IMGKit: Python library of HTML to IMG wrapper | Hacker News Reader

IMGKit: Python library of HTML to IMG wrapper

github.com

65 pointsJack061424 comments

Open articleView on HN

What would be a good reason to use this over PhantomJS?

If you wanted a different engine: Webkit.

Edit: I was mistaken, they both back to Webkit.

PhantomJS is also WebKit...

Python

PhantomJS is a JavaScript API. This is a Python API.

As much as I despise the nodejs dependency hell, PhantomJS is a better solution when compared to wkhtml, so a Python-PhantomJS wrapper would be a significantly better product imho.

PhantomJS has stand-alone binaries that you can just call from Python if you wanted.

That's all this project does.

At least for screenshots, the headless support recently introduced into chrome might be the most straightforward solution, with the fewest pieces.

Some detail: https://docs.google.com/document/d/1VTcYz4q_x0f1O5IVrvRX4u1D...

Can this create/upsample an image to a specified resolution/zoom level? Say I want a 2000x5000 pixel image of the way a website looks when the window size is 400x1000.

I looked through the manual but didn't see explicit examples.

If it's jsut upsampling you could do that with PIL/Pillow, after saving the output image.

Not if you want photos of how responsive layouts behave at higher resolutions.

Which is why the parents specifically are just talking about upsampling the image, though

I want it rendered at the higher resolution, not upsampled from a lower-resolution image. Should have made that more clear.

You can see manual of wkhtmltoimage:)

How well does this handle SVGs? I used PhantomJS for headless rendering before, but it didn't properly render SVG linear gradients.

Does the text remain text or does it become an image?

It creates an image. Think of it like taking a screenshot of the website.

For the sake of preserving the structure, I'd prefer WeasyPrint.

Yes, you can make text become image

I think the title of this is a bit misleading. It is a wrapper lib around the cli of wkhtmltopdf (which the GitHub repo states).

Oh god. wkhtmltopdf sucks on so many levels, Webkit itself being one of them as their fork is totally outdated... sometimes I really think about rewriting it using Chromium Embedded Framework.

I found there is no python API of this feature, so I wrote it, this lib is based on ruby pdfkit

You're better off using PhantomJS to take screenshots.

There's still no great solution for Flash if you have to support that. I did a lot of research on this previously.

Reply on news.ycombinator.com