fix(zip): skip directory entries in ZipConverter

[nileshpatil6]

What

When a ZIP archive contains explicit directory entries (e.g. created with zipfile.ZipFile.mkdir() or tools like Info-ZIP), zipObj.namelist() returns those entries alongside actual files.

Before this fix, ZipConverter passed every name to zipObj.read(), which returns an empty bytes object for directory entries. That empty stream then went through convert_stream, and the result produced a spurious ## File: subdir/ heading with no content.

Fix

Add an is_dir() guard before reading and converting each entry:

if zipObj.getinfo(name).is_dir():
    continue

Reproduction

import zipfile, io
from markitdown import MarkItDown

buf = io.BytesIO()
with zipfile.ZipFile(buf, 'w') as z:
    z.mkdir('subdir')
    z.writestr('subdir/note.txt', 'hello')
buf.seek(0)

md = MarkItDown()
result = md.convert_stream(buf, stream_info_kwargs={'extension': '.zip'})
print(result.markdown)
# Before: includes empty '## File: subdir/' section
# After:  only '## File: subdir/note.txt' with content