trevorcampbell/shortex

LaTeX header file with useful definitions

shortex's Issues

Some more minor todos

• updated copyright (to include Jonathan and Jeff)
• no need for @ in \cadlag definition
• add var greek letters to \lowercaseGreek

\dee -> \d

from another style file I saw

Add a "squish math" command

We should add a command that reduces whitespace in a line of math (especially around symbols like +, -, =, which tend to be quite whitespace heavy.

Example -- not sure if this is possible, especially with the & inside, but ideally something like

\[
\squish{A &= b + c}
\]

backref hyperref

probably worth including an option for the shortex package that adds the backref option to the hyperref package. It's not something I've used a lot so I'd need to play with it a bit, but from the one demo I've seen it seems like it might be very useful to include.

Add an example document to the end of the documentation

As in the title. Serves as both an example and a unit test for style file changes

why no \hh?

Is there a reason to exclude h here?

shortex/shortex.sty

Line 639 in 1443cc0

\charalphabetmacro{h}{\hat}{\lowerCaseRomanLettersNoMHT}

I replaced \lowerCaseRomanLettersNoMHT with a similar \lowerCaseRomanLettersNoMT and did not have issues using \hh.

package option / default for creating the "usual" set of character style macros

My instinct is to have the default be "all the usual stuff" so that people aren't surprised when things don't work (I just had a user have this happen and ask me what was going on), and then have a lightweight option or something like that to avoid all the defs.

commenter names/colors

Right now we have \PRB, \RMK, \mPRB, \mRMK for problem/remark in-text/margin, and \HLT for highlighting text. I'm working on a paper right now with multiple people doing editing, and I find it super helpful to have comments denoted by initials.

I wonder if we can do something like

\usepackage[commenters=TC,JH,JN]{shortex}

and then have commenting commands (perhaps with commenter-specific colors!) like

\cTC   (comment TC)
\mcTC  (margin comment TC)
\pTC  (problem TC)
\mpTC  (margin problem TC)
\hTC   (highlight TC)

and furthermore, I wonder if the "problem" type comments are kind of pointless. When editing I don't feel the need to distinguish between "comment" and "problem" -- I just use comment for everything. Perhaps instead of "problem" we have bold comment like \bcTC if you want something to stick out more.

remove xargs?

Currently it seems to be unused. Need to investigate but could probably just remove it.

pp(, pp[, and cppl

I remember way back when I implemented these I wanted to do something like \pp{ or \pp\{ (instead of \cppl ) but couldn't get away with it. Just creating an issue to keep a record of that idea. Any ideas how to do that @jhuggins @jnegrea ?

new bracketing stuff

@jhuggins how do the new bracketing commands around here:

shortex/shortex.sty

Line 117 in c5ec2d0

\newcommand{\lcr}[3]{\left #1 #2 \right #3} % autosize

shortex/shortex.sty

Line 142 in c5ec2d0

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

compare to the \pp ones I added a while back:

shortex/shortex.sty

Line 134 in c5ec2d0

\def\pp({\left(}

?

Can you pull the new README from master into this branch and add some examples too?

Automatic environment macro generation, star versions

As described here: https://tex.stackexchange.com/questions/376375/using-ifstar-to-define-a-star-variant/376380

make autonum the default (option for noautonum)

as in the title. I think right now no autonum is default, but maybe I'm mistaken. Need to look into it.

replace eucal with mathrsfs

I never really paid much attention to this (since originally adapting the style file from other people), but shortex includes eucal. That package apparently replaces \mathscr with a script that sort of fits the same niche as \mathcal but just looks worse IMHO. I also don't see this script used very often in papers.

I would suggest we use mathrsfs instead, which is quite distinct from \mathcal, generally looks good, and is used somewhat frequently in papers.

(obviously all of my assessments are based on my own subjective opinion and particular literature bubble -- but at the very least, eucal is visually similar to mathcal, so I'd prefer having the more distinct mathrsfs)

Subscript special chars needs braces

I'm not sure if this is so important -- not even sure if I want to change this -- but figured I'd document it here. Currently,

f_\scS

will cause LaTeX to hit an error:

[21]
! Missing { inserted.
<to be read again> 
                   \let 
l.726 \]

In contrast:

f_{\scS}

works.

Use `\DeclarePairedDelimiter` rather than `\lcrx`?

As described in the math tools documentation, \DeclarePairedDelimiter does a lot the things we have done manually, but more cleanly if you just want to toggle between using \left\right or not adjusting the size. Right now \lcrx does not lead to good looking superscripts and subscripts, for example.

Overline, widetilde

Should implement

\ol____

and

\wt____

macros

Add CI tests?

After v0.1 releases, we should probably:

• create some unit tests for package functionality. I would need to investigate the right way to do that in LaTeX...
• create some GitHub actions to automate those tests on PR etc.

add levy-prokhorov distance

as in the title

too much space between theorem-like and display eqn

I'm not sure if this is something that is particular to the style file I'm using currently (neurips_2024.sty) or if it's more general, but I encountered this recently:

Reprex:

\[
blah
\]
\bthm
asdf
\ethm

results in too much space between the display and the theorem-like:

derivative commands

I like \der, but the fact that partial derivatives are \D is a bit odd (from a semantic naming point of view).

Maybe we should change to \pder? Other ideas welcome.

inconsistency in environment defs

some are \long\def, some \newcommand. Need to investigate a bit and decide on a consistent style for these (though may need inconsitency in some cases. My LaTeX-fu isn't good enough at the moment to know)

todos on new `\f...` macros

• testing
• add more annotation types (e.g. more fonts and accents)
• make it a package option
• consider changing to \s...

Jupyterbook documentation?

Right now we have a combination of a markdown readme and a separate PDF documentation.

Might be worthwhile to just do it all in Jupyterbook or something similar so we can write one documentation and have it as both the README and a compiled PDF.

I think this is a bit low priority at the moment, so I'll punt to some time after v0.1.

New set defines

@jhuggins I see you've added some new sets like

shortex/shortex.sty

Line 175 in c5ec2d0

\newcommand{\posReals}{\reals_+}

We should come up with a set of modifiers (like \b... and \h... for symbols) for common modifications to spaces. Goal is to:

• improve tex readability
• save time typing

e.g. camelcase is OK for readability but might make typing a pain (need to hit shift constantly). What do you think is the right tradeoff there?

Perhaps would it make sense to do something like \ex... for extended, so e.g. \exreals and \p... for positive, e.g. \preals ?

Some missing prefixed symbols

In the style file there are some commented out prefixed symbols (e.g. some of the \b... symbols). I imagine this is because of a clash with some base latex command.

We should figure out what to do about these -- ideally keep the super short prefixes we have now and get rid of the preceding definition, if it doesn't muck anything basic up...

typesetting matrices

Right now we typeset matrices via

\bmat
...
\ema 

which is a shortcut for \begin{pmatrix}...\end{pmatrix}. I want to use square brackets around my matrices -- I think it's more standard/common anyway (\begin{bmatrix}...\end{bmatrix}). Can we do

\bmat
\emat

for square brackets (or maybe \bbmat...\ebmat), and

\bpmat
\epmat

for parentheses?

Potential bugs in `\f` and `\parsefontstylestrings`

• it seems in the definition of \f I can't use \bm instead of \mathbf (which treats greeks better and maintains style on romans)
• the \parsefontstylestrings command doesn't work with greek symbols in the alphabet (at least, not the two obvious things I tried, \parse...{{blahblah}}{\alpha\beta\gamma} and \parse...{{blahblah}}{{alpha}{beta}{gamma}} )

Some of the default font style/accent combinations don't work

For example, \bdR doesn't seem to work currently. This is a must-fix before v0.1

Hessians

Currently shortex.pdf uses \hes but shorter.sty uses \hess.

Add mention of what packages *not to* include in docs

People have told me theyre getting option clashes, likely due to their inclusion of hyperref etc in their individual document. We should tell people what packages to remove from their own preamble in the documentation (and have an example in the "usage" section of the readme)

\P preceded by alignment reverts to pilcrow

Right now we have \P defined as

But if \P is preceded by an alignment character, like

Then it typesets as a pilcrow:

Adding almost anything, e.g. \hspace{0cm}, prior to \P fixes it:

`\beta` gets interpreted as `\bar{\eta}`

I think this macro needs fine tuning

shortex/shortex.sty

Line 612 in 167bb34

\texalphabetmacro{b}{\boldorbar}{\lowerCaseGreekLetters}

HLT should just change the text color

What the title says. Currently it also placed additional characters before the highlighted text.

arxiv green/red boxes are back...

I noticed recently that my blackhypersetup doesn't prevent those ugly green/red boxes from appearing in papers on arxiv. I think arxiv must have updated their process recently (as it also appears you can have a non-flat folder structure now too...)

A student of mine fixed it for one paper by doing

\hypersetup{
breaklinks=true,
colorlinks=true,
pdfusetitle=true
}

My guess is that the breaklinks=true command was the important one there, but some testing required.

draft mode

Add an option to pass through to graphicx for draft mode

Establish a contributing guide

We should establish some principles for defining new commands / shortcuts. E.g. a list of priorities when creating new commands:

• brevity
• consistency
• readability
• others?

Something I feel pretty strongly about: this style file is for "power-users," so I personally prefer brevity and consistency over plain-english readability (but usually brevity engenders readability so these aren't mutually exclusive goals). The other reason I put brevity at the top is for health reasons -- the less I type (especially special characters which require ctrl / shift / etc), the less likely I am to cause a repetitive stress injury to hands/wrists/etc.

E.g. if I could choose between \exreals vs \extendedReals vs \er I would have a strong preference for \exreals, which I think has the best tradeoff between readability and brevity of the three.

But if, say, \exnats was not possible to use (because of a previous basic unmutable latex define) while \extnats was, then I'd prefer \extreals for consistency's sake.

This is all open to discussion obviously, but I figured I'd try to give a sense for my style preferences.

Another (somewhat related) idea is to make sure every definition in the style file has a "long-form" that is readable, and then shortcut versions. Not sure whether this is a good idea or will just create a bunch of work to make things consistent.

Finally -- is it possible to define commands with macros? I.e. define the "bold prefix" operator, and then apply it to all characters. So if we ever want to change that prefix operator we just change one line in the style file instead of having to do it for all characters. This might also let us keep one "list of characters for which modifiers apply".

scurl not in the documentation

scurl is defined but missing from docs

v/var greeks in lowercaseGreeks?

I see we have a few shortened greek letters and variants (eps, ups, veps, vtheta, etc)

need to think about whether to remove these, include them in lowercaseGreeks, or leave them in but not include them

some additional packages

Someone reached out to me and suggested that we add:

• xr (and xr-hyper) for splitting main and appendix for conference submissions
• thmtools, especially for restateable theorems

I'm not sold on adding them necessarily, but figured I'd document it here for now pending further thought.

Add shorthands for abbreviations with lots of punctuation

Concretely:

\newcommand{\eg}{e.g.,\ }
\newcommand{\etal}{et al.\xspace}
\newcommand{\ie}{i.e.,\ }
\newcommand{\aka}{a.k.a.\ }

breqn?

I'm not sure how reliable this package is, or how well it plays with cleveref and autonum (I suspect not but who knows), but breqn seems like it might be another thing to add to shortex. It removes all the boilerplate around breaking equations.

Anyway, something to keep in mind

Filling out examples.tex

We need to provide examples for all the style file features in example.tex

reorganize the style file

Right now the file is organized in terms of "topic" (algorithms, urls, figures, etc). I think it would be better to match the documentation and organize in terms of types of shortcut/macro (so all environments together, for example). This will help with consistency (all similar defs together) and easier to change later to make automated macro defns out of a large block of manual defns.

algpseudocode comments break

Hi! Thanks for the great package. algpseudocode comments aren't working for me. The following breaks

\begin{algorithmic}
\State $i \gets 10$ \Comment{hi}
\end{algorithmic}

The log shows

Environment align* undefined. \State $i \gets 10$ \Comment{hi}