The word order in Chinese is sometimes misarranged when parsing PDF

### Bug

The word order in Chinese is sometimes misarranged when parsing PDF

[“人工智能+传媒”：推动媒...人工智能+传媒”研讨会综述_曹雅芳.pdf](https://github.com/user-attachments/files/19523672/%2B.%2B._.pdf)

For example

the title in PDF is {“人工智能+传媒” ： 推动媒体深融的重要引擎}
but the parse result is {## '人工智能+传媒' 动媒体深融的重要引擎 ： 推}
the correct result should be {## '人工智能+传媒'： 推动媒体深融的重要引擎}

### Steps to reproduce

converter = DocumentConverter()
url = str(req.url)
result = converter.convert(source=url)

### Docling version

Docling version: 2.28.4
Docling Core version: 2.24.1
Docling IBM Models version: 3.4.1
Docling Parse version: 4.0.0
Python: cpython-312 (3.12.9)
Platform: macOS-15.3.2-arm64-arm-64bit

### Python version

Python 3.12.9