Skip to content

fix(search): check the last shortened read window against its budget #9987

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
waleedlatif1 merged 1 commit into from
Nov 13, 2026
Merged

fix(search): check the last shortened read window against its budget #9987

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 16 additions & 0 deletions apps/sim/lib/sim-search/live/read-window.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
/**
* @vitest-environment node
*/
import { describe, expect, it } from 'vitest'
import { encodedWindowEnd } from '@/lib/sim-search/live/read-window'

const encoded = (text: string) => JSON.stringify(text).length - 2

describe('encodedWindowEnd', () => {
it('returns the longest window it found to fit when escapes lead the text', () => {
const content = `${'"'.repeat(1_000)}${'x'.repeat(20_000)}`
const end = encodedWindowEnd(content, 0, 20_000, 8_000)
expect(end).toBe(19_000)
expect(encoded(content.slice(0, end))).toBeLessThanOrEqual(20_000)
})
})
6 changes: 5 additions & 1 deletion apps/sim/lib/sim-search/live/read-window.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,9 @@
/** Characters one standard read returns, the page budget indexed document reads use. */
export const READ_WINDOW_CHARACTERS = 8000

/** Times a window is shortened to fit its budget before the minimum is used instead. */
const MAX_SHRINKS = 3

/**
* Where a window of `content` from `start` ends so its text, as JSON encodes it, fits `budget`
* characters. JSON escapes quotes, backslashes and line breaks, so a window is shortened until its
Expand All @@ -13,9 +16,10 @@ export function encodedWindowEnd(
minimumEnd: number
): number {
let end = Math.min(content.length, start + budget)
for (let attempt = 0; attempt < 3 && end > minimumEnd; attempt++) {
for (let shrinks = 0; end > minimumEnd; shrinks++) {
const encoded = JSON.stringify(content.slice(start, end)).length - 2
if (encoded <= budget) return end
if (shrinks === MAX_SHRINKS) break
end = start + Math.floor(((end - start) * budget) / encoded)
}
return minimumEnd
Expand Down
Loading