zesdex/crates/zesdex-utils/src/slug.rs

#![allow(
    clippy::cast_possible_truncation,
    clippy::cast_sign_loss,
    clippy::cast_precision_loss,
    clippy::cast_possible_wrap
)]

use std::path::{Path, PathBuf};

const MAX_SLUG_LENGTH: usize = 80;

/// Convert an arbitrary string into a URL / filesystem-safe slug.
///
/// The algorithm:
/// 1. Lowercase the input.
/// 2. Replace any sequence of non-alphanumeric characters (except `-` and `_`)
///    with a single `-`.
/// 3. Strip leading/trailing `-`.
/// 4. If the result is empty, return `None`.
/// 5. Truncate to 80 characters, breaking at the last full word if possible.
///
/// Returns `None` if the slug would be completely empty.
pub fn slugify(s: &str) -> Option<String> {
    if s.is_empty() {
        return None;
    }

    let lower = s.to_lowercase();

    // Replace non-alphanumeric (except dash/underscore) sequences with '-'
    let mut slug = String::with_capacity(lower.len());
    let mut prev_was_sep = false;

    for c in lower.chars() {
        if c.is_alphanumeric() {
            slug.push(c);
            prev_was_sep = false;
        } else if !prev_was_sep {
            slug.push('-');
            prev_was_sep = true;
        }
        // else skip consecutive separators
    }

    // Strip leading/trailing dashes
    let slug = slug.trim_matches('-').to_string();

    if slug.is_empty() {
        return None;
    }

    // Truncate to MAX_SLUG_LENGTH
    let slug = if slug.len() > MAX_SLUG_LENGTH {
        let mut truncated: String = slug.chars().take(MAX_SLUG_LENGTH).collect();

        // Trim trailing dash from broken word boundary
        while truncated.ends_with('-') {
            truncated.pop();
        }

        if truncated.is_empty() {
            // If trimming removed everything, take the raw max-length prefix
            slug.chars().take(MAX_SLUG_LENGTH).collect()
        } else {
            truncated
        }
    } else {
        slug
    };

    Some(slug)
}

/// Join `base` with a slugified version of `name`.
///
/// If `slugify(name)` returns `None`, the name is used as-is (lowercased).
pub fn slug_path(base: &Path, name: &str) -> PathBuf {
    match slugify(name) {
        Some(slug) => base.join(slug),
        None => base.join(name.to_lowercase()),
    }
}

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

    #[test]
    fn test_slugify_basic() {
        assert_eq!(slugify("Hello World"), Some("hello-world".into()));
    }

    #[test]
    fn test_slugify_special_chars() {
        assert_eq!(slugify("Hello, World! #2"), Some("hello-world-2".into()));
    }

    #[test]
    fn test_slugify_empty() {
        assert_eq!(slugify(""), None);
    }

    #[test]
    fn test_slugify_only_separators() {
        assert_eq!(slugify("!!! @@"), None);
    }

    #[test]
    fn test_slugify_collapse() {
        assert_eq!(slugify("a   b---c___d"), Some("a-b-c-d".into()));
    }

    #[test]
    fn test_slugify_leading_trailing() {
        assert_eq!(slugify("---hello---"), Some("hello".into()));
    }

    #[test]
    fn test_slugify_dash_underscore_as_separator() {
        assert_eq!(slugify("my-slug_here"), Some("my-slug-here".into()));
    }

    #[test]
    fn test_slugify_truncate() {
        let long = "a".repeat(100);
        let slug = slugify(&long);
        assert!(slug.is_some());
        assert!(slug.as_ref().unwrap().len() <= MAX_SLUG_LENGTH);
    }

    #[test]
    fn test_slug_path() {
        let base = Path::new("/tmp");
        assert_eq!(slug_path(base, "Hello World"), Path::new("/tmp/hello-world"));
    }
}