fix(search): prevent discovery restarts and stream results efficiently

[bajrangCoder]

No description provided.

[greptile-apps]

Confidence Score: 5/5

The PR appears safe to merge, with both previous review findings resolved and no actionable new defects identified.

Summary

• Bounds project-discovery waiting at five seconds and preserves the resulting snapshot.
• Adds worker acknowledgements and processing watchdogs for result backpressure and timeout handling.
• Streams native and fallback results in bounded per-file batches.
• Correctly associates repeated, interleaved result groups with their file metadata.
• Adds coverage for discovery timing, streaming, Unicode empty matches, native cancellation, and regex deadlines.

Diagram

sequenceDiagram
    participant UI as Search UI
    participant Discovery as File discovery
    participant Worker as Search worker
    participant View as Result view

    UI->>Discovery: waitForFileList()
    alt ready within five seconds
        Discovery-->>UI: ready
    else still scanning
        UI->>UI: take current snapshot
        UI->>UI: warn results may be incomplete
    end
    UI->>Worker: search snapshot
    loop Each bounded result batch
        Worker-->>UI: search-result
        UI->>View: append matches and file metadata
        UI-->>Worker: result-ack
    end
    Worker-->>UI: done-searching

Loading

_{Reviews (2) · Last reviewed commit: "fix"}