rdbms-playground/src/input_render.rs

//! Render-side helpers for the input panel
//! (ADR-0022 stage 2 — token-class colouring of the live
//! input field with cursor injection).
//!
//! The functions here are pure: given an input string, a
//! cursor byte position, and a theme, they return a sequence
//! of `StyledRun`s describing each contiguous span with its
//! ratatui style. `ui::render_input_panel` converts these to
//! `Span<'_>`s at render time.
//!
//! Cursor handling:
//!   - Cursor inside a token splits that token's run into
//!     before/under/after, with `under` carrying the token's
//!     colour plus `Modifier::REVERSED`.
//!   - Cursor on a whitespace gap between tokens splits the
//!     gap the same way.
//!   - Cursor at end-of-input is represented as a trailing
//!     run with empty byte range; the renderer treats that as
//!     "inverted space".
//!
//! Per ADR-0022 §2/§3, this is the silent always-on layer.
//! The error overlay (stage 4) and hint panel (stage 5)
//! compose with these runs without fighting them.

use ratatui::style::{Modifier, Style};

use crate::dsl::lexer::lex;
use crate::theme::Theme;

/// A run of text with its byte range in the source and the
/// ratatui style it should render with. The text itself is
/// not stored — callers slice `source[byte_range.0..byte_range.1]`.
///
/// An empty byte range (`(n, n)`) represents the end-of-input
/// cursor and is rendered as an inverted space.
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct StyledRun {
    pub byte_range: (usize, usize),
    pub style: Style,
}

impl StyledRun {
    /// The text this run covers in `source`. Empty for the
    /// end-of-input cursor sentinel.
    #[must_use]
    pub fn text<'a>(&self, source: &'a str) -> &'a str {
        &source[self.byte_range.0..self.byte_range.1]
    }
}

/// Build the run sequence for the input panel.
///
/// Lexes `input`, assigns each token its `theme.token_color`,
/// preserves whitespace gaps as `theme.fg` runs, then injects
/// the cursor at `cursor_byte` (clamped to `input.len()`).
#[must_use]
pub fn render_input_runs(input: &str, cursor_byte: usize, theme: &Theme) -> Vec<StyledRun> {
    let mut runs = lex_to_runs(input, theme);
    inject_cursor(&mut runs, input, cursor_byte, theme);
    runs
}

/// Cursor-less variant: tokenises `input` into styled runs
/// covering the full byte range, with no inverted cursor.
/// Used by the echo-line renderer (ADR-0022 §5) where there's
/// no cursor to show.
#[must_use]
pub fn lex_to_runs(input: &str, theme: &Theme) -> Vec<StyledRun> {
    base_runs(input, theme)
}

fn base_runs(input: &str, theme: &Theme) -> Vec<StyledRun> {
    let tokens = lex(input);
    let mut runs = Vec::with_capacity(tokens.len() * 2);
    let mut pos = 0;
    for tok in tokens {
        let (start, end) = tok.span;
        if pos < start {
            // Whitespace gap before this token.
            runs.push(StyledRun {
                byte_range: (pos, start),
                style: Style::default().fg(theme.fg),
            });
        }
        runs.push(StyledRun {
            byte_range: (start, end),
            style: Style::default().fg(theme.token_color(&tok.kind)),
        });
        pos = end;
    }
    if pos < input.len() {
        runs.push(StyledRun {
            byte_range: (pos, input.len()),
            style: Style::default().fg(theme.fg),
        });
    }
    runs
}

fn inject_cursor(
    runs: &mut Vec<StyledRun>,
    input: &str,
    cursor_byte: usize,
    theme: &Theme,
) {
    let cursor_byte = cursor_byte.min(input.len());

    // End-of-input cursor: append the empty-range sentinel.
    if cursor_byte == input.len() {
        runs.push(StyledRun {
            byte_range: (input.len(), input.len()),
            style: Style::default()
                .fg(theme.fg)
                .add_modifier(Modifier::REVERSED),
        });
        return;
    }

    let idx = runs
        .iter()
        .position(|r| r.byte_range.0 <= cursor_byte && cursor_byte < r.byte_range.1)
        .expect("cursor_byte < input.len() ⇒ some run contains it");
    let target = runs[idx].clone();
    let (start, end) = target.byte_range;

    // Walk to the next char boundary so a multi-byte UTF-8
    // codepoint is treated as a single visual unit at the
    // cursor.
    let mut char_end = cursor_byte + 1;
    while char_end < input.len() && !input.is_char_boundary(char_end) {
        char_end += 1;
    }

    let mut replacement: Vec<StyledRun> = Vec::with_capacity(3);
    if start < cursor_byte {
        replacement.push(StyledRun {
            byte_range: (start, cursor_byte),
            style: target.style,
        });
    }
    replacement.push(StyledRun {
        byte_range: (cursor_byte, char_end),
        style: target.style.add_modifier(Modifier::REVERSED),
    });
    if char_end < end {
        replacement.push(StyledRun {
            byte_range: (char_end, end),
            style: target.style,
        });
    }

    runs.splice(idx..=idx, replacement);
}

#[cfg(test)]
mod tests {
    use super::*;
    use pretty_assertions::assert_eq;

    fn dark() -> Theme {
        Theme::dark()
    }

    fn reversed(r: &StyledRun) -> bool {
        r.style.add_modifier.contains(Modifier::REVERSED)
    }

    #[test]
    fn empty_input_renders_only_the_end_of_input_cursor() {
        let runs = render_input_runs("", 0, &dark());
        assert_eq!(runs.len(), 1);
        assert_eq!(runs[0].byte_range, (0, 0));
        assert!(reversed(&runs[0]));
    }

    #[test]
    fn keyword_token_takes_keyword_colour() {
        let theme = dark();
        let runs = render_input_runs("create", 6, &theme);
        // Token + end-of-input cursor.
        assert_eq!(runs.len(), 2);
        assert_eq!(runs[0].byte_range, (0, 6));
        assert_eq!(runs[0].style.fg, Some(theme.tok_keyword));
        assert!(!reversed(&runs[0]));
        assert!(reversed(&runs[1]));
    }

    #[test]
    fn cursor_inside_token_splits_into_three_runs_keeping_colour() {
        let theme = dark();
        let runs = render_input_runs("create", 3, &theme);
        assert_eq!(runs.len(), 3);
        assert_eq!(runs[0].byte_range, (0, 3));
        assert_eq!(runs[1].byte_range, (3, 4));
        assert_eq!(runs[2].byte_range, (4, 6));
        // All three keep the keyword colour.
        for r in &runs {
            assert_eq!(r.style.fg, Some(theme.tok_keyword));
        }
        assert!(!reversed(&runs[0]));
        assert!(reversed(&runs[1]));
        assert!(!reversed(&runs[2]));
    }

    #[test]
    fn cursor_on_whitespace_inverts_a_single_space() {
        let theme = dark();
        // "create table" has whitespace at byte 6.
        let runs = render_input_runs("create table", 6, &theme);
        // base: keyword, ws(6,7), keyword. After cursor injection
        // at the start of ws: under=(6,7) REVERSED. The
        // before/after slices are empty so we get 3 runs total.
        assert_eq!(runs.len(), 3);
        let r_under: Vec<_> = runs.iter().filter(|r| reversed(r)).collect();
        assert_eq!(r_under.len(), 1);
        assert_eq!(r_under[0].byte_range, (6, 7));
        assert_eq!(r_under[0].style.fg, Some(theme.fg));
    }

    #[test]
    fn lex_error_token_renders_in_error_colour() {
        let theme = dark();
        let runs = render_input_runs("$", 1, &theme);
        // Error token (0,1), then end-of-input cursor (1,1).
        assert_eq!(runs.len(), 2);
        assert_eq!(runs[0].style.fg, Some(theme.tok_error));
    }

    #[test]
    fn whitespace_between_tokens_takes_default_fg() {
        let theme = dark();
        let runs = render_input_runs("create table", 12, &theme);
        // base: keyword(0,6), ws(6,7), keyword(7,12). Plus
        // end-of-input cursor (12,12) = 4 runs.
        assert_eq!(runs.len(), 4);
        assert_eq!(runs[1].byte_range, (6, 7));
        assert_eq!(runs[1].style.fg, Some(theme.fg));
        assert_eq!(runs[3].byte_range, (12, 12));
        assert!(reversed(&runs[3]));
    }

    #[test]
    fn cursor_inside_multi_byte_string_literal_advances_to_char_boundary() {
        let theme = dark();
        // 'café' = ['(0)', c(1), a(2), f(3), é(4-5), '(6)] — é is 2 bytes.
        // Cursor at byte 4: inside é. char_end advances to 6.
        let runs = render_input_runs("'café'", 4, &theme);
        let r_under: Vec<_> = runs.iter().filter(|r| reversed(r)).collect();
        assert_eq!(r_under.len(), 1);
        assert_eq!(r_under[0].byte_range, (4, 6));
    }

    #[test]
    fn end_of_input_cursor_is_an_empty_range() {
        let runs = render_input_runs("create", 6, &dark());
        let last = runs.last().expect("non-empty");
        assert_eq!(last.byte_range, (6, 6));
        assert!(reversed(last));
    }

    #[test]
    fn full_command_lexes_to_each_token_class() {
        let theme = dark();
        let runs = render_input_runs(
            "insert into T values (1, 'hi', null) --all-rows",
            47,
            &theme,
        );
        // Spot-check: there's at least one run of each token class.
        let fgs: Vec<_> = runs.iter().filter_map(|r| r.style.fg).collect();
        assert!(fgs.contains(&theme.tok_keyword)); // insert / into / values / null
        assert!(fgs.contains(&theme.tok_identifier)); // T / hi (string is separate)
        assert!(fgs.contains(&theme.tok_number)); // 1
        assert!(fgs.contains(&theme.tok_string)); // 'hi'
        assert!(fgs.contains(&theme.tok_punct)); // ( , )
        assert!(fgs.contains(&theme.tok_flag)); // --all-rows
    }
}