Add --batch option as an alternative to STDIN

[rolandwalker]

Description

Add --batch option as an alternative to STDIN, factoring out a statements_from_filehandle() function which can also be used to improve the functionality of the source command.

The initial behavior when both STDIN is not-a-tty and --batch is in effect is to warn, and use --batch. Similarly, the priority of the --execute option was clarified with warnings. These may be revisited: we could consider different behavior, as well as exiting with an error in case of warning.

Motivation: display of a prompt_toolkit progress bar during batch execution. If the input is from STDIN, we can't know how many statements are the goal, at least without creating a tempfile. As mentioned above, this also lays the basis to improve the source command, which reads the entire file into memory.

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]

No correctness or security regressions stood out in the PR diff.

Residual testing gaps (worth adding before merge):

1. Missing coverage for --batch - (explicit STDIN path), including precedence behavior when STDIN is non-tty.

   • Relevant logic: main.py:2565

2. Missing coverage for the new MAX_MULTILINE_BATCH_STATEMENT guard raising ValueError in the factored parser.

   • Relevant logic: batch_utils.py:14

3. No direct unit tests for statements_from_filehandle() splitting behavior across multi-statement input and trailing non-terminated statement.

   • Relevant logic: batch_utils.py:8

I couldn’t run tests locally because pytest/uv aren’t installed in this environment.

[github-actions]

1. Regression: --batch does not expand ~ paths
   File: mycli/main.py:2570
   click.open_file(batch_file) uses the literal string, so --batch '~/script.sql' fails (especially when quoted), while source already does os.path.expanduser(...) in mycli/main.py:483.
   Action: expand user paths before open, e.g. click.open_file(os.path.expanduser(batch_file)).

2. Missing edge-case coverage for new batch flow
   Files: test/test_main.py:1313, mycli/packages/batch_utils.py:8
   Current tests cover happy path and option precedence with --execute, but not key new edges:

• --batch - should read stdin explicitly.
• both piped stdin and --batch file should verify warning + file wins behavior.
• MAX_MULTILINE_BATCH_STATEMENT overflow path (ValueError handling) should exit cleanly.
• statements_from_filehandle() should be unit-tested directly for trailing-no-semicolon and multi-statement boundaries.

No security-specific issues stood out in this diff.

Assumption: I could not run pytest here (pytest not installed in this environment), so findings are from static review only.