Adding new features to the Typst template #9970

christopherkenny · 2024-07-10T22:32:51Z

This implements the changes suggested in #9956, with the exception of the filecolor/urlcolor one. These would require adding some regex to guess the link types. This is theoretically possible to do, but it wasn't clear to me that this is a good thing to put in a default template. Happy to adjust if you have thoughts on this.

Some things to note:

I'm converting colors by passing them as content, as I was seeing pandoc escape # if that was included. An alternative here is to just do something like: rgb(linkcolor.replace("\\#", "#")).
I set the default fonts for math and code ("raw") to fonts that are bundled with Typst. These need not be those fonts if there are more familiar pandoc preferences.

christopherkenny · 2024-08-22T00:28:35Z

Hi @jgm, I don't want to bother you if this is already on your radar, but should I add you as a reviewer (for when you have time)? I'm not 100% sure of the norms for contributions here.

jgm · 2024-08-23T18:21:54Z

Sorry for the delay on this!

jgm

Thanks for this PR, and sorry I took so long to review it. I made a number of queries and comments, which you can look over.

jgm · 2024-08-23T18:23:40Z

data/templates/default.typst

+$if(mathfont)$
+  mathfont: ($for(mathfont)$"$mathfont$",$endfor$),
+$endif$
+$if(codefont)$
+  codefont: ($for(codefont)$"$codefont$",$endfor$),
+$endif$


Curious why you used $for$ here. mathfont would normally be expected to be a string in pandoc templates.

This is a bit of an odd one, but Typst uses a vector of fonts because it automatically falls back to the next font if one is not found. This use of $for$ keeps that idea, where you might set (for example) EB Garamond and then Garamond. If EB Garamond is downloaded, then the template uses that. If not, it falls back to the (less featured, but more common) Garamond. This goes to the design choice that Typst will cycle through all the fonts chosen until it finds one available or ends up back at the default.

jgm · 2024-08-23T18:24:52Z

data/templates/default.typst

+$if(citecolor)$
+  citecolor: [$citecolor$],
+$endif$


It works in content like this, with the escaped #?

citecolor: [\#ffffff]

You've tested I assume?

I can confirm that this works, as I'm currently using it in a custom template. If there's something that you don't like about it, happy to brainstorm alternatives.

data/templates/template.typst

jgm · 2024-08-23T18:28:21Z

data/templates/template.typst

-    author: authors.map(author => content-to-string(author.name)),
+    author: authors.map(author => content-to-string(author.name)).join(", ", last: ", and "),


I don't like this because and is an English word; anyone writing in another language would need to change the template. We could consider using &, or just use commas.

Happy to go either way here. I pushed a change to make it &. I can make it anything that you think is least problematic, even removing the last special case.

jgm · 2024-08-23T18:29:24Z

data/templates/template.typst

+    leading: linestretch * 0.65em
+  )


What does the leading do here, and why is it needed with linestretch? Is the result the same as Typst's defaults when the default linestretch is used?

leading sets the space between lines. Here, mixing it with linestretch ensures that a value of 1 * 0.65em = 0.65em or Typst's default single-spaced. For double spacing, then 1.3 = 2 * 0.65em. Here, this setting just makes it explicit so that setting 1 for single, etc, all map onto what is expected.

jgm · 2024-08-23T18:31:02Z

data/templates/template.typst

-      #text(weight: "bold", size: 1.5em)[#title]
+      #text(weight: "bold", size: 1.5em)[#title #if thanks != none {
+            footnote(thanks, numbering: "*")
+            counter(footnote).update(n => n - 1)


I assume this is needed because otherwise the first note would be 2? That's a bit unexpected when numbering is explicitly specified, but maybe Typst works that way? (I'd consider it a bug if so.)

Yes, so the numbering is a wrapper on top of a counter for footnotes. The numbers for footnotes are the counting numbers (1, 2, ...) and don't update even when you change the style. So this simply says: use * for the title, then set it back to 1 so that it can use 1, 2, ... for future footnotes.

christopherkenny · 2025-03-19T02:41:17Z

Sorry for the long delay on this. I've been on the academic job market and a bit crazed. I'm using Quarto + pandoc + Typst for my dissertation and will resolve these over the next few days while in the midst of this.

christopherkenny · 2025-03-19T03:07:40Z

d05ca30 handles the merging with the pagenumber change

christopherkenny · 2025-05-03T02:52:41Z

Hi @jgm. I've rerequested your review for when you have a chance. These are all now set to be less opinionated and better fit with the default typst settings, especially removing colors and fonts unless specified.

christopherkenny · 2025-05-03T03:41:51Z

Apologies, spoke too soon on the last message. Just went through and triple checked things and now it is ready for re-review.

jgm · 2025-05-11T23:02:57Z

data/templates/template.typst

-  }
+    #if abstract != none {
+      block(inset: 2em)[
+        #text(weight: "semibold")[Abstract] #h(1em) #abstract


It would be nice not to force the use of an English word "Abstract" here. People write in other languages, too. If typst doesn't have a way of retrieving a localization of this, we colud introduce $abstract-title$ , and I could set it to a localized default in the Tyspt writer.

jgm · 2025-05-11T23:03:25Z

looks pretty good -- I made one comment about "Abstract."

jgm · 2025-05-11T23:50:43Z

The other thing that needs to be done is modifying MANUAL.txt so that the new variables are documented.
There should be a section "Variables for Typst."

jgm · 2025-05-15T14:24:38Z

TODO (for me): set abstract-title variable to a localized default in the typst writer.

jgm · 2025-05-15T14:27:18Z

  abstractTitle <- translateTerm Abstract

and then when setting template variables,

                  defField "abstract-title" abstractTitle .

christopherkenny · 2025-05-15T14:37:46Z

The MANUAL.txt and abstract-title defaults are now updated. Thanks for the review!

jgm · 2025-05-16T22:34:26Z

Looks like the branch has to be rebased on to main in order to be merged cleanly.

christopherkenny · 2025-05-23T02:39:45Z

So, I just rebased onto main, but I have a strong suspicion that this is not what you wanted. Before doing this, I made a backup branch which is just a copy prior to me rebasing. I can open a clean PR from that if this is not what you wanted.

rnwst · 2025-05-24T03:06:49Z

@christopherkenny -- I had a look at what may have gone wrong with your rebase. My theory is as follows (please correct me if any of this is wrong!):

During your work on this branch, you used the 'Update branch' button a few times to merge jgm:main into christopherkenny:typst-template.
Your pandoc fork hasn't been synced since June 2024 and so the latest commit on christopherkenny:main is correspondingly old.
When you rebased onto main, you therefore rebased onto a commit from June 2024 rather than the latest main. In addition, you ran rebase without the --rebase-merges option, so your merge commits (created previously via the 'Update branch' button) were replaced with all of the respective commits from jgm:main, which is how you ended up with 662 commits on your branch. If the --rebase-merges option had been used, I think the rebase would have had no effect (since the rebase would've been done on the same commit on jgm:main that your branch was branching out from anyway).

To fix this, you could restore your previous version of this branch (via your backup branch), sync your fork (and do a git pull to make sure your local main is up to date as well), and perform an interactive rebase on main, during which you would drop all merge commits.

christopherkenny · 2025-05-24T03:14:20Z

Yeah, that sounds exactly like what happened. Thanks for the help @rnwst. I will get this cleaned up tomorrow.

christopherkenny · 2025-05-24T03:34:49Z

Thanks @rnwst, that was exactly it. All cleaned up now. Thanks for the help!

rnwst · 2025-05-24T12:25:10Z

Looks like you have opted to undo and then redo (after merging jgm:main) the commit that would be causing a merge conflict, instead of rebasing the whole branch on top of latest jgm:main, which would result in a cleaner commit history. Some of the commit messages could also be tidied up, and perhaps some or all of the commits squashed. But I don't know if @jgm is as pedantic as I am, so let's wait to see what he says.

jgm · 2025-05-27T03:41:39Z

I was thinking this could all be rebased into one commit (unless there's some strong reason for having more than one commit for this change).

christopherkenny added 5 commits July 8, 2024 22:08

add thanks to title

2960681

add linestretch (aka leading) in par

c5011d8

add linestretch to test

3b86a98

add font pieces

608bbc2

add link colors

7b5bf8e

agbeans24 approved these changes Aug 13, 2024

View reviewed changes

jgm requested changes Aug 23, 2024

View reviewed changes

jgm force-pushed the main branch from 3ff001b to 3da4d41 Compare December 18, 2024 23:02

christopherkenny added 2 commits March 18, 2025 22:48

Avoid english "and" -> in favor of "&"

932cffa

Merge branch 'main' into typst-template

d05ca30

christopherkenny added 5 commits May 2, 2025 21:22

Merge branch 'jgm:main' into typst-template

9ef0721

fix writer test typst template

85ce47c

fix whitespace

5dfac7c

don't set fonts explicitly

69a0fbf

better conditional, remove colors too

761d16a

christopherkenny requested a review from jgm May 3, 2025 02:51

christopherkenny added 4 commits May 2, 2025 22:59

fix brace

816e016

filecolor / place were lost in merge. fixes that

c98c343

remove duplicate line, fix place spacing

00ea462

should be a block, not a place

512f320

jgm reviewed May 11, 2025

View reviewed changes

jgm force-pushed the main branch from 60c147d to bfcff3e Compare May 12, 2025 00:38

handle $abstract-title$

06aeec6

christopherkenny force-pushed the typst-template branch from eb63b85 to 587dcd4 Compare May 23, 2025 02:31

christopherkenny added 2 commits May 23, 2025 23:27

Merge branch 'jgm:main' into typst-template

963e5b6

Update MANUAL.txt

c86ebe6

christopherkenny force-pushed the typst-template branch from aa96126 to c86ebe6 Compare May 24, 2025 03:33

		author: authors.map(author => content-to-string(author.name)),
		author: authors.map(author => content-to-string(author.name)).join(", ", last: ", and "),

Uh oh!

Adding new features to the Typst template #9970

Are you sure you want to change the base?

Adding new features to the Typst template #9970

Uh oh!

Conversation

christopherkenny commented Jul 10, 2024

Uh oh!

christopherkenny commented Aug 22, 2024

Uh oh!

jgm commented Aug 23, 2024

Uh oh!

jgm left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

christopherkenny commented Mar 19, 2025

Uh oh!

christopherkenny commented Mar 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

christopherkenny commented May 3, 2025

Uh oh!

christopherkenny commented May 3, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jgm commented May 11, 2025

Uh oh!

jgm commented May 11, 2025

Uh oh!

jgm commented May 15, 2025

Uh oh!

jgm commented May 15, 2025

Uh oh!

christopherkenny commented May 15, 2025

Uh oh!

jgm commented May 16, 2025

Uh oh!

christopherkenny commented May 23, 2025

Uh oh!

rnwst commented May 24, 2025

Uh oh!

christopherkenny commented May 24, 2025

Uh oh!

christopherkenny commented May 24, 2025

Uh oh!

rnwst commented May 24, 2025

Uh oh!

jgm commented May 27, 2025

Uh oh!

Uh oh!

christopherkenny commented Mar 19, 2025 •

edited

Loading