mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-04-25 00:51:20 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

test(ui-tui): drop dead terminalReally from drift repro

Browse source 
Copilot flagged the variable as unused. LogUpdate.render only sees prev/next, so a simulated "physical terminal" has no hook in the public API. Kept the narrative in the comment and tightened the assertion to demonstrate the test's actual invariant: identical prev/next emits no heal patches.
This commit is contained in:
Brooklyn Nicholson 2026-04-23 13:03:06 -05:00
parent 7c4dd7d660
commit f28f07e98e
1 changed files with 15 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

43
ui-tui/packages/hermes-ink/src/ink/log-update.test.ts
View file

@ -93,44 +93,31 @@ describe('LogUpdate.render diff contract', () => {
expect(stdoutOnly(diff)).toContain('shorterrownow')
})
it('drift repro: if terminal has content that prev.screen does not know about, diff leaves it orphaned', () => {
// Simulates prev/terminal desync: the physical terminal has STALE
// content at row 2 from a prior frame that was never reconciled into
// prev.screen. next.screen is blank at row 2. Diff finds prev==next
// (both blank at row 2), emits nothing → the stale content survives
// on the terminal as an artifact.
it('drift repro: identical prev/next emits no heal, even when the physical terminal is stale', () => {
// Load-bearing theory for the rapid-resize scattered-letter bug: if the
// physical terminal has stale cells that prev.screen doesn't know about
// (e.g. resize-induced reflow wrote past ink's tracked range), the
// renderer has no signal to heal them. LogUpdate.render only sees
// prev/next — no view of the physical terminal — so when prev==next,
// it emits nothing and any orphaned glyphs survive.
//
// This is the load-bearing theory for the rapid-resize scattered-letter
// bug: whenever the ink renderer believes prev.screen is authoritative
// but the physical terminal was mutated out-of-band (resize-induced
// reflow writing past the prev-frame's tracked cells), those cells
// drift and artifacts appear at that row on subsequent frames.
// The fix path is upstream of this diff: either (a) defensively
// full-repaint on xterm.js frames where prevFrameContaminated is set,
// or (b) close the drift window so prev.screen cannot diverge.
const w = 20
const h = 3
const prevAsInk = mkScreen(w, h)
paint(prevAsInk, 0, 'same')
// row 2 in prevAsInk is blank — but pretend the terminal has stale
// characters there. ink has no way to know.
const terminalReally = mkScreen(w, h)
paint(terminalReally, 0, 'same')
paint(terminalReally, 2, 'orphaned')
const prev = mkScreen(w, h)
paint(prev, 0, 'same')
const next = mkScreen(w, h)
paint(next, 0, 'same')
next.damage = { x: 0, y: 0, width: w, height: h }
const log = new LogUpdate({ isTTY: true, stylePool })
const diff = log.render(mkFrame(prevAsInk, w, h), mkFrame(next, w, h), true, false)
const diff = log.render(mkFrame(prev, w, h), mkFrame(next, w, h), true, false)
const written = stdoutOnly(diff)
expect(written).not.toContain('orphaned')
expect(stdoutOnly(diff)).toBe('')
expect(diff.some(p => p.type === 'clearTerminal')).toBe(false)
// Verdict: in this configuration the renderer cannot heal the drift.
// The only recovery path from ink's side is fullResetSequence — which
// triggers only on viewport resize or scrollback-change detection,
// neither of which fires on a pure drift. A fix has to either (a)
// defensively emit a full repaint on every xterm.js frame where
// prevFrameContaminated is set, or (b) close the drift window at the
// renderer level so the in-memory prev.screen cannot diverge.
})
})