Function clean_graphics

src/table.py:2298–2343

Detect and join rectangles of "connected" vector graphics.

(npaths=None)

        return False

    def clean_graphics(npaths=None):
        """Detect and join rectangles of "connected" vector graphics."""
        if npaths is None:
            allpaths = page.get_drawings()
        else:  # accept passed-in vector graphics
            allpaths = npaths[:]  # paths relevant for table detection
        paths = []
        for p in allpaths:
            # If only looking at lines, we ignore fill-only paths,
            # except simulated lines (i.e. small width or height).
            if (
                lines_strict
                and p["type"] == "f"
                and p["rect"].width > snap_x
                and p["rect"].height > snap_y
            ):
                continue
            paths.append(p)

        # start with all vector graphics rectangles
        prects = sorted(set([p["rect"] for p in paths]), key=lambda r: (r.y1, r.x0))
        new_rects = []  # the final list of joined rectangles
        # ----------------------------------------------------------------
        # Strategy: Join rectangles that "almost touch" each other.
        # Extend first rectangle with any other that is a "neighbor".
        # Then move it to the final list and continue with the rest.
        # ----------------------------------------------------------------
        while prects:  # the algorithm will empty this list
            prect0 = prects[0]  # copy of first rectangle (performance reasons!)
            repeat = True
            while repeat:  # this loop extends first rect in list
                repeat = False  # set to true again if some other rect touches
                for i in range(len(prects) - 1, 0, -1):  # run backwards
                    if are_neighbors(prect0, prects[i]):  # close enough to rect 0?
                        prect0 |= prects[i].tl  # extend rect 0
                        prect0 |= prects[i].br  # extend rect 0
                        del prects[i]  # delete this rect
                        repeat = True  # keep checking the rest

            # move rect 0 over to result list if there is some text in it
            if chars_in_rect(CHARS, prect0):
                # contains text, so accept it as a table bbox candidate
                new_rects.append(prect0)
            del prects[0]  # remove from rect list

        return new_rects, paths

    bboxes, paths = clean_graphics(npaths=paths)