TeX, LaTeX and Friends

Skillmon

08:53

@PauloCereda we need another package, dogsettings.

@barbarabeeton we had our first snow on Friday, but it wasn't nearly enough for a good snowball fight either, and currently it's snowing here, but it melts as soon as it hits the ground.

Paulo Cereda

09:10

@Skillmon ooh

David Carlisle

09:47

faq.gutenberg-asso.fr

2

Joseph Wright

@DavidCarlisle Nice

David Carlisle

@JosephWright C'est bien que je parle couramment le français

Joseph Wright

@DavidCarlisle Oui

egreg

Someone started chapter 8 of the TeXbook.

David Carlisle

10:05

@egreg They should just start at appendix D like any normal person

2

Jasper Habicht

Oh boy, what is with the close votes list today? Does it ever end?

Pablo González L

11:43

@JosephWright Since you are in variant cleaning mode, I think some of them can be cleaned in l3doc, for example: \cs_generate_variant:Nn \tl_set_rescan:Nnn { NnV }.

Joseph Wright

@PabloGonzálezL I'll look at that

mickep

12:58

And I who thought ^^ was some smiley...

egreg

@mickep That's ^_^ but, alas, it's illegal in TeX.

mickep

@egreg (^^)

Skillmon

@mickep (i?

mickep

@Skillmon O_o

Skillmon

@mickep °,..,°

mickep

13:11

@Skillmon Ohhh

Pablo González L

@JosephWright Thanks

Hi, short question `expl3`...Can I do this using `clist_map_inline:nn` or some other way?
```
\cs_set_protected:Npn \__my_tmp:nn #1 #2
{
\keys_define:nn { mymodule / #1 }
{
key-a .tl_set:c = { l_my_key_#2_tl }
}
}
\__my_tmp:nn { one } { i }
\__my_tmp:nn { two } { ii }
\__my_tmp:nn { three } { iii }
\__my_tmp:nn { four } { iv }
```

Jasper Habicht

13:32

@PabloGonzálezL Is the question how to translate digits to numerals (like 1 to "one")?

Pablo González L

13:47

No, actually I want to do something like this:
```
\cs_set_protected:Npn \my_tmp:n #1
{
\tl_new:c { l_my_counter_#1_tl }
}
{ \clist_map_inline:nn { i, ii, iii, iv } { \__my_tmp:n {#1} }
```
but using two arguments `\__my_tmp:nn`.

Joseph Wright

@PabloGonzálezL If you want multiple arguments, you either need to use the pairwise seq functions or (more likely) do your own recursion using \q_recursion_tail/stop

Pablo González L

@JosephWright Any example at hand in siunitx to guide me...these are for a list environment and I have to repeat those lines several times :(

Joseph Wright

\ExplSyntaxOn
\cs_new_protected:Npn \__pg_iterator:nn #1#2
  {
    \quark_if_recursion_tail_stop:n {#1}
    \keys_define:nn { pg / #1 } { #2 .code:n = { } }
    \__pg_iterator:nn
  }
 \__pg_iterator:nn
   { one } { i }
   { two } { ii }
   { three } { iii }
   { four } { iv }
   { \q_recursion_tail } { ? }
   \q_recursion_stop

@PabloGonzálezL ^^^

Pablo González L

@JosephWright Great, thanks a lot...I think I can adapt that to what I need :D

Alan Munn

14:20

@campa Channeling @DavidCarlisle I see. :)

David Carlisle

@AlanMunn @samcarter will be impressed with the tikzness shown in both the answers there.

samcarter

@DavidCarlisle Of course I am -- it is always impressive to see a true TikZ gold badge holder answer a tikz question:)

Skillmon

14:43

@PabloGonzálezL alternatively:

\cs_set_protected:Npn \__my_tmp:nn #1 #2
  { \keys_define:nn { mymodule / #1 } { key-a .tl_set:c = { l_my_key_#2_tl } } }
\clist_map_inline:nn { {one} {i}, {two} {ii}, {three} {iii}, {four} {iv} }
  { \__my_tmp:nn #1 }

A loop, similar to @JosephWright's but faster:

\cs_set_protected:Npn \__my_tmp:nnw #1 #2 #3#
  {
    \keys_define:nn { mymodule / #1 } { key-a .tl_set:c = { l_my_key_#2_tl } }
    #3 \__my_tmp:nnw
  }
\__my_tmp:nnw
  {one} {i}
  {two} {ii}
  {three} {iii}
  {four} {iv}
  \use_none:nn {}

campa

@AlanMunn Erm, I must confess that I don't quite know what you mean with "channeling" in this context...

David Carlisle

@campa he means that I have infiltrated your psyche and causing you to answer tikz questions with picture mode

3

samcarter

15:13

@campa Be careful, you are only one step away from eating roasted duck for dinner :)

Joseph Wright

@Skillmon Oh no, the Bruno trick :(

Skillmon

@JosephWright I wasn't aware it had a name :P But why the long face?

campa

@samcarter Uhm, I was actually thinking about chicken this evening... coincidence or slow infiltration?

2

Skillmon

@campa he's turning you around!

samcarter

@campa oh no, the bird-eating has spread :)

campa

15:27

4

Joseph Wright

15:38

@Skillmon Tends to be a bit fragile if you need to change the nature of the iteration

Paulo Cereda

15:52

@campa oh no

David Carlisle

17:15

@JosephWright is alignat supposed to work (for math grabbing)

\DocumentMetadata{
testphase={
    phase-III
    ,math,
  },
pdfversion=2.0,
uncompress}

\documentclass{article}

\usepackage{amsmath}


\begin{document}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

\section{Equation groups with mutual alignment}


xxx
\begin{alignat}{2}
x& = y_1-y_2+y_3-y_5+y_8-\dots
                  &\quad&x\\
 & = y'\circ y^*  && xx\\
 & = y(0) y'      && xx\\
\end{alignat}
%

\end{document}

Joseph Wright

@DavidCarlisle ER, yes ...

David Carlisle

Runaway argument?
\let \split \insplit@ \DN@ {\align@ \st@rredfalse }\fi \fi \collect@body \ETC.
! File ended while scanning use of \__math_grab_dollardollar:w.
<inserted text>
                \par

Ulrike Fischer

17:27

@DavidCarlisle the registration seems to be missing. If you add after begin document the following it works (but gives one line too much but also without the tagging code)

\ExplSyntaxOn
 \math_register_halign_env:nn {alignat}{}
\ExplSyntaxOff

Joseph Wright

@UlrikeFischer Could well be: this was tricky ...

David Carlisle

@UlrikeFischer ah thanks

Ulrike Fischer

@DavidCarlisle is it expected that one gets a fourth line?

David Carlisle

@UlrikeFischer yes the last \\ are wrong I'd added them earlier to avoid make4ht dropping lines (which is no longer required)

@UlrikeFischer so I think I can tag amsldoc with mathml AF

Ulrike Fischer

@DavidCarlisle wow ;-)

Joseph Wright

17:35

@DavidCarlisle Wow!

David Carlisle

@UlrikeFischer I just need to back out some tests and experiments and see if it still works....

Joseph Wright

@UlrikeFischer Will you fix or should I?

Ulrike Fischer

@JosephWright I will add it to my branch where I'm moving to sockets etc.

Joseph Wright

@UlrikeFischer Cool

mickep

@DavidCarlisle Nice!

David Carlisle

17:50

1066 and all that

! Package tagpdf Error: The number of automatic begin (1065) and end (1066)
(tagpdf)                text-unit para hooks differ!

But only one error in 46 pages not too bad, will look again later.

Ulrike Fischer

18:06

@DavidCarlisle that is from amsldoc?

Ulrike Fischer

18:38

@DavidCarlisle the class naturally breaks all the tagging of the sectioning commands, but the para error is from the {\samepage verbatim } on line 1477. That hides the \@doendpe setting. If they would use the proper environment \begin{samepage}...\end{samepage} it would work.

Pablo González L

19:06

@Skillmon Thank you very much :D ... your alternative is quite understandable :D

David Carlisle

@UlrikeFischer ooh just finished eating ~~@PauloCereda~~ chicken, let me look...

Joseph Wright

19:18

Loads of documentation tune-ups; I wonder why

Ulrike Fischer

@JosephWright some people seem to read documentation, I wonder why ;-)

Pablo González L

Documentation ....mmm...I read in @JosephWright blog that the correct thing to do is to use \NewDocumentCommand \foo (without braces), but in xparse it appears \NewDocumentCommand {\foo } :D

David Carlisle

19:38

@PabloGonzálezL well both work:-)

Skillmon

@JosephWright they're not as negatively influenced by @DavidCarlisle as @campa is.

campa

@Skillmon My free will is fighting back: In the end it's been a caprese salad for dinner...

Udi Fogiel

Hello all, is \file_if_exists:n(TF) expand the first argument (the name of the file) similar to \input?

Joseph Wright

@UdiFogiel Yes

Udi Fogiel

@JosephWright Great thanks!

Skillmon

19:48

@JosephWright I'd have said no, \file_if_exists:nTF isn't defined... :P

Udi Fogiel

@Skillmon yap, I meant \file_if_exist:n(TF) :)

Skillmon

@UdiFogiel yes, I know :)

David Carlisle

@UdiFogiel more \InputIfFileExists (\input would give an error on missing file)

David Carlisle

20:32

@UlrikeFischer I checked in amsldoc.pdf even though it doesn't work `! Missing $ inserted.
<inserted text>
$
l.1282 \end
{minipage}
` if I scroll past that I get a pdf which pdf-xchange is happy with but ngpdf doesn't like at all (I used a modified class that removed all tehe changes to lists and sections, so i just get book class defaults) Giving up for tonight (hence teh checkin)

Ulrike Fischer

@DavidCarlisle yes, I got an error there too. minipage inside other stuff is still a problem (and math tagging inside the tabular is not activated anyway).

David Carlisle

@UlrikeFischer I may just try to simplify the layout a bit there so it gets past that.

Ulrike Fischer

20:55

@DavidCarlisle actually it is not the minipage but the display math inside it: \[\int_{A_1}\int_{A_2}\dotsi.\]

David Carlisle

@UlrikeFischer yes, I tried replacing the nested minipage by p columsn but same

Ulrike Fischer

@DavidCarlisle it is the problem that math grabbing is deactivated and [ is mapped to equation* which tries to get its body.

\DocumentMetadata{testphase={math}}
\documentclass{article}
\begin{document}
\ExplSyntaxOn
\bool_set_true:N \l__math_collected_bool
\ExplSyntaxOff
\[\int_{A_1}\int_{A_2}\dotsi.\]
\end{document}

Once we activate math grabbing inside it should work (or if we define \[ better.)

David Carlisle

@UlrikeFischer ah thanks, I'll have a play and see if I can get a version working now.

Joseph Wright

@UlrikeFischer WE should look again at this I guess

@UlrikeFischer Think I said we really do have to grab everything: I suspect that would be better as a straight kernel change without needing tagging even loaded

Transcript for

$TeX - LaTeX$ TeX, LaTeX and Friends