items.sh: return early when no Taps directory

[cho-m]

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same change?
• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your changes? Here's an example.
• Have you successfully run brew style with your changes locally?
• Have you successfully run brew typecheck with your changes locally?
• Have you successfully run brew tests with your changes locally?

---

Workaround for removing errors output in #14254. Will need API usage for actual completion support, but haven't had chance to look into this.

EDIT: Updated to allow completion from cached JSON to fix #14254

Haven't tested on HOMEBREW_INSTALL_FROM_API setup, but seemed to work when I changed Taps in following line to invalid directory:

brew/Library/Homebrew/items.sh

Line 21 in 884c4be

find "${HOMEBREW_REPOSITORY}/Library/Taps" \

---

e.g. changing above to Taps2 and trying without this PR:

❯ brew formulae
find: /opt/homebrew/Library/Taps2: No such file or directory

❯ brew info find: /opt/homebrew/Library/Taps2: No such file or directory
find: /opt/homebrew/Library/Taps2: No such file or directory
brew

with PR:

❯ brew formulae


❯ brew info <TAB>
<NOTHING>

[BrewTestBot]

Review period will end on 2022-12-19 at 21:44:38 UTC.

[BrewTestBot]

Review period ended.

[cho-m]

Concept idea for reading cached JSON using ruby without brew libraries.

This currently doesn't handle any error inside Ruby. Particularly, a corrupt JSON file ends up dumping the entire contents of formula.json to terminal which can be quite lengthy. EDIT: hiding stderr should be enough for this.

I had added grep . since ${formulae} can be "" so it leads to blank line. Not sure if there is better way to handle this.

[MikeMcQuaid]

This makes sense! Happy to try it out.

[Rylan12]

I wonder if we should do this calculation when we download HOMEBREW_CACHE/api/formula.json, instead. That way we only need to do it once per unique cache file (I'm assuming this line is relatively slow)

[MikeMcQuaid]

@Rylan12 That seems like a good idea! Let's merge as-is for now and iterate to improve performance here 💪🏻

[MikeMcQuaid]

Thanks again @cho-m!