ccd-cli 1.0.0-alpha.9

pub(crate) const SURFACED_REFERENCE_EXTENSIONS: &[&str] = &[
    ".rs", ".md", ".toml", ".json", ".sh", ".yml", ".yaml", ".py", ".ts", ".tsx", ".go", ".txt",
    ".sql", ".rb", ".java", ".c", ".cpp", ".h", ".hpp", ".html", ".css", ".lock",
];

pub(crate) const SURFACED_EXTENSIONLESS_FILENAMES: &[&str] = &[
    "Dockerfile",
    "Makefile",
    "LICENSE",
    "COPYING",
    "NOTICE",
    "README",
    "CHANGELOG",
    "CONTRIBUTING",
    "AUTHORS",
    "MAINTAINERS",
    "CODEOWNERS",
    "Gemfile",
    "Rakefile",
    "Procfile",
];

fn looks_like_surfaced_reference(token: &str) -> bool {
    if token.len() < 2 {
        return false;
    }
    if token.starts_with("http://") || token.starts_with("https://") {
        return false;
    }
    if token.contains('/') {
        return true;
    }
    if SURFACED_REFERENCE_EXTENSIONS
        .iter()
        .any(|ext| token.ends_with(ext))
    {
        return true;
    }
    SURFACED_EXTENSIONLESS_FILENAMES.contains(&token)
}

fn strip_sentence_period(text: &str) -> &str {
    let mut chars = text.char_indices().rev();
    let Some((last_idx, last_ch)) = chars.next() else {
        return text;
    };
    if last_ch != '.' {
        return text;
    }
    match chars.next() {
        Some((_, prev))
            if prev.is_alphanumeric() || matches!(prev, '`' | '"' | '\'' | ')' | ']' | '}') =>
        {
            &text[..last_idx]
        }
        _ => text,
    }
}

fn normalize_prose_token(token: &str) -> &str {
    let mut current = token;
    loop {
        let stripped = current
            .trim_matches(|c: char| matches!(c, '`' | '"' | '\'' | ',' | ';' | ':' | '!' | '?'));
        let stripped = strip_sentence_period(stripped);
        if stripped == current {
            return current;
        }
        current = stripped;
    }
}

pub(crate) fn collect_prose_path_candidates(text: &str, out: &mut Vec<String>) {
    for raw in text.split(|c: char| {
        c.is_whitespace() || matches!(c, ',' | ';' | '(' | ')' | '[' | ']' | '{' | '}')
    }) {
        let trimmed = normalize_prose_token(raw);
        if looks_like_surfaced_reference(trimmed) {
            out.push(trimmed.to_owned());
        }
    }
}

pub(crate) fn extract_key_files_candidates(text: &str) -> Vec<String> {
    let trimmed = text.trim();
    let mut candidates: Vec<String> = Vec::new();

    if let Some(rest) = trimmed.strip_prefix('`') {
        if let Some(end) = rest.find('`') {
            let path = &rest[..end];
            if !path.is_empty() {
                candidates.push(path.to_owned());
                return candidates;
            }
        }
    }

    let stripped = trimmed.trim_matches(|c: char| matches!(c, '`' | '"' | '\''));
    if !stripped.is_empty() {
        candidates.push(stripped.to_owned());
    }

    if let Some((prefix, _)) = stripped.split_once(" - ") {
        let prefix = prefix.trim();
        if !prefix.is_empty() && prefix != stripped {
            candidates.push(prefix.to_owned());
        }
    }

    candidates
}