Negative effect of running .png() on .jpg files?

I am in the process of optimizing lots of images, a mixture of both .png and .jpg images. Can I just run all the files through the same function?

sharp(files[c])
  .png({...})
  .jpeg({...})
  .toBuffer(...)

So if the file is a PNG does it ignore the .jpeg() method and if it’s a JPG does it ignore the .png() method? Or do I need to determine the mime type ahead of time and use the appropriate method?

Is there any downside to doing this? Any negative affect on the files? Does it take a considerably longer time than if I checked the mime type ahead of time and used only the appropriate method for each mime type?

I am writing something that will be processing 1000’s of images at a time so I just want to get an idea of processing time one way or the other. The documentation just isn’t clear on this subject and I don’t see any examples or tests where both .png() and .jpeg() is used in the same chain.

[lovell]

@polarathene Thanks for reporting, have created #1528 to track this.

[polarathene]

@lovell great, so I shouldn’t need to try reproduce it any further? Is it a small fix to create? If you’re pressed for time I could try give it a go perhaps?

It should roughly involve knowing if a force: true for an image format in the chain had been declared earlier? Assuming you have something like an array of the format option objects that should be fairly easy to identify? 😃