frep-core 0.1.7

use std::io::BufRead;

#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
pub enum LineEnding {
    /// No line ending (typically the last line of a file)
    None,
    /// Unix/Linux/macOS line ending (`\n`)
    Lf,
    /// Windows line ending (`\r\n`)
    CrLf,
}

impl LineEnding {
    #[inline]
    pub fn as_str(self) -> &'static str {
        match self {
            LineEnding::None => "",
            LineEnding::Lf => "\n",
            LineEnding::CrLf => "\r\n",
        }
    }

    #[inline]
    pub fn as_bytes(self) -> &'static [u8] {
        match self {
            LineEnding::None => b"",
            LineEnding::Lf => b"\n",
            LineEnding::CrLf => b"\r\n",
        }
    }
}

/// An iterator that reads lines from a `BufRead` source while preserving line endings.
///
/// Unlike the standard library's `lines()` iterator which strips line endings,
/// this iterator returns tuples of `(content, line_ending)` where the content is
/// returned as bytes and the line ending is preserved as a separate enum value.
///
/// Callers are responsible for UTF-8 validation if they need to work with the content
/// as text. When the content is known to be valid UTF-8, it can be converted using
/// `String::from_utf8()` or `String::from_utf8_lossy()`.
pub struct LinesSplitEndings<R> {
    reader: R,
    buffer: Vec<u8>,
}

impl<R: BufRead> LinesSplitEndings<R> {
    /// Creates a new `LinesSplitEndings` iterator from any type that implements `BufRead`.
    pub fn new(reader: R) -> Self {
        Self {
            reader,
            buffer: vec![],
        }
    }
}

impl<R: BufRead> Iterator for LinesSplitEndings<R> {
    type Item = std::io::Result<(Vec<u8>, LineEnding)>;

    fn next(&mut self) -> Option<Self::Item> {
        self.buffer.clear();
        match self.reader.read_until(b'\n', &mut self.buffer) {
            Ok(0) => None, // EOF
            Ok(_) => {
                let (content, ending) = split_line_ending(&self.buffer);
                Some(Ok((content.to_vec(), ending)))
            }
            Err(e) => Some(Err(e)),
        }
    }
}

/// Extension trait that adds the `lines_with_endings()` method to any `BufRead` implementation.
///
/// # Examples
///
/// ```
/// use std::io::Cursor;
/// use frep_core::line_reader::BufReadExt;
///
/// let cursor = Cursor::new(b"hello\nworld\r\n");
///
/// for line_result in cursor.lines_with_endings() {
///     let (content, ending) = line_result?;
///     println!("Content: '{}', Ending: '{:?}'", String::from_utf8_lossy(&content), ending);
/// }
/// # Ok::<(), std::io::Error>(())
/// ```
pub trait BufReadExt: BufRead {
    /// Returns an iterator that yields lines with their endings preserved.
    ///
    /// Each item yielded by the iterator is a `Result<(Vec<u8>, LineEnding), io::Error>`
    /// where the first element is the line content as bytes and the second is the line ending type.
    fn lines_with_endings(self) -> LinesSplitEndings<Self>
    where
        Self: Sized,
    {
        LinesSplitEndings::new(self)
    }
}

impl<R: BufRead> BufReadExt for R {}

/// Splits a line into its content and line ending parts.
///
/// # Examples
///
/// ```
/// use frep_core::line_reader::{split_line_ending, LineEnding};
///
/// assert_eq!(split_line_ending(b"hello\n"), (&b"hello"[..], LineEnding::Lf));
/// assert_eq!(split_line_ending(b"hello\r\n"), (&b"hello"[..], LineEnding::CrLf));
/// assert_eq!(split_line_ending(b"hello"), (&b"hello"[..], LineEnding::None));
/// ```
#[inline]
pub fn split_line_ending(line: &[u8]) -> (&[u8], LineEnding) {
    let len = line.len();
    if len == 0 {
        return (line, LineEnding::None);
    }

    if line[len - 1] == b'\n' {
        if len >= 2 && line[len - 2] == b'\r' {
            (&line[..len - 2], LineEnding::CrLf)
        } else {
            (&line[..len - 1], LineEnding::Lf)
        }
    } else {
        (line, LineEnding::None)
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use std::io::Cursor;

    #[test]
    fn test_split_line_ending_empty() {
        assert_eq!(split_line_ending(b""), ("".as_bytes(), LineEnding::None));
    }

    #[test]
    fn test_split_line_ending_no_ending() {
        assert_eq!(
            split_line_ending(b"hello world"),
            ("hello world".as_bytes(), LineEnding::None)
        );
    }

    #[test]
    fn test_split_line_ending_lf() {
        assert_eq!(
            split_line_ending("hello\n".as_bytes()),
            ("hello".as_bytes(), LineEnding::Lf)
        );
        assert_eq!(
            split_line_ending("\n".as_bytes()),
            ("".as_bytes(), LineEnding::Lf)
        );
    }

    #[test]
    fn test_split_line_ending_crlf() {
        assert_eq!(
            split_line_ending("hello\r\n".as_bytes()),
            ("hello".as_bytes(), LineEnding::CrLf)
        );
        assert_eq!(
            split_line_ending("\r\n".as_bytes()),
            ("".as_bytes(), LineEnding::CrLf)
        );
    }

    #[test]
    fn test_split_line_ending_unicode() {
        assert_eq!(
            split_line_ending("héllo 世界\n".as_bytes()),
            ("héllo 世界".as_bytes(), LineEnding::Lf)
        );
        assert_eq!(
            split_line_ending("héllo 世界\r\n".as_bytes()),
            ("héllo 世界".as_bytes(), LineEnding::CrLf)
        );
    }

    #[test]
    fn test_lines_split_endings_empty() {
        let cursor = Cursor::new("");
        let mut lines = LinesSplitEndings::new(cursor);
        assert!(lines.next().is_none());
    }

    #[test]
    fn test_lines_split_endings_single_line_no_ending() {
        let cursor = Cursor::new("hello");
        let mut lines = LinesSplitEndings::new(cursor);

        let result = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result.0).unwrap(), "hello");
        assert_eq!(result.1, LineEnding::None);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_lines_split_endings_single_line_with_lf() {
        let cursor = Cursor::new("hello\n");
        let mut lines = LinesSplitEndings::new(cursor);

        let result = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result.0).unwrap(), "hello");
        assert_eq!(result.1, LineEnding::Lf);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_lines_split_endings_multiple_lines_mixed() {
        let cursor = Cursor::new("line1\nline2\r\nline3\n\nline5");
        let mut lines = LinesSplitEndings::new(cursor);

        let result1 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result1.0).unwrap(), "line1");
        assert_eq!(result1.1, LineEnding::Lf);

        let result2 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result2.0).unwrap(), "line2");
        assert_eq!(result2.1, LineEnding::CrLf);

        let result3 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result3.0).unwrap(), "line3");
        assert_eq!(result3.1, LineEnding::Lf);

        let result4 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result4.0).unwrap(), "");
        assert_eq!(result4.1, LineEnding::Lf);

        let result5 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result5.0).unwrap(), "line5");
        assert_eq!(result5.1, LineEnding::None);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_lines_split_endings_empty_lines() {
        let cursor = Cursor::new("\n\r\n\r\n");
        let mut lines = LinesSplitEndings::new(cursor);

        let result1 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result1.0).unwrap(), "");
        assert_eq!(result1.1, LineEnding::Lf);

        let result2 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result2.0).unwrap(), "");
        assert_eq!(result2.1, LineEnding::CrLf);

        let result3 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result3.0).unwrap(), "");
        assert_eq!(result3.1, LineEnding::CrLf);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_buf_read_ext_trait() {
        let cursor = Cursor::new("hello\nworld\r\n");
        let mut lines = cursor.lines_with_endings();

        let result1 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result1.0).unwrap(), "hello");
        assert_eq!(result1.1, LineEnding::Lf);

        let result2 = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result2.0).unwrap(), "world");
        assert_eq!(result2.1, LineEnding::CrLf);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_large_line() {
        let content = "a".repeat(10000);
        let line = format!("{content}\n");
        let cursor = Cursor::new(line);
        let mut lines = LinesSplitEndings::new(cursor);

        let result = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result.0).unwrap(), content);
        assert_eq!(result.1, LineEnding::Lf);

        assert!(lines.next().is_none());
    }

    #[test]
    fn test_unicode_content() {
        let content = "Hello 世界 🦀 Rust";
        let line = format!("{content}\r\n");
        let cursor = Cursor::new(line);
        let mut lines = LinesSplitEndings::new(cursor);

        let result = lines.next().unwrap().unwrap();
        assert_eq!(String::from_utf8(result.0).unwrap(), content);
        assert_eq!(result.1, LineEnding::CrLf);

        assert!(lines.next().is_none());
    }
}