vmui: auto-suggestion stops working after the server returns big number of results

[valyala]

Describe the bug

Auto-suggestion stops working after the server returns big number of results, even the number of returned results is smaller than the limit=1000 hardcoded at vmui

To Reproduce

1. Open https://play.victoriametrics.com/select/accounting/1/6a716b0f-38bc-4856-90ce-448fd713e3fe/prometheus/graph/
2. Make sure the Autocomplete toggle is disabled.
3. Open network inspector in the browser and clear the history of network queries there.
4. Type p in the query input field at vmui and press ctrl+space
5. Note that vmui sends the request to https://play.victoriametrics.com/select/accounting/1/6a716b0f-38bc-4856-90ce-448fd713e3fe/prometheus/api/v1/label/__name__/values?match%5B%5D=%7B__name__%3D%7E%22.*p.*%22%7D&limit=1000&start=1706652000&end=1706738399.999 and obtains a list of metric names with p char. The list contains 777 items, e.g. its' size is smaller than the limit=1000.
6. Note that vmui doesn't show a popup with auto-suggestion list of metric names with p char. This isn't OK, since it is expected that it must show the first 50 items from the list with the warning like 50 out of 777 items are displayed - type more chars for more specific suggestions.
7. Type rocess_cpu in the query input field, so now it contains process_cpu, and press ctrl+enter again.
8. vmui doesn't send any requests to the server. This is expected behaviour, since it already has an exhaustive list of items with the p char, which has been loaded at the step 5.
9. vmui doesn't show a short list with metric names containing process_cpu in their names. This is unexpected behavior.

Version

The latest commit in master branch.

Logs

No response

Screenshots

No response

Used command-line flags

No response

Additional information

Related issue - #5472

[valyala]

This issue has been fixed in v1.98.0