fix(search): allow LOAD * in FT.AGGREGATE #2726

[watersRand]

Description

This pull request adds support for the LOAD * syntax within the FT.AGGREGATE command.

Purpose:
According to the [RediSearch documentation](https://redis.io/commands/ft.aggregate/), the LOAD clause typically expects a specific number of arguments (nargs). However, it also supports a wildcard * to project all document attributes into the aggregation pipeline.

Problem:
The previous implementation utilized pushVariadicWithLength, which automatically prepended a numeric count to the arguments. When passing *, the library generated LOAD 1 *, resulting in a Redis protocol error.

Solution:
The parseAggregateOptions function now includes type narrowing to handle the wildcard case:

• If LOAD is exactly '*', the parser pushes the literal string * as the argument count substitute.
• For all other cases (individual strings, objects, or arrays of fields), the original pushVariadicWithLength logic is maintained to ensure backward compatibility.

Checklist

• Does npm test pass with this change (including linting)?
• Is the new or changed code fully tested?
• Is a documentation update included (if this change modifies existing APIs, or introduces new ones)?

---

Note

Medium Risk
Changes FT.AGGREGATE argument construction to special-case LOAD, which can affect query serialization and compatibility with RediSearch. While covered by a new unit test, this touches core command parsing logic used broadly by clients.

Overview
Adds LOAD * support for FT.AGGREGATE. FtAggregateOptions.LOAD now accepts the wildcard '*', and parseAggregateOptions emits LOAD * directly instead of LOAD 1 * (while keeping the existing counted-args behavior for specific fields/aliases and arrays).

Updates tests to cover the new wildcard projection case.

^{Reviewed by Cursor Bugbot for commit 095ae1c. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{Reviewed by Cursor Bugbot for commit 095ae1c. Configure here.}

[cursor]

Missing newline causes } if on same line

Low Severity

The closing brace of the LOAD block and the TIMEOUT if statement are on the same line (} if (), making it visually ambiguous whether this is an independent if or was intended to be an else if. The original code had these as clearly separated independent if blocks with a blank line between them. A future maintainer could easily mistake this for a missing else and introduce a regression.

 

^{Reviewed by Cursor Bugbot for commit 095ae1c. Configure here.}

[nkaradzhov]

@watersRand Thanks, this looks good!