Add --max-count to stop after the first N matches

[Gulliby]

Requested by gulliby

Adds --max-count N (-m N) to stop after the first N matching records, like grep -m.

Before: logslice printed every match and read the input to its end, even when only the first few records were wanted from a very large log.

After: the tool stops reading as soon as N records have matched (after all filters, across files in argument order). Later files are not opened. --count prints at most N, and --summary summarizes only those N records.

$ logslice errors.log --max-count 2
2026-10-10 12:00:00 ERROR Disk full
2026-10-10 12:00:05 ERROR Timeout

An invalid value (0, 01, -1, abc, 1000000000, empty) is an argument error with exit status 2.

How: parse_max_count validator in src/logslice/filters.py, the option and the early stop in src/logslice/cli.py, README paragraph and options-table row, one new row in the help test, and 34 new cases in tests/test_max_count.py (291 tests in all).