I use pandoc to convert html to pdf with my personal latex documentclass, in which images not exists would be replaced with boxes.

For
```
<img src="not-exists.jpg">
```
if I convert it to latex first, and then to pdf, everything works as expected.

But if I convert it to pdf directly, pandoc reports a warning: could not fetch resource, replacing image with description.

I hope it does what the "html->latex->pdf" process did, without checking the resource. Is this customizable?

--