pdfminer.six: pdf text bbox don't match its real location.

Hi, I am using pdfminer.six==20181108,

I found this recently that: Sometimes a parsed PDF will have plenty of "\t"s in its texts. when that happens, the LRTextBoxes are not matching the real text location on the pdf page (char bbox are too large for that char itself, and covering adjacent char locations.)

Is there a way to solve this by setting up some params in LAParams? I guess? or has anyone else met this issue before.

this is my simple function for pdf text extraction.

from pdfminer.pdfdocument import PDFDocument
from pdfminer.pdfpage import PDFPage
from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
from pdfminer.layout import LAParams
from pdfminer.converter import PDFPageAggregator

    def load_pdf(pdf_stream):
        doc = PDFDocument(PDFParser(pdf_stream))
        rsrcmgr = PDFResourceManager()
        device = PDFPageAggregator(rsrcmgr, laparams=LAParams(detect_vertical=False))
        interpreter = PDFPageInterpreter(rsrcmgr, device)
        pages = PDFPage.create_pages(doc)

        pgs = []
        for _, pg in enumerate(pages):
                interpreter.process_page(pg)
                pgs.append(device.get_result())
        return pgs```