test_pdf.py 文件源码

def pdf_as_matrix(buff, border):
    """\
    Reads the path in the PDF and returns it as list of 0, 1 lists.

    :param io.BytesIO buff: Buffer to read the matrix from.
    """
    pdf = buff.getvalue()
    h, w = re.search(br'/MediaBox \[0 0 ([0-9]+) ([0-9]+)\]', pdf,
                     flags=re.MULTILINE).groups()
    if h != w:
        raise ValueError('Expected equal height/width, got height="{}" width="{}"'.format(h, w))
    size = int(w) - 2 * border

    graphic = _find_graphic(buff)
    res = [[0] * size for i in range(size)]
    for x1, y1, x2, y2 in re.findall(r'\s*(\-?\d+)\s+(\-?\d+)\s+m\s+'
                                        r'(\-?\d+)\s+(\-?\d+)\s+l', graphic):
        x1, y1, x2, y2 = [int(i) for i in (x1, y1, x2, y2)]
        y = abs(y1)
        res[y][x1:x2] = [1] * (x2 - x1)
    return res