Bruno Le Floch comments

Results 146 comments of


                                            Bruno Le Floch

user-defined fp functions do not error if incorrect number of arguments given

I've implemented this idea, but only in a separate branch, pending discussions. This would possibly break user code if people rely on the possibility to silently ignore extra/missing arguments. Is...

Provide \exp_after:nN combining expanded, unexpanded and expandafter primitives

I have no strong opinion on name. Looking through the interface3 pdf index, `w` only appears in signatures as the last letter, except for `\exp_after:wN`. (There are also many internal...

Provide \exp_after:nN combining expanded, unexpanded and expandafter primitives

What signature?

Provide \exp_after:nN combining expanded, unexpanded and expandafter primitives

Sorry, but `\exp_after:nwN` makes no sense for what I am proposing. It could make sense for the following ``` \cs_new:Npn \exp_after:nwN #1 { \tex_expanded:D { \exp_not:n {#1} \exp_after:wN } \exp_after:wN...

Incorrect evaluation/display when applying l3fp user function to variable without value

User-defined functions call `\@@_exp_after_symbolic_f:nw` to correctly expand after them. This auxiliary is meant for use on expressions that have already been evaluated, so it suppresses evaluation, except (for some reason)...

Incorrect evaluation/display when applying l3fp user function to variable without value

I hesitate about the best behaviour. For now in the commit 97ed26c7bb400e0eb107baac41bc961037eee968 I've simply suppressed evaluation of user-defined functions when their argument is symbolic, so that `\fp_new_function:n { cube }...