Quote:
Originally Posted by tn4w
For all the users.
|
The you in my comment was not you personally but a generic you as in a calibre user.
Umm... just how is calibre going to validate the cover image for all the users?
I won't ask you to write the code but perhaps you could generate a simple flow diagram of how you would program the image insertion code to be able to distinguish a "good" cover from a "bad" cover.
Don't forget to allow for tasks such as distinguishing image quality, image size, colour depth, etc. that we currently use wetware to handle. Make sure that you include the code paths needed to analyze the image file to locate strings of text to match to the author(s), editor(s), book title, series title and index, etc. to ensure that you have a cover image that matches the book. You may also want to familiarize yourself with the current state of the art in image recognition and the current software patents in this area.
In simple words, what you are asking for is not a simple task.
For your original query, with the current state of calibre development, how would you know that the URL you want to use to download the image has a valid, reasonably sized image if you haven't visited the page and examined the image for yourself?