feature request: support dynamic data resolver

[meritozh]

can we support dynamic data resolver so i can browser a comic archive without extracting first.

image api like:

type MangaPage = ImagePage | HtmlPage | ImagePageResolver;

export interface ImagePageResolver extends BasePage {
  type: "image";
  resolver: (string) -> string; // input: filename, output: base64 data?
  width?: number;
  height?: number;
  alt?: string;
}