Function get_text_words

src/utils.py:86–162

Return the text words as a list with the bbox for each word. Args: page: pymupdf.Page clip: (rect-like) area on page to consider flags: (int) control the amount of data parsed into the textpage. textpage: (pymupdf.TextPage) either passed-in or None. sort:

(
    page: pymupdf.Page,
    clip: rect_like = None,
    flags: OptInt = None,
    textpage: pymupdf.TextPage = None,
    sort: bool = False,
    delimiters=None,
    tolerance=3,
)

def get_text_words(
    page: pymupdf.Page,
    clip: rect_like = None,
    flags: OptInt = None,
    textpage: pymupdf.TextPage = None,
    sort: bool = False,
    delimiters=None,
    tolerance=3,
) -> list:
    """Return the text words as a list with the bbox for each word.

    Args:
        page: pymupdf.Page
        clip: (rect-like) area on page to consider
        flags: (int) control the amount of data parsed into the textpage.
        textpage: (pymupdf.TextPage) either passed-in or None.
        sort: (bool) sort the words in reading sequence.
        delimiters: (str,list) characters to use as word delimiters.
        tolerance: (float) consider words to be part of the same line if
            top or bottom coordinate are not larger than this. Relevant
            only if sort=True.

    Returns:
        Word tuples (x0, y0, x1, y1, "word", bno, lno, wno).
    """

    def sort_words(words):
        """Sort words line-wise, forgiving small deviations."""
        words.sort(key=lambda w: (w[3], w[0]))
        nwords = []  # final word list
        line = [words[0]]  # collects words roughly in same line
        lrect = pymupdf.Rect(words[0][:4])  # start the line rectangle
        for w in words[1:]:
            wrect = pymupdf.Rect(w[:4])
            if (
                abs(wrect.y0 - lrect.y0) <= tolerance
                or abs(wrect.y1 - lrect.y1) <= tolerance
            ):
                line.append(w)
                lrect |= wrect
            else:
                line.sort(key=lambda w: w[0])  # sort words in line l-t-r
                nwords.extend(line)  # append to final words list
                line = [w]  # start next line
                lrect = wrect  # start next line rect

        line.sort(key=lambda w: w[0])  # sort words in line l-t-r
        nwords.extend(line)  # append to final words list

        return nwords

    pymupdf.CheckParent(page)
    if flags is None:
        flags = pymupdf.TEXTFLAGS_WORDS
    tp = textpage
    if tp is None:
        tp = page.get_textpage(clip=clip, flags=flags)
    elif getattr(tp, "parent") != page: