Unicode replacement

[QuentinAndre11]

Hello !
I am facing a weird problem: when I extract the text of my PDF with extract_text or extract_text_simple, I have some of my characters replaced by �, which should be my desired behavior.
However, if I do not use extract_text but directly the layout detection (I need it because I need more complicated logic on the extracted textboxes), when I iterate on page.objects["textboxhorizontal"] the "text" property does not contain the � anymore, how comes ? What block of the code answers to this difference ?
I use the latest stable version from pypi.
I cannot give you the PDF but I hope the description of the problem is sufficient.
Have a nice day !

[QuentinAndre11]

I actually found out that these characters were stored in textlinehorizontal but not in textboxhorizontal, so nevermind