373,14 @@ impl NuDataFrame {
.map(|col| (col.name().to_string(), col.into_iter()))
.collect::<Vec<(String, std::vec::IntoIter<Value>)>>();
let has_index = self.has_index();
let values = (0..size)
.map(|i| {
let mut record = Record::new();
record.push("index", Value::int((i + from_row) as i64, span));
if !has_index {
record.push("index", Value::int((i + from_row) as i64, span));
}
for (name, col) in &mut iterators {
record.push(name.clone(), col.next().unwrap_or(Value::nothing(span)));