yewstack/yew · error · syn::Error

unsupported literal

Error message

unsupported literal

What it means

`HtmlNode::parse` classifies each child literal; a token that lexes as a literal but cannot be parsed into a known kind lands in `Lit::Verbatim` and is rejected as 'unsupported literal' (html_node.rs:28-32). Typical inputs are numeric literals with unrecognized suffixes (e.g. `1u0`) — tokens rustc would also reject, but the macro reports them first with a span pinned to the literal itself.

Source

Thrown at packages/yew-macro/src/html_tree/html_node.rs:30

pub enum HtmlNode {
    Literal(Box<Lit>),
    Expression(Box<Expr>),
}

impl Parse for HtmlNode {
    fn parse(input: ParseStream) -> Result<Self> {
        let node = if HtmlNode::peek(input.cursor()).is_some() {
            let lit = input.parse()?;
            match lit {
                Lit::ByteStr(lit) => {
                    return Err(syn::Error::new(
                        lit.span(),
                        "byte-strings can't be converted to HTML text
                         note: remove the `b` prefix or convert this to a `String`",
                    ));
                }
                Lit::Verbatim(lit) => {
                    return Err(syn::Error::new(lit.span(), "unsupported literal"));
                }
                _ => (),
            }
            HtmlNode::Literal(Box::new(lit))
        } else {
            HtmlNode::Expression(Box::new(input.parse()?))
        };

        Ok(node)
    }
}

impl PeekValue<()> for HtmlNode {
    fn peek(cursor: Cursor) -> Option<()> {
        cursor.literal().map(|_| ()).or_else(|| {
            let (ident, _) = cursor.ident()?;
            match ident.to_string().as_str() {
                "true" | "false" => Some(()),

View on GitHub (pinned to 0e4a05472f)

Solutions

1. Fix the suffix to a valid one (`u8`, `i32`, `f64`, …) or remove it entirely
2. If the child is meant to be displayed text, quote it: `{ "1" }`
3. If the tokens come from your own macro, emit the number and suffix as separate, valid tokens instead of one malformed literal

Example fix

// before
html! { <p>{ 1u0 }</p> }

// after
html! { <p>{ 1u8 }</p> }

When it happens

Trigger: A child like `html! { 1u0 }` — an integer or float literal whose suffix is not a valid primitive suffix (`u8`, `i32`, `f64`, …), so syn cannot classify it.

Common situations: Typos in literal suffixes (`1u6`, `2.0f64x`), macro-generated code gluing a suffix onto a number, or pasting experimental literal syntax from elsewhere into a template.

Related errors

• byte-strings can't be converted to HTML text note: remove th
• string literals must not contain more than one class (hint:
• expected a valid closing tag for component note: found openi
• missing condition for `if` expression
• this `if` expression has a condition, but no block

AI-assisted analysis of yewstack/yew@0e4a05472f (2026-08-22). Data as JSON: /api/errors/ddfcebae0bdd8da1. Report an issue: GitHub.