Fix IndexError in check_table_values_for_dict on empty tables (fixes #712)

[Leonard013]

Motivation

check_table_values_for_dict crashes with IndexError: list index out of range
whenever it is run on a table that has a column with zero rows. Empty DynamicTable /
EventsTable objects are legitimate NWB constructs, so nwbinspector currently crashes
on any file that contains one.

Root cause (diagnosed by @oruebel in #712), in src/nwbinspector/checks/_tables.py,
the per-column skip condition:

for column in table.columns:
    if not hasattr(column, "data") or isinstance(column, VectorIndex) or not isinstance(column.data[0], str):
        continue

column.data[0] is evaluated with no empty-length guard, so a zero-length column
raises IndexError.

What

Add a len(column.data) == 0 guard to the skip condition, placed before the
column.data[0] access so Python's or short-circuits on empty columns:

for column in table.columns:
    if (
        not hasattr(column, "data")
        or isinstance(column, VectorIndex)
        or len(column.data) == 0
        or not isinstance(column.data[0], str)
    ):
        continue

The multi-line form matches the sibling check check_col_not_nan directly below in the
same file. black --line-length 120 and ruff both pass.

Why this does not over-skip non-empty columns

or evaluates left-to-right and stops at the first truthy operand:

• Empty column (len(column.data) == 0 is True): short-circuits and continues,
  so column.data[0] is never evaluated — the crash is avoided.
• Non-empty column (len(column.data) == 0 is False): evaluation falls through to
  not isinstance(column.data[0], str), exactly as before. The new term is False for
  every non-empty column, so it changes nothing about which non-empty columns get
  scanned, and the dict-detection logic is untouched.

Only genuinely empty columns are skipped.

Tests

Added test_check_table_values_for_dict_empty_column to tests/unit_tests/test_tables.py:

def test_check_table_values_for_dict_empty_column():
    table = DynamicTable(name="test_table", description="")
    table.add_column(name="test_column", description="")
    assert check_table_values_for_dict(table=table) is None

• Before the fix: fails with IndexError at the diagnosed line.
• After the fix: passes.

The existing test_check_table_values_for_dict_*_fail tests continue to pass, confirming
a non-empty column containing a dict-string still triggers the check (no over-skip).
pytest tests/unit_tests/test_tables.py → 48 passed; broad slice → 340 passed, 7 skipped.

Follow-up (out of scope for #712)

Scoped to check_table_values_for_dict per @oruebel's diagnosis. Three sibling checks in
the same file carry the same latent empty-table IndexError and would still crash on a
full run over an empty table — worth a separate follow-up PR if desired:

• check_column_binary_capability — column.data[0]
• check_col_not_nan — column[0]
• check_table_time_columns_are_not_negative — table[column_name][0]

Happy to open that follow-up.

CHANGELOG

Added an entry under # v0.7.3 (Upcoming) → ### Fixes.