fix: handle scalar XPath results in Selector.xpath()

[mercael91]

Summary

lxml's xpath() returns a scalar (float/bool) for XPath functions like count() and boolean(). The existing code tried to iterate over the scalar, causing TypeError: 'float' object is not iterable.

Change

Add an isinstance() check before the iteration loop to return scalar results as a single-item string list.

# Added before the iteration loop
if isinstance(results, (float, int, bool)):
    return SelectorList([], _strings=[str(results)])

Repro

from pyscrappy import Selector

Selector("<div><p>1</p><p>2</p></div>").xpath("count(//p)").getall()
# Before: TypeError: 'float' object is not iterable
# After: ['2.0']

Fixes #136

[vedaant00]

Verified locally: the fix is right, checking the scalar on the whole result and routing it through _strings is exactly the clean way to handle count()/boolean(). CI lint was red on trailing whitespace in the new blank lines, so I pushed a commit that fixes that and adds two regression tests (count and boolean) since #136 asked for coverage; both fail on the pre-fix code. 483 passed. Nice fix @mercael91.