def _strip_special_divs()

in obelics/processors/dom_tree_simplificator.py [0:0]

• 19 lines of code
• 12 McCabe index (conditional complexity)

    def _strip_special_divs(self, selectolax_tree):
        special_div_ids = ["footer", "header", "navigation", "nav", "navbar", "menu"]
        modification = True
        while modification:
            # Traverse the tree to find one node to remove, and remove it right then
            # to avoid the recursivity problem with `decompose`
            found_a_node = False
            for node in selectolax_tree.root.traverse():
                if node.tag == "div":
                    attributes = node.attributes
                    if (
                        ("id" in attributes and attributes["id"] in special_div_ids)
                        or ("class" in attributes and attributes["class"] in special_div_ids)
                        or ("title" in attributes and attributes["title"] in special_div_ids)
                    ):
                        node.decompose()
                        found_a_node = True
                        break
            if not found_a_node:
                modification = False
        return selectolax_tree