from __future__ import annotations
import itertools
from typing import TYPE_CHECKING
from .api import input_to_element
from .utils import extract_text, iterwalk_limited, take
if TYPE_CHECKING:
from collections.abc import Iterable
from lxml.html import HtmlElement
from . import SelectorOrElement
[docs]def extract_brand_name(node: SelectorOrElement, search_depth: int = 0) -> str | None:
"""Extract a brand name from a node that contains it.
It tries element text and image alt and title attributes.
:param node: Node including the brand name.
:param search_depth: Max depth for searching images.
:return: The brand name or None.
"""
_BRAND_LENGHT_LIMIT = 50
node = input_to_element(node)
extracted = _extract_brand(node, search_depth)
short = (b for b in extracted if b and len(b) < _BRAND_LENGHT_LIMIT)
results = take(short, 1)
return results[0] if results else None
def _extract_brand(node: HtmlElement, search_depth: int = 0) -> Iterable[str | None]:
if node.tag == "img":
return extract_image_text(node, 0)
value = extract_text(node)
if value:
return [value]
return extract_image_text(node, search_depth)
def extract_image_text(node: HtmlElement, search_depth: int = 0) -> Iterable[str]:
def extract_text_from_image(node: HtmlElement) -> Iterable[str | None]:
for attrib in ["alt", "title"]:
yield (node.attrib.get(attrib) or "").strip()
nodes = iterwalk_limited(node, search_depth)
images = filter(lambda n: n.tag == "img", nodes)
attribs = map(extract_text_from_image, images)
flat_attribs = itertools.chain.from_iterable(attribs)
return (a for a in flat_attribs if a)