OneToNElementsMap got into inconsistent state

Error message

OneToNElementsMap got into inconsistent state

What it means

An internal invariant check in OneToNElementsMap: parent_to_children listed a child index that has no entry in child_to_parent. The two maps must always agree, so this panic indicates a corrupted internal state — a bug in the map's insert/remove logic, not bad user input.

Solutions

Reproduce with the offending config and file a bug report with the config snippet — this is an internal invariant violation in eww's scope bookkeeping.
Work around by fully restarting the daemon (`eww kill && eww daemon`) instead of hot-reloading the dynamic config.
Update eww to the latest version in case the insert/remove inconsistency was already fixed.
Simplify the config: avoid widgets/for_each constructs that churn scope identities on every reload to reduce exposure.

Defensive patterns

Strategy: validation

Validate before calling

// sanity-check the map before querying
assert!(parent_to_children.keys().all(|c| child_to_parent.contains_key(c)), "map out of sync");

Try / catch

// this is an unrecoverable internal invariant; catch_unwind around config reload keeps the daemon alive
let r = std::panic::catch_unwind(AssertUnwindSafe(|| graph.get_children_edges_of(idx)));

Prevention

Keep parent_to_children and child_to_parent mutations paired in a single method
Write property tests asserting the two maps mirror each other after insert/remove
Avoid rapid config reloads on broken configs until the bug is fixed
Report the triggering config upstream

When it happens

Trigger: Calling get_children_edges_of(index) after a sequence of operations that removed a child's parent entry (or inserted the child into parent_to_children without a matching child_to_parent entry), e.g. during scope-graph rebuilds on config reload.

Common situations: Hit while reloading eww configs with dynamic scopes (for_each/windows), typically during rapid config reloads or hot-reload races that partially tear down scope mappings.

Understand the failure class

Background: "This is a bug, please report it": internal invariant violations, unreachable panics, and SNH errors explained — this error's family across 47 libraries.

Related errors

AI-assisted analysis of elkowar/eww@48f5aa8b37 (2026-09-08). Data as JSON: /api/errors/09e804a2deb069d4. Report an issue: GitHub.

Appendix: source

Thrown at crates/eww/src/state/one_to_n_elements_map.rs:65

    pub fn get_parent_edge_of(&self, index: I) -> Option<&(I, T)> {
        self.child_to_parent.get(&index)
    }

    pub fn get_parent_edge_mut(&mut self, index: I) -> Option<&mut (I, T)> {
        self.child_to_parent.get_mut(&index)
    }

    #[allow(unused)]
    pub fn get_children_of(&self, index: I) -> HashSet<I> {
        self.parent_to_children.get(&index).cloned().unwrap_or_default()
    }

    /// Return the children and edges to those children of a given scope
    pub fn get_children_edges_of(&self, index: I) -> Vec<(I, &T)> {
        let mut result = Vec::new();
        if let Some(children) = self.parent_to_children.get(&index) {
            for child_scope in children {
                let (_, edge) = self.child_to_parent.get(child_scope).expect("OneToNElementsMap got into inconsistent state");
                result.push((*child_scope, edge));
            }
        }
        result
    }

    #[cfg_attr(not(debug_assertions), allow(dead_code))]
    pub fn validate(&self) -> Result<()> {
        for (parent, children) in &self.parent_to_children {
            for child in children {
                if let Some((parent_2, _)) = self.child_to_parent.get(child) {
                    if parent_2 != parent {
                        bail!(
                            "parent_to_child stored mapping from {:?} to {:?}, but child_to_parent contained mapping to {:?} \
                             instead",
                            parent,
                            child,
                            parent_2

View on GitHub (pinned to 48f5aa8b37)