How can I get a regex to match #emph[text]?

Questions
1

in my template, I use

  show regex("^—.*") : it => {
    set text(style: "normal")
    align(right, it)
  }

To catch text that apear at the end of my quotes like:

#quote(block: true, quotes: auto,)[#par()[Subhuti, what is called Buddhism, is not Buddhism.]
#par()[— The Buddha, #emph[Diamond Pranjaparamita Sutta]]
]

But it only catches — The Buddha, and stop after that.

I have tried different regex including escaping # with \ or \\ but nothing works.
Thanks

2

Is there a reason why you’re not just using the attribution field?

#quote(
  block: true,
  quotes: auto,
  attribution: [The Buddha, _Diamond Pranjaparamita Sutta_]
)[
  Subhuti, what is called Buddhism, is not Buddhism.
]

page-1
page-11500×222 75.9 KB

3

Thank you for your reply @ssotoen

I actually should have specified the context of my quesiton as I knew about the atribution (which is a nice thing in Typst)

The context of my question is from mdbook-typst and particularly my branch of it. I have been using mdbook-typst to transform my mdbook to a final PDF A5 with custom typst template and it works really well. I can type my book in Obsidian with a mdbook structure and after compiling, I get a nice PDF with footnote, images, quotes etc. transformed from markdown to typst. This is a little dream for me, as I work on a boox tablet and I can generate my book from my obsidian vault via Termux.

Anyway, the thing is that there is no attribution in markdown by default, so I ended up doing:

> Text  
> Text  
>  
> -- Atribution

in my markdown. Then via the typst template I was hoping to left align all my attributions in my quotes in markdown, hence my question.

When I do

  show quote: it => {
    panic(it.body)
  }

in my template, I get this in my terminal:

error: panicked with: sequence(
  sequence(
    parbreak(),
    [Subhuti, what is called Buddhism, is not Buddhism.],
    parbreak(),
  ),
  [ ],
  sequence(
    parbreak(),
    sequence(
      [— The Buddha,],
      [ ],
      emph(body: [Diamond Pranjaparamita Sutta]),
    ),
    parbreak(),
  ),
  [ ],
)

I thought I could access the different fields in the it.body in the show rule to apply different style via if function. But the outcome is that sequence(...), is there a way to access the elements of that sequence? (there is an interesting example here that is even able to give an attribution)

4

You can get the sequence’s items with .children. To build back a sequence there’s no exposed API but as a hack you can get the constructor with [].func(). Example:

#show quote.where(block: true, attribution: none): it => {
  let elems = it.body.children
  let attr-start = elems.position(x => x == [--])
  if attr-start == none {
    return it
  }
  let body = elems.slice(0, attr-start)
  let attr = elems.slice(attr-start + 1)
  let sequence = [].func()
  quote(block: true, attribution: sequence(attr), sequence(body))
}

#quote(block: true)[
  X

  -- Y
]

though in your case it seems you’ll need more elaborate code to parse the sequence.

5

Ok thank you, this seems to do the trick

#show quote.where(block: true, attribution: none): it => {
    let elems = it.body.children
    let last-line = elems.at(elems.len()-2).children.at(1)
    let last-line-array = elems.at(elems.len()-2).children
    let first-char 

    if last-line.has("children") {
      first-char = last-line.children.at(0).text.codepoints().at(0)
    } else {
      first-char = last-line.text.codepoints().at(0)
    }

    if first-char != "—" {
    return it
    }

    let body = elems.slice(0, elems.len()-2)
    let sequence = [].func()
    
    last-line-array = last-line-array.slice(1)
    last-line-array = last-line-array.join().children.map(e => {
      if e.has("text") {e.text.replace("— ", "")}
      else {e}
    })
    quote(block: true, attribution: sequence(last-line-array), sequence(body))
  }

the problem was that if you have some italic or something else in your attribution, then instead of having a simple text as content, you have an array. Maybe there is a easier way to treat both case, but I did it separatly.

Thanks for your help

Here is a update version that takes in account en dash as well (which are dash-dash in obsidian)

#show quote.where(block: true, attribution: none): it => {

    let elems = it.body.children
    let last-line = elems.at(elems.len()-2).children.at(1)
    let last-line-array = elems.at(elems.len()-2).children
    let first-char 

    if last-line.has("children") {
      first-char = last-line.children.at(0).text.codepoints().at(0)
    } else {
      first-char = last-line.text.codepoints().at(0)
    }

    if first-char != "—" and first-char != "–" { // second case is for -- used in markdown
    return it
    }

    let body = elems.slice(0, elems.len()-2)
    let sequence = [].func()

      last-line-array = last-line-array.slice(1).join().children
      last-line-array = last-line-array.enumerate().map(((i, e)) => {
      if i == 0 and e == [–] {""} // this is for -- used in markdown
      else if i == 1 and e == [ ] and last-line-array.at(0) == [–] {""}
      else if e.has("text") {e.text.replace("— ", "")} // this is for — used in markdown
      else {e}
    })

    quote(block: true, attribution: sequence(last-line-array), sequence(body))
  }
1 Like