fix RESP3 mismatches

[mgravell]

A few commands give different result shapes between RESP2 and RESP3, and require custom processing; this PR fixes those, making them work with either shape, interchangeably; usually this means arrays becoming maps, which by itself would be fine, but: the actual nesting/indenting/etc is changed at the same time - for example, a single flat array with a "stride" might become a jagged nesting of map[key, array]. This can vary between server versions - for example, search on v7 using RESP3 uses a different layout to v8 using RESP3; the implementation here attempts to handle things based on inspection of what arrives, rather than hard coding versions to layouts.

Affected areas:

• streams
• time-series
• search
• json
• spell-checking

To ensure future compliance, we change [RunPerProtocol] to be implicit (RESP2+RESP3) on all integration tests (tests can still use explicit [RunPerProtocol] to change the combination, but this would be rare; the default of "test everything on RESP2 and RESP3" is a good one).

---

Note

Medium Risk
Updates core response-parsing logic to accept multiple RESP3 map/array layouts across Redis module commands; mistakes could cause subtle data-shape regressions at runtime. Changes are mostly additive/branching, but touch widely-used parsing paths for search, timeseries, streams, and JSON.

Overview
Fixes multiple RESP3 vs RESP2 response-shape mismatches by expanding ResponseParser to detect RESP3 Map/wrapped-array layouts and normalize them into the existing object model (notably TimeSeries MGET/MRANGE, rules/labels, streams, FT spellcheck/config dictionaries, and profile search/aggregate parsing across v7/v8 layouts).

Updates JSON commands to use shared parsers for JSON.TYPE and JSON.NUMINCRBY so RESP3 arrays and RESP2 JSON-string payloads both work.

Test infrastructure now defaults integration tests to run under both RESP2 and RESP3, infers Redis server version from the first shared connection, and re-validates skip logic at runtime via AssertVersion; several tests were adjusted accordingly (including unique JSON keys and TimeSeries test namespaces/constructors).

^{Reviewed by Cursor Bugbot for commit b426103. Bugbot is set up for automated code reviews on this repo. Configure here.}

[jit-ci]

🛡️ Jit Security Scan Results

✅ No security findings were detected in this PR

---

^{Security scan by Jit}

[cursor]

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

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 6e65015. Configure here.}