loongson/pypi: links for web-extract