Bruno Le Floch

What do we do with malformed utf8? It's not l3regex's job to detect it or fix it, but it would be good to have some clear rule about what to...

It seems the recommended approach is to raise an error as early as possible when encountering malformed utf8, so I'll try to do that. I think the main change will...

Correct. Quoting the doc: % The option |(?i)| makes the match case insensitive (identifying % \texttt{A}--\texttt{Z} with \texttt{a}--\texttt{z}; no Unicode support % yet). The code that maps A-Z to a-z...

> What I think will be a bit problematic are character ranges, because the code > seems to use (currently) the ASCII codepoint for a given char, so |A-Z| is...

Just a historical note: we used to have `\peek_catcode_ignore_spaces:N`, `\peek_catcode_remove_ignore_spaces:N` (now listed in `doc/l3obsolete.txt`), so the sentence probably dates from back then.

It occurs to me that `\iow_term:x` or more generally `\iow_now:Nx` expand their argument through a normal `x`-expansion and thus their argument is not expanded within the `\write` statement itself. Thus,...

It occurs to me that we might be able to make `\mode_if_write:TF` do something useful in messages etc at the expense of more code. Then it would be realistic to...

Perhaps call it `\mode_if_shipout:TF` (in relation with `\lua_shipout_x:n` and `\iow_shipout_x:Nn`).

The internal representation of floating point numbers in `l3fp` does not include any notion of trailing zeros. So this padding can only occur at display time.

It seems #1360 (which suggests to add checking to `\prop_if_empty:NTF` and friends) is relevant to this follow-up PR you mention. I second Joseph's point that the debugging code needs not...