extern/nushell
1
0
Fork 1
mirror of https://github.com/nushell/nushell.git synced 2024-11-22 08:23:24 +01:00
Code Issues Packages Projects Releases Wiki Activity

from ssv --aligned-columns should separate lines by character index instead of byte index (#8558)

Browse Source 
# Description

## Symptom
Lines which are input into `from ssv --aligned-columns` are split
incorrectly of they contain utf-8 characters which have the length of
multiple bytes. Notice how the values of the `Bars` column bleeds into
the `Security` column in the following output (the big grey areas are
censored data ;) ):

![before-patch](https://user-images.githubusercontent.com/17351844/226757737-be7ca493-5c64-4a91-9153-984df515bb8c.png)

## Problem
The function behind `from ssv --aligned-columns` splits lines into
fields by byte index (which is default behavior of str.get(...) in Rust)
instead of character index. If the header row has a different length in
bytes than the remaining table rows, the split is executed incorrectly.

## Solution
The function behind `from ssv --aligned-columns1 now separates lines by
character index instead of byte index. This productes the following
(correct) output (the big grey areas are censored data ;) ):

![after-patch](https://user-images.githubusercontent.com/17351844/226757850-7acaebf3-2d40-4f85-b76e-64e465254bda.png)
This commit is contained in:
Christian Friedow 2023-03-22 23:54:18 +01:00 committed by GitHub
parent c48e9cdf5b
commit 66ad83c15c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 11 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
crates/nu-command/src/formats/from/ssv.rs
View File

@ -85,15 +85,23 @@ fn parse_aligned_columns<'a>(
.iter()
.enumerate()
.map(|(i, (header_name, start_position))| {
let char_index_start = match l.char_indices().nth(*start_position) {
Some(idx) => idx.0,
None => *start_position,
};
let val = match headers.get(i + 1) {
Some((_, end)) => {
if *end < l.len() {
l.get(*start_position..*end)
let char_index_end = match l.char_indices().nth(*end) {
Some(idx) => idx.0,
None => *end,
};
l.get(char_index_start..char_index_end)
} else {
l.get(*start_position..)
l.get(char_index_start..)
}
}
None => l.get(*start_position..),
None => l.get(char_index_start..),
}
.unwrap_or("")
.trim()