What causes "'{raw_id}' no found" in gitbutlerapp/gitbutler?

Re-run the listing command to get fresh hunk ids, then retry the selection

'{raw_id}' no found

Error message

'{raw_id}' no found

What it means

filter_uncommitted_hunks re-parses every raw hunk id from the IdMap against the current diff context (parse_using_context). An empty result means the id matches no CLI hunk anymore - the worktree changed between listing and selection. The message's 'no found' means 'not found'.

Source

Thrown at crates/but/src/utils/diff_rendering.rs:1161

fn filter_uncommitted_hunks<'a, F>(
    ctx: &'a Context,
    id_map: &'a IdMap,
    mut filter: F,
) -> anyhow::Result<Vec<(&'a str, Arc<CliId>, &'a UncommittedHunk)>>
where
    F: FnMut(&but_core::SingleHunk) -> bool,
{
    let mut uncommitted_hunks = id_map
        .uncommitted_hunks
        .iter()
        .filter(move |(_, hunk)| filter(&hunk.hunk))
        .map(|(raw_id, hunk)| {
            let mut cli_ids = id_map.parse_using_context(raw_id, ctx)?;
            if cli_ids.len() == 1 {
                Ok((&**raw_id, Arc::new(cli_ids.remove(0)), hunk))
            } else if cli_ids.is_empty() {
                bail!("'{raw_id}' no found")
            } else {
                bail!(
                    "'{raw_id}' resolved to more than one hunk ({})",
                    cli_ids.len()
                )
            }
        })
        .collect::<anyhow::Result<Vec<_>>>()?;

    uncommitted_hunks.sort_by(|(id_a, _, hunk_a), (id_b, _, hunk_b)| {
        (
            &hunk_a.hunk.path,
            hunk_a
                .hunk
                .hunk_header
                .as_ref()
                .map(|header| header.old_start),
            id_a,

View on GitHub (pinned to caf1f223d3)

Solutions

Re-run the listing command to get fresh hunk ids, then retry the selection
Ensure no file changes happen between listing and selecting hunks
If another process touched the worktree, let it finish, then refresh the ids

Defensive patterns

Strategy: retry

Try / catch

match filter_uncommitted_hunks(ctx, &id_map, filter) {
    Err(e) if e.to_string().contains("no found") => {
        // worktree changed: recompute IdMap, re-list hunk ids, reselect once
    }
    r => r,
}

Prevention

Re-resolve ids in the same operation that listed them
Treat any worktree write as invalidating cached hunk ids
When this fires, show the user a fresh listing instead of the raw error

When it happens

Trigger: Listing hunk ids, then modifying/staging/reverting files before selecting one; a concurrent process (editor, formatter, another but invocation) alters the worktree; ids come from a stale listing.

Common situations: Interactive hunk-selection flows (squash/discard) where the user edits files mid-session, long-lived shells holding old ids, CI retrying an operation after files changed.

Related errors

AI-assisted analysis of gitbutlerapp/gitbutler@caf1f223d3 (2026-08-20). Data as JSON: /api/errors/3e3dd2ac564380ea. Report an issue: GitHub.