Crate crlf_to_lf_inplace 

crlf-to-lf-inplace

Fast in-place CRLF to LF line ending conversion for Rust strings. Uses memchr for good performance without custom SIMD.

Features

• No-std compatible with optional std support
• Zero-allocation in-place conversion
• Fast CRLF detection using memchr
• UTF-8 safe conversion

Installation

Add this to your Cargo.toml:

[dependencies]
crlf-to-lf-inplace = "0.1.0"

No-std builds can disable the default std feature:

[dependencies]
crlf-to-lf-inplace = { version = "0.1.0", default-features = false }

Feature Flags

Feature	Description
std	Enable standard library support (disabled by default for no_std compatibility)

Usage

Use crlf_to_lf_inplace for String values, or crlf_to_lf_inplace_bytes for raw byte buffers. The byte-slice API returns the new logical length, so truncate or slice the buffer using the returned length.

Basic Example

use crlf_to_lf_inplace::crlf_to_lf_inplace;

let mut text = String::from("line1\r\nline2\r\n");
crlf_to_lf_inplace(&mut text);
assert_eq!(text, "line1\nline2\n");

Advanced Example

use crlf_to_lf_inplace::crlf_to_lf_inplace_bytes;

let mut buffer = b"foo\r\nbar\r\nbaz\r\n".to_vec();
let new_len = crlf_to_lf_inplace_bytes(&mut buffer);
buffer.truncate(new_len);
assert_eq!(buffer.as_slice(), b"foo\nbar\nbaz\n");

Performance

Benchmarked on AMD Ryzen 9 9950X3D (single-threaded) with ~71 KiB of concatenated Markdown files (29 files from real opencode configuration):

Input Type	Time	Throughput
CRLF (conversion needed)	9.72 µs	7.2 GiB/s
LF only (no conversion)	378 ns	179 GiB/s

The implementation uses memchr for fast \r scanning with a two-phase approach: first locating the initial CRLF, then processing remaining matches. Files already using LF line endings benefit from an early-exit path.

Run benchmarks with:

cargo bench

License

Licensed under MIT

Functions

crlf_to_lf_inplace

Convert CRLF sequences to LF in-place within a String.

crlf_to_lf_inplace_bytes

Convert CRLF sequences to LF in-place within a byte slice.