Incremental parsing cannot parse contents if data is split at certain positions

[Theodore Chang]

Python : sys.version_info(major=3, minor=10, micro=11, releaselevel='final', serial=0)
lxml.etree : (5, 1, 0, 0)
libxml used : (2, 10, 3)
libxml compiled : (2, 10, 3)
libxslt used : (1, 1, 37)
libxslt compiled : (1, 1, 37)

See also: https://github.com/rajatomar788/pywebcopy/issues/123

If the fed data is split inside a `href`, nothing can be parsed.

See example:

Wrong:

```python
    from lxml import etree

    parser = etree.HTMLPullParser()
    for data in (b'<root><a href="2011-03-13_', b'135411/">2011-03-13_135411/</a></root>',):
        parser.feed(data)
        for _, elem in parser.read_events():
            print(elem.tag) # nothing
    parser.close()
```

Expected:

```python
    from lxml import etree

    parser = etree.HTMLPullParser()
    for data in (b'<root><a href="2011-03-13_135411/">2011-03-13_135411/</a></root>',):
        parser.feed(data)
        for _, elem in parser.read_events():
            print(elem.tag) # a root
    parser.close()
```

[scoder (scoder)]

Works for me with libxml2 2.12.6. The binary wheels of lxml 5.1 should be using 2.12.5, I guess that's ok as well.

[Theodore Chang (tlcfem)]

It looks like on Windows, the precompiled binary is downloaded during installation.

https://github.com/lxml/lxml/blob/82a42601e3a7f1a44aadc5a948c2b6fde3e0e407/buildlibxml.py#L34C1-L41C6

```py
# use pre-built libraries on Windows

def download_and_extract_windows_binaries(destdir):
    url = "https://api.github.com/repos/lxml/libxml2-win-binaries/releases?per_page=5"
```

Maybe consider bumping up version in libxml2-win-binaries repo?

[scoder (scoder)]

Ah, right, this seems to require libxml2 2.11+. I've added your test as a regression test for newer versions and updated the Windows libraries.

https://github.com/lxml/lxml/commit/807fd66704471434c8f1e2f9b6da66497ce43590
https://github.com/lxml/libxml2-win-binaries/commit/8e1f55e596ed14660064c7c19837f4171f5b0842