Documentation improvement about "stop downloading the image after a few kilobytes"

[rogerdielrton]

Hello.
In "usage" doc, when you say:
"You can optionally check the buffer lengths & stop downloading the image after
a few kilobytes. You don't need to download the entire image"
Could you be most explicit about how many (minimum) kb per file type is required to
detect image dimensions?
Could you update the documentation, with that info and an example?
Thanks in advance.

[rogerdielrton]

Something similar to:
https://github.com/arnaud-lb/imagesize.js
Get the size of an image without reading or downloading it entirely

[jspenc72]

Please do improve documentation for this.

[kurijov]

+1

[rauschma]

Two ideas come to mind:

• A constant that tells you how much data is enough for all image formats to determine the dimensions.
• A function that checks whether enough data has already been loaded.

[andressrg]

@rogerdielrton http-image-size does what you want to accomplish.

It even uses image-size.

[derwaldgeist]

+1
Also would like to see that, since performance improvement is important. My use-case is to detect the 5 largest images in a page, so I do not need the image content.

[derwaldgeist]

@rogerdielrton Well, this implementation just relies on the first chunk to contain all relevant data. The question still is: is this sufficient in every case?

[Alino]

+1

[puzrin]

https://github.com/nodeca/probe-image-size - it's specially designed to download only the necessary part of image.

[AlastairTaft]

+1