badness 0.3.0

//! `title-capitalization`: an unprotected acronym (or mid-word capital) in a
//! title-like field.
//!
//! Many bibliography styles lowercase title text that is not brace-protected, so an
//! acronym like `DNA` renders as `dna` unless written `{DNA}`. This rule flags
//! capitals that such a style would clobber. A [`Severity::Warning`]; report-only
//! (wrapping in braces is a meaning-preserving edit, but choosing *what* to protect
//! is the author's call).
//!
//! **Acronym heuristic (not strict).** Only capitals likely to be acronyms are
//! flagged, to stay quiet on ordinary Title-Cased databases:
//! - a run of **two or more** consecutive capitals (`DNA`, `LaTeX`'s `L`+`T`… —
//!   any `[A-Z]{2,}` run), or
//! - a **single** capital in the *middle* of a word (`iPhone`'s `P`),
//!
//! A lone capital starting a word (`The`, `Quick`) is left alone. Only text at
//! brace depth 0 inside the value is considered — content already inside a nested
//! `{…}` group is protected and skipped. Bare `LITERAL` pieces (macros/numbers) are
//! not scanned.

use std::path::PathBuf;

use crate::bib::ast::{field_name, field_value};
use crate::bib::syntax::{SyntaxElement, SyntaxKind, SyntaxNode};
use crate::linter::diagnostic::{Diagnostic, Severity};

use super::{BibRule, BibRuleContext};

/// Field names whose value is title-like prose subject to style lowercasing.
const TITLE_FIELDS: &[&str] = &[
    "title",
    "subtitle",
    "booktitle",
    "booksubtitle",
    "maintitle",
    "mainsubtitle",
    "journaltitle",
    "issuetitle",
    "issuesubtitle",
    "eventtitle",
    "shorttitle",
];

pub struct TitleCapitalization;

impl BibRule for TitleCapitalization {
    fn id(&self) -> &'static str {
        "title-capitalization"
    }

    fn default_severity(&self) -> Severity {
        Severity::Warning
    }

    fn interests(&self) -> &'static [SyntaxKind] {
        &[SyntaxKind::FIELD]
    }

    fn check(&self, el: &SyntaxElement, _ctx: &BibRuleContext<'_>, sink: &mut Vec<Diagnostic>) {
        let Some(field) = el.as_node() else {
            return;
        };
        let Some(name) = field_name(field) else {
            return;
        };
        let name_lc = name.to_lowercase();
        if !TITLE_FIELDS.contains(&name_lc.as_str()) {
            return;
        }
        let Some(value) = field_value(field) else {
            return;
        };
        for piece in value.children() {
            let (inner, base) = match piece.kind() {
                SyntaxKind::BRACE_GROUP => match inner_of(&piece, '{', '}') {
                    Some(parts) => parts,
                    None => continue,
                },
                SyntaxKind::QUOTED => match inner_of(&piece, '"', '"') {
                    Some(parts) => parts,
                    None => continue,
                },
                // Bare LITERAL pieces are macros/numbers, not title prose.
                _ => continue,
            };
            for (start, end, run) in unprotected_acronyms(&inner) {
                sink.push(Diagnostic {
                    rule: self.id(),
                    severity: self.default_severity(),
                    path: PathBuf::new(),
                    start: base + start,
                    end: base + end,
                    message: format!(
                        "unprotected capitals `{run}` in `{name_lc}`; wrap in braces (`{{{run}}}`) \
                         to keep case under title-lowercasing styles"
                    ),
                    fix: None,
                });
            }
        }
    }
}

/// The inner text of a delimited piece (between `open` and `close`) and the byte
/// offset of that inner text in the document. Returns `None` for a piece missing
/// its closing delimiter (a recovery artifact) so a partial group is not scanned.
fn inner_of(node: &SyntaxNode, open: char, close: char) -> Option<(String, usize)> {
    let text = node.to_string();
    let stripped = text.strip_prefix(open)?.strip_suffix(close)?;
    let base = usize::from(node.text_range().start()) + open.len_utf8();
    Some((stripped.to_string(), base))
}

/// Find the byte ranges (relative to the start of `text`) of acronym-like capital
/// runs at brace depth 0, with the matched substring. See the module docs for the
/// heuristic.
fn unprotected_acronyms(text: &str) -> Vec<(usize, usize, String)> {
    let mut hits = Vec::new();
    let mut depth: i32 = 0;
    // Byte offset of the previous char and whether it was an alphabetic letter at
    // depth 0 (to detect a mid-word single capital).
    let mut prev_alpha = false;
    let chars: Vec<(usize, char)> = text.char_indices().collect();
    let mut i = 0;
    while i < chars.len() {
        let (off, ch) = chars[i];
        match ch {
            '{' => {
                depth += 1;
                prev_alpha = false;
                i += 1;
                continue;
            }
            '}' => {
                depth -= 1;
                prev_alpha = false;
                i += 1;
                continue;
            }
            _ => {}
        }
        if depth == 0 && ch.is_ascii_uppercase() {
            // Extend a maximal run of consecutive ASCII capitals at this depth.
            let run_start = off;
            let prev_alpha_at_run = prev_alpha;
            let mut j = i;
            while j < chars.len() && chars[j].1.is_ascii_uppercase() {
                j += 1;
            }
            let run_len = j - i;
            let run_end = chars.get(j).map(|&(o, _)| o).unwrap_or(text.len());
            if run_len >= 2 || prev_alpha_at_run {
                hits.push((run_start, run_end, text[run_start..run_end].to_string()));
            }
            prev_alpha = true;
            i = j;
            continue;
        }
        prev_alpha = depth == 0 && ch.is_alphabetic();
        i += 1;
    }
    hits
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::bib::parse;
    use crate::bib::semantic::Model;

    fn findings(src: &str) -> Vec<Diagnostic> {
        let root = parse(src).syntax();
        let model = Model::build(&root);
        let ctx = BibRuleContext {
            path: std::path::Path::new("x.bib"),
            root: &root,
            model: &model,
            db: crate::bib::semantic::builtin(),
        };
        let mut out = Vec::new();
        for el in root.descendants_with_tokens() {
            if TitleCapitalization.interests().contains(&el.kind()) {
                TitleCapitalization.check(&el, &ctx, &mut out);
            }
        }
        out
    }

    #[test]
    fn flags_bare_acronym() {
        let out = findings("@article{k, title = {The DNA helix}}\n");
        assert_eq!(out.len(), 1);
        assert_eq!(out[0].rule, "title-capitalization");
        assert!(out[0].message.contains("DNA"));
    }

    #[test]
    fn braced_acronym_is_protected() {
        assert!(findings("@article{k, title = {The {DNA} helix}}\n").is_empty());
    }

    #[test]
    fn ordinary_title_case_is_fine() {
        assert!(findings("@article{k, title = {The Quick Brown Fox}}\n").is_empty());
    }

    #[test]
    fn flags_midword_capital() {
        let out = findings("@article{k, title = {The iPhone era}}\n");
        assert_eq!(out.len(), 1);
        assert!(out[0].message.contains('P'));
    }

    #[test]
    fn ignores_non_title_field() {
        assert!(findings("@article{k, author = {DNA Smith}}\n").is_empty());
    }

    #[test]
    fn flags_in_quoted_value() {
        let out = findings("@article{k, title = \"A DNA study\"}\n");
        assert_eq!(out.len(), 1);
    }

    #[test]
    fn quoted_with_braced_acronym_is_protected() {
        assert!(findings("@article{k, title = \"A {DNA} study\"}\n").is_empty());
    }

    #[test]
    fn underlines_the_run() {
        let src = "@article{k, title = {The DNA helix}}\n";
        let out = findings(src);
        assert_eq!(out.len(), 1);
        assert_eq!(&src[out[0].start..out[0].end], "DNA");
    }

    #[test]
    fn flags_each_acronym() {
        let out = findings("@article{k, title = {RNA and DNA}}\n");
        assert_eq!(out.len(), 2);
    }
}