« first day (4074 days earlier)      last day (874 days later) » 

4:02 AM
Just recursively compare the tokens one by one.
Maybe I should post an answer there for completion... but is there already a package/answer that does that?
(not for any practical purposes)
 
 
4 hours later…
8:06 AM
@user202729 Handling spaces makes it impossible, at least we never found a way
 
@DavidCarlisle wait, this madness happens more than once in a lifetime?
@user202729 yes, etl
 
8:22 AM
@user202729 (and I wrote it, so I took the liberty to post an answer, sorry)
@JosephWright spaces are the hard part? Doubt that, or do you mean multiple consecutive spaces?
 
@Skillmon Yes
 
8:43 AM
Well it's ok, using package is more elegant than writing code manually
(actually package is just code of other people, but still)
@JosephWright As expected. (Never look into it.)
TeX parsing rules is complex enough without that.
I remember last time getting very stuck with tl_map_function not preserve spaces (tokcycle helps.)
"Should have been included in expl3. Packages are extremely hard to find"
 
9:07 AM
@user202729 That's by-design - something like the case-changer (where we do preserve spaces) is slower
 
@user202729 you can use \etl_act:nnnnnnnnnnnn (didn't count the ns :))
 
9:54 AM
@user202729 Need some justification - we have 'cool ideas' about that really don't have applications ....
 
 
2 hours later…
12:07 PM
@Skillmon Well to be fair C and similar languages doesn't have specifying parameters by name too
@JosephWright This thing is a "relatively basic" thing to do but it doesn't have built in
To be fair you can do something similar to tokcycle in LaTeX3 using the built in functions too while keeping the code relatively simple (I guess?)
 
12:31 PM
@Skillmon Why are the stars in etl documentation look italic (compared to those in interface3.pdf)? It looks so weird
 
 
1 hour later…
1:47 PM
@user202729 I’m not seeing this as ‘basic’ I guess because most of the time string comparisons work, and it’s rare in my experience to need tl comparison in an expansion context
 
 
1 hour later…
2:56 PM
@DavidCarlisle and @JosephWright -- Happy Boxing Day.
 
 
2 hours later…
5:20 PM
@user202729 If the strings are handled as macro-arguments, handling implicit/explicit spaces is not a big deal as only explicit space tokens cannot be grabbed as undelimited macro-arguments but leading explicit space-tokens of macro-arguments can be detected. The hard problem is finding an expandable way of, e.g., distinguishing explicit non-active character token from its active pendant when that is let equal to that non-active character token.
@user202729 Same problem with one-letter-control-sequences whose name equals the explicit character token which they are let equal to while \escapechar is negative. Handling of explicit catcode-1/2-character-tokens is cumbersome also but is feasible if you are willing to have a bumch of code only for doing brace-hacks.
 
5:37 PM
@user202729 Scenario 1: \begingroup\catcode`\Z=13\def\temp{\endgroup\letZ= }\temp Z -- how to expandably distinguish active-Z from letter-Z under these conditions without using macros that process delimited arguments? (\show Z in the first case delivers > the letter Z. and in the second case delivers > Z=the letter Z., so the phrases delived by \show slightly differ. But this is not the case for \meaning and \show is not expandable.
@user202729 Scenario 2: \let\!=!\escapechar=-1 -- how to expandably distinguish \! from ! under these conditions without using macros that process delimited arguments?
 
 
3 hours later…
9:00 PM
user image
5
Merry Christmas folks!
 
9:41 PM
@Werner merrychristmassamtsirhcyrrem
 
10:18 PM
\def\blame{@UlrikeFischer\blame}\blame
 
10:37 PM
@DavidCarlisle blame capacity exceeded, sorry [blame grouping levels=255]
2
 
@UlrikeFischer I could use \begingroup groups and blame you for that as well?
 
@user202729 different font
@user202729 well, but Hungarian notation comes close in verbosity.
 
@DavidCarlisle Sorry was a bit distracted -- my husband was asking about the skittles championship down under --. Lively discussion about (color/begin) groups ;-)
 

« first day (4074 days earlier)      last day (874 days later) »