TL;DR

Using [convert][] to assemble multiple images in a PDF file.

In recent posts (Crop an image with ImageMagick, No voids, please) I introduced some small activity I’ve done recently, which is:

• take a bunch of images, all with the same structure
• isolate one part of the image with some cropping operation
• remove spaces from the resulting file names
• turn all the cropped, renamed files into a single PDF file.

Here’s the final script for doing the last step:

This also makes it clear why I wanted to get rid of the spaces: I’m just taking all the file names and putting them inside the command line.

This approach has the obvious shortcoming of being limited by the shell command line size. You know, the kind of stuff for which you would resort to [xargs][].

In this case, though, I’m anticipating that I’ll be joining only a few (say less than 50) files, so I doubt that the command line will become too lengthy.

With this, my whole pipeline became something like this:

ls inputs*png | crop 10,10 100,100 | no-voids | imgs2pdf stuff.pdf

I’m not entirely sure that the output on standard output should be the filename, though. On the one hand, it’s the actual output of this stage in the pipeline; on the other hand, in this case the most useful following step in the pipeline would be to get rid of the cropped inputs, which I’m losing here.

While I ponder about this… please stay safe!