fix: stop rg after file limit

[ry2009]

What does this PR do?

• Stop ripgrep scans once the file limit is reached and kill the rg process.

  • Disable symlink follow + hidden scanning by default for glob/list (only
    enable when targeted).
  • Pass limit through the debug rg path.

  Fixes Glob search uses 100% of CPU #5220

How did you verify your code works?

• Manual: Bun script calling Ripgrep.files({ limit: 100 }) on /tmp/oc-rg-test
  (20k files + symlink loop); returns 100 quickly, no lingering rg processes.

[github-actions]

The following comment was made by an LLM, it may be inaccurate:

Based on my search, I found one potentially related PR:

Related PR:

• glob: stop ripgrep when results sufficient and ignore heavy dirs #5266 - "glob: stop ripgrep when results sufficient and ignore heavy dirs"
  • glob: stop ripgrep when results sufficient and ignore heavy dirs #5266
  • This appears to address a similar concern about stopping ripgrep execution early when sufficient results are found, which is related to your PR's goal of stopping rg after file limit is reached.

No duplicate PRs found that are currently open and addressing the exact same issue as PR #9220.

[ry2009]

Also cleaner then my previous perf fix #5266

[konovalov-nk]

This is cursed as well...
See my comment in other thread