thoughtsync/desktop/src-tauri/src/local/derive.rs

//! Deriving `[[wiki-links]]` and `#tags` from a note's body — the local mirror of
//! what the server computes on save. Pure string scanning (no regex dependency),
//! kept in lockstep with the frontend's inline rules (see frontend notes/markdown.ts):
//!
//! - `[[link]]`: `[[` … `]]` with no brackets inside, inner text trimmed. Used to
//!   compute backlinks at query time (links are derived, never stored/synced).
//! - `#tag`: `#` at a word boundary followed by tag characters (letter first).
//!   On save these become labels attached with `via_tag = true`.
//!
//! Both dedupe case-insensitively, preserving first-seen order.

/// Extract the trimmed inner text of every `[[wiki-link]]` in `body`.
pub fn extract_links(body: &str) -> Vec<String> {
    let bytes = body.as_bytes();
    let mut out: Vec<String> = Vec::new();
    let mut i = 0;
    while i + 1 < bytes.len() {
        if bytes[i] == b'[' && bytes[i + 1] == b'[' {
            if let Some(rel) = body[i + 2..].find("]]") {
                let inner = &body[i + 2..i + 2 + rel];
                // Mirror the frontend's `[^[\]]+`: no stray brackets inside.
                if !inner.contains('[') && !inner.contains(']') {
                    let t = inner.trim();
                    if !t.is_empty() {
                        push_unique(&mut out, t);
                    }
                }
                i += 2 + rel + 2;
                continue;
            }
        }
        i += 1;
    }
    out
}

/// Extract every `#tag` name (without the leading `#`) from `body`.
pub fn extract_tags(body: &str) -> Vec<String> {
    let chars: Vec<char> = body.chars().collect();
    let mut out: Vec<String> = Vec::new();
    let mut i = 0;
    while i < chars.len() {
        if chars[i] == '#' {
            let boundary = i == 0 || (!is_tag_char(chars[i - 1]) && chars[i - 1] != '#');
            // A tag must start with a letter (so "#1" or a bare "#" is not a tag).
            if boundary && i + 1 < chars.len() && chars[i + 1].is_alphabetic() {
                let mut j = i + 1;
                while j < chars.len() && is_tag_char(chars[j]) {
                    j += 1;
                }
                let tag: String = chars[i + 1..j].iter().collect();
                push_unique(&mut out, &tag);
                i = j;
                continue;
            }
        }
        i += 1;
    }
    out
}

fn is_tag_char(c: char) -> bool {
    c.is_alphanumeric() || c == '_' || c == '-'
}

fn push_unique(out: &mut Vec<String>, candidate: &str) {
    if !out.iter().any(|x| x.eq_ignore_ascii_case(candidate)) {
        out.push(candidate.to_string());
    }
}

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

    #[test]
    fn links_basic_and_trim() {
        assert_eq!(extract_links("see [[ Alpha ]] and [[Beta]]"), vec!["Alpha", "Beta"]);
    }

    #[test]
    fn links_dedupe_case_insensitive_first_seen() {
        assert_eq!(extract_links("[[Note]] then [[note]] again"), vec!["Note"]);
    }

    #[test]
    fn links_ignore_malformed_and_nested_brackets() {
        assert_eq!(extract_links("[[a[b]] [[]] [ [x] ] plain"), Vec::<String>::new());
        assert_eq!(extract_links("[[ok]] [[a]b]]"), vec!["ok"]);
    }

    #[test]
    fn tags_basic() {
        assert_eq!(extract_tags("a #todo and #Work-item_2 here"), vec!["todo", "Work-item_2"]);
    }

    #[test]
    fn tags_require_letter_start_and_boundary() {
        // "#1" (digit) and an in-word "#" (email-ish) are not tags.
        assert_eq!(extract_tags("#1 nope a#b no but #Yes"), vec!["Yes"]);
    }

    #[test]
    fn tags_dedupe_case_insensitive() {
        assert_eq!(extract_tags("#Home #home #HOME"), vec!["Home"]);
    }

    #[test]
    fn empty_body() {
        assert!(extract_links("").is_empty());
        assert!(extract_tags("").is_empty());
    }
}