Content OR String inside link function

[trail ginkgo]

I am creating a cover page. And you can set your email. But that email can be of type content or string. This is because you can use the custom function sensitive() which allows the user to censor information when a flag is set. This means that the content is SOMETIMES censored. But the string can be used in case the user does not want to censor anything at all.

So where is my code that I have tried so far:

if type(author.email) == content {
    link("mailto:" + author.email)
} else {
    link("mailto:" + author.email)[#author.email]
}

When its not content, it works fine. But I cant get the content working for the life of me.

I would love to just say, this content is now a string, and all things that are not characters you can ignore

This for example (not email but name) can look like 2 things:

And I want the same for the email.

this is the branch im working on right now

Email part is in the tempalte file on line 235

[tender adder]

if a part of the email is wrapped in some other content structure (sensitive), then you need to recursively go through the content (email) to just get the string of the email

it is required to link to the email

Though, do note that censoring the email would not really help in this way as link will still link to the correct email

[trail ginkgo]

ugh haha

well i do have access to the cencorded flag

how do you do the recursive thing?

because i could not do that if the flag is set, and not put anything on the doc

[trail ginkgo]

yeah i gave up on it

lol

[vale dirge]

you cant censor an URI

the string parameter of link is where it links to, not what is displayed

[trail ginkgo]

In the sensitive func I replace the characters

so the link will be a link, but invalid

[vale dirge]

yeah but that doesnt make technical sense

you cant make a link to (square) (square)

if ya get me

like you cant tell someone "visit my website at https:// (black rectangle).com"

[trail ginkgo]

yeah

[vale dirge]

rather what you need to do is to remove the link call

[trail ginkgo]

but if you dont want to censored you dont put it in the func

or dont set the flag

[vale dirge]

let me see if i understand then

are you trying to make it so #censor[aaaa #link("mailto:...", body: [a]) bbb] will automatically delete (/ turn useless) links inside it?

[trail ginkgo]

if the censor flag is set yes. The censor func I made only changes the text if a flag is set.

[vale dirge]

okay

then that changes the perspective I had

haha

then what you need is something like

#let censor(x) = {
  show link: it => it.at("body", default: if type(it.target) == str { "x" * it.target.len() } else { "xxx" })
  // your other show rules to show black boxes
  x
}

just an example

you can also just show link: none

if you just want to get rid of it

[trail valley]

There is a snippet that lets you extract plain text from some content: https://sitandr.github.io/typst-examples-book/book/typstonomicon/extract_plain_text.html

I'm not sure how directly applicable it is (if you want to use it do check it; you're trying to ensure privacy so don't just trust it) but the important part is that it recurses over the kinds of contents it encounters, and accumulates the stuff it finds into a value (here, a string). What I'd do is probably adapt that so that instead of (or in addition to) returning plain text, it returns whether a sensitive value appears anywhere in the content.

[trail ginkgo]

yo thanks! :D

ill defo check it out

[vale dirge]

I think show rules would be much more reliable

But that snippet i guess can be interesting to use in other contexts

[trail valley]

so let's say you have #email[#sensitive[username]@gmail.com] and this should become [email protected] or #link("[email protected]") depending on whether censoring is enabled, would that be achievable with show rules? That's what I understood it is, in minimal form and maybe a bit of my spin on it, that Tom wants.

[vale dirge]

The idea is to use links for emails, and then you just use #show link: ...

[vale dirge]

As seen here

Though you could have to adapt the code to only target links containing mailto

[tender adder]

well, that would only work if it was used as censor(link(...)), from their above example it seems like the usage is link(...censor(...)...)

[trail valley]

ah ok, after playing around with it I think I see what you mean. Although I thought one goal was that if the link text (i.e. possibly sensitive content) was an email, the link destination would default to be just that. That's how it sounded over in quick-questions - and that, iiuc, would require walking the content.

PS: I the snippet should say it.dest instead of it.target, just something to watch out for @trail ginkgo

[trail valley]

the usage would be link(...sensitive(...)...) (or maybe something like link(...)[...#sensitive(...)...]) and censor would recognize the presence of sensitive content.

[tender adder]

ah, right. I was thinking to still extract the link if there's arbitrary content, which is probably not needed here

[trail ginkgo]

I have come one step closer

My current problem is as follows:

I have a content object which does not contain the field text. It only has a function which is my sensitive function. If I just do it it works but I can not concatenate it with other strings.

I want to basically get the result of the function

not a reference to the function itself

// author: ntjess
#let stringify-by-func(it) = {
  let func = it.func()
  return if func in (parbreak, pagebreak, linebreak) {
    "\n"
  } else if func == smartquote {
    if it.double { "\"" } else { "'" } // "
  } else if it.fields() == (:) {
    // a fieldless element is either specially represented (and caught earlier) or doesn't have text
    ""
  } else {
    panic("Not sure how to handle type `" + repr(func) + "`")
  }
}

#let plain-text(it) = {
  return if type(it) == str {
    it
  } else if type(it) == content {
    let text_str = ""
    if it.has("children") {
      for child in it.children {
        if child.has("text") {
          text_str = text_str + child.text
        } else {
          text_str = text_str + "REDACTED"
        }
      }
    }
    text_str
  } else if it == [ ] {
    " "
  } else if it.has("children") {
    it.children.map(plain-text).join()
  } else if it.has("body") {
    plain-text(it.body)
  } else if it.has("text") {
    it.text
  } else {
    stringify-by-func(it)
  }
}

email: [#sensitive("some")\@mail.com],

well

that works half

now it only censors lol

gosh darnit haha

i need to get rid of that func ref somehow

BUT, all the other functionality works

Updated code

// author: ntjess
#let stringify-by-func(it) = {
  let func = it.func()
  return if func in (parbreak, pagebreak, linebreak) {
    "\n"
  } else if func == smartquote {
    if it.double { "\"" } else { "'" } // "
  } else if it.fields() == (:) {
    // a fieldless element is either specially represented (and caught earlier) or doesn't have text
    ""
  } else {
    panic("Not sure how to handle type `" + repr(func) + "`")
  }
}

#let plain-text(it) = {
  return if type(it) == str {
    it
  } else if type(it) == content {
    let text_str = ""
    if it.has("children") {
      for child in it.children {
        if child.has("text") {
          text_str = text_str + child.text
        } else {
          // The issue is here.
          // Somehow convert this to a string
          text_str = text_str + "REDACTED"
        }
      }
    }
    text_str
  } else if it == [ ] {
    " "
  } else if it.has("children") {
    it.children.map(plain-text).join()
  } else if it.has("body") {
    plain-text(it.body)
  } else if it.has("text") {
    it.text
  } else {
    stringify-by-func(it)
  }
}

so close yet so far

[vale dirge]

Then you'd have to use a show rule on text

I.e.
#show regex("\S+@\S+"): none for example would remove all emails theoretically

In practice there’s a bug

Where you can’t match on symbols used by syntax

For example @ is used for ref

So that might not work