[fix] Track watch iterations independently to ensure proper time tracking (#1565)

[scottnemes]

Description

This fixes a bug where if the watch query is ran as part of a multi-query statement it will return the incorrect run time for the first iteration. I.e.

Before:

After:

Resolves #1565

Checklist

• I added this contribution to the changelog.md file.
• I added my name to the AUTHORS file (or it's already there).
• To lint and format the code, I ran

  uv run ruff check && uv run ruff format && uv run mypy --install-types .

[github-actions]

Findings (Correctness)

1. Watch sleep adjustment is still applied to every watch result after the first, even within the same watch iteration.
   In mycli/main.py:915, watch_count flips to >0 after the first watch result and never changes. That means for a watch query that produces multiple SQL results per iteration (e.g., watch 5 select 1; select 2;), the second result in the same iteration will still subtract the sleep time, under-reporting execution time.
   Suggested fix: only apply the sleep offset on the first result of each watch iteration. Options:
   • Tag the first SQL in each watch iteration in watch_query (e.g., command["watch_iteration"]=iteration_id or command["watch_is_first"]=True) and key off that in output_res.
   • Or track per-iteration boundaries in output_res by incrementing a watch_iteration_id from the generator rather than a simple boolean.

Tests / Edge Cases

1. Add a regression test for multi-statement watch: watch 5 select 1; select 2;
   Assertions: only the first result of each watch iteration should subtract the sleep duration.
2. Add a regression test for mixed multi-query input: select 1; watch 5 select 2;
   Assertion: the first watch iteration should not offset time because of the preceding non-watch statement.

No security concerns found in this change.

[rolandwalker]

Sweet!