> We could also have placeholders for
> command-line arguments, e.g. "(SOMETHING %i %o)" means invoke
> SOMETHING with a single command-line argument where %i is substituted
> by the name of a file with the code to be highlighted ("input") and %o
> with the name of the file to write the results ("output").Makes sense to me. It is way better than adding a new var. Should we use %i and %o or something more verbose like %INPUT and %OUTPUT? There is another point that is good to mention: If *HIGHLIGHT_SYNTAX* is set to anything other than "highlight" or "pygments" it defaults to *source-highlight*. So, if the user set "./myscript.pl" as *HIGHLIGHT_SYNTAX* without the placeholders the tex2any script will perform the highlight using the *source-highlight*. Which may confuse users. Since this is an experimental feature, we could introduce a breaking change and always expect for a command to be provided in order to simplify the implementation and remove this ambiguity. But IDK what is the project polices over breaking changes. I'll be glad to send a v2 with these changes once we clarify these points.
