fix(glossy): fix forward linking (broken themes)

Fixes [#19](https://github.com/swaits/typst-collection/issues/19).

Themes must emit `entry.label` for terms to be able to forward-link to
their glossary entrie.

.build.yml

@@ -1,13 +1,27 @@
 image: alpine/edge
+packages:
+  - git
+
 secrets:
   - 88cce185-d2ec-4573-9546-e33f36f79bbf
-sources:
+
-  - git+ssh://git@git.sr.ht/~swaits/typst-collection
+environment:
+  # repo and user names on both sr.ht and github
+  REPO: typst-collection
+  USER: swaits
+
 tasks:
-  - mirror-to-github: |
+  - fetch-all-branches: |
-      cd ~/typst-collection
+      cd "$HOME/$REPO"
-      ssh-keyscan -t rsa github.com >> ~/.ssh/known_hosts
-      # Explicitly fetch all branches & tags
       git fetch origin '+refs/heads/*:refs/heads/*' '+refs/tags/*:refs/tags/*' --prune
-      git remote add github git@github.com:swaits/typst-collection.git
+
+  - get-github-public-key: |
+      ssh-keyscan -t rsa github.com >> "$HOME/.ssh/known_hosts"
+
+  - add-github-remote: |
+      cd "$HOME/$REPO"
+      git remote add github "git@github.com:$USER/$REPO.git"
+
+  - mirror-to-github: |
+      cd "$HOME/$REPO"
       git push --mirror github

glossy/README.md

@@ -41,7 +41,7 @@ different goals:
 ### Import the package
 
 ```typst
-#import "@preview/glossy:0.4.0": *
+#import "@preview/glossy:0.5.1": *
 ```
 
 ### Defining Glossary Terms
@@ -186,8 +186,9 @@ Display the glossary using the `glossary()` function:
 
 ```typst
 #glossary(
-  title: "Web Development Glossary", // Optional: defaults to Glossary
+  title: "Web Development Glossary", // Optional: defaults to Glossary theme:
-  theme: my-theme, // Optionaal: defaults to theme-twocol
+  my-theme, // Optional: defaults to theme-academic
+  sort: true, // Optional: whether or not to sort the glossary
   ignore-case: false, // Optional: ignore case when sorting terms
   groups: ("Web")  // Optional: Filter to specific groups
 )
@@ -252,7 +253,7 @@ Customize glossary appearance by defining a theme with three functions:
   // Individual glossary entry
   entry: (entry, index, total) => {
     // index = entry index, total = total entries in group
-    let output = [#entry.short]
+    let output = [#entry.short#entry.label] // **NOTE:** Label here!
     if entry.long != none {
       output = [#output -- #entry.long]
     }
@@ -279,6 +280,9 @@ Entry fields available to themes:
 - `label`: Term's dictionary label
 - `pages`: Linked page numbers where term appears
 
+**NOTE:** If the theme does not emit `entry.label`, linking form terms to their
+glossary entry will not work.
+
 ## License
 
 This project is licensed under the MIT License.

glossy/bin/autotest.sh

@@ -0,0 +1,6 @@
+watchexec \
+  --watch . \
+  --clear \
+  --ignore 'tests/**/diff/**' \
+  --ignore 'tests/**/out/**' \
+  "typst-test r"

glossy/src/gloss.typ

@@ -23,7 +23,7 @@
 // Returns:
 //   dictionary: Normalized dictionary with all expected keys populated
 //
-// Throws:
+// Panics:
 //   - If 'short' key is missing when using normal (non-abbreviated) syntax
 //   - If 'short' value is not a string
 //   - If 'short' value is an empty string
@@ -97,7 +97,7 @@
 // Returns:
 //   dictionary: The glossary entry associated with the key
 //
-// Throws:
+// Panics:
 //   - If the key does not exist in the glossary
 //
 #let __get_entry(key) = {
@@ -153,7 +153,7 @@
 //
 // Parameters:
 //   key (string): The glossary entry key
-// 
+//
 // Returns:
 //   boolean: If the term is visible in the glossary (and the label is unique)
 //
@@ -187,7 +187,7 @@
 //   content: The fully formatted term, including optional article, capitalization,
 //            usage tracking metadata, and the chosen form (short, long, or both).
 //
-#let __gls(key, modifiers: array, show-term: function, term-links: true) = {
+#let __gls(key, modifiers: array, format-term: function, show-term: function, term-links: true) = {
   // ---------------------------------------------------------------------------
   // Check for illegal modifier combinations
   // ---------------------------------------------------------------------------
@@ -248,27 +248,6 @@
     }
   }
 
-  // get_term(mode): Returns the appropriate term string based on the mode.
-  // mode is one of "short", "long", or "both".
-  //   - "short": Returns the short form.
-  //   - "long": Returns the long form.
-  //   - "both": Returns "Long form (short form)".
-  // We rely on prior logic to ensure that when mode = "both" or "long", a long form is available.
-  let get_term = (mode) => {
-    let short_form = pluralize_term(entry.short, entry.plural)
-    let long_form = pluralize_term(entry.long, entry.longplural) // none safe here
-    if mode == "short" {
-      // Just the short form.
-      short_form
-    } else if mode == "long" {
-      // Just the long form.
-      long_form
-    } else {
-      // mode == "both": "Long form (short form)".
-      long_form + " (" + short_form + ")"
-    }
-  }
-
   // get_article(mode): Returns the appropriate article string based on the mode.
   //   - If wants_article is false, returns an empty string.
   //   - If mode is "short", return the short article.
@@ -337,19 +316,33 @@
   // Determine mode using the helper function
   let mode = determine_mode(wants_both, wants_long, wants_short, is_first_use, long_available)
 
-  // Get the article and term for the chosen mode, then apply capitalization if requested
+  // Pluralize (if requested)
-  let (article, term) = capitalize_term(get_article(mode), get_term(mode))
+  let short-form = pluralize_term(entry.short, entry.plural)
+  let long-form = pluralize_term(entry.long, entry.longplural) // `none` safe here
+
+  // Apply format-term() to get the term
+  let formatted-term = format-term(mode, short-form, long-form)
+  if type(formatted-term) != "string" {
+    // This is because we still need to capitalize the term, and we cannot do
+    // that with content, thus requiring a string here.
+    // TODO: consider if we want to capitalize *before* this. First intuition is
+    // "no".
+    panic("Your cutsom format-term() function must return a string.")
+  }
+
+  // Get the article, then capitalize either the article or term (if requested)
+  let (article, term) = capitalize_term(get_article(mode), formatted-term)
 
   // ---------------------------------------------------------------------------
   // Construct and return the final output
   // ---------------------------------------------------------------------------
   context {
-    let text = if term-links and __has_glossary_entry(key) {
+    let linked-term = if term-links and __has_glossary_entry(key) {
-      link(label(key), term) 
+      link(label(key), term)
     } else {
       term
     }
-    [#article#show-term(text)#metadata(term)#__term_label(key)]
+    [#article#show-term(linked-term)#metadata(term)#__term_label(key)]
   // |^^^^^^^|^^^^^^^^^      |^^^^^^^^      |^^^^^^^^^^^^
   // \_art.  |               |              |
   //         \_ apply user formatting function to the term
@@ -384,7 +377,48 @@
     })
     .dedup(key: ((loc, page)) => page) // deduplicate by page
     .map(((loc, page)) => link(loc, page)) // create links
-    .join(", ")    
+    .join(", ")
+}
+
+// Formats a term string based on the specified mode.
+//
+// Given mode, long-form, and short-form, this function returns the formatted
+// term based on the mode. When mode is ...
+//
+// - "short" -> returns "short-form"
+// - "long" -> returns "long-form"
+// - "both" -> returns "short-form (long-form)"
+//
+// Parameters:
+//   mode (string): The mode in which to format the term. Possible values are "short", "long", or "both".
+//   short-form (string): The short form of the term.
+//   long-form (string): The long form of the term.
+//
+// Returns:
+//   string: The formatted term based on the specified mode.
+//
+#let __default-format-term(mode, short-form, long-form) = {
+    if mode == "short" {
+      short-form
+    } else if mode == "long" {
+      long-form
+    } else {
+      // mode assumed to be "both"
+      long-form + " (" + short-form + ")"
+    }
+}
+
+// Styles a term to control its display in a document.
+//
+// Parameters:
+//   term-body (content'ish): the unstyled output from __gls()
+//
+// Returns:
+//   content: display ready content (assumed based on `term-body`)
+//
+#let __default-show-term(term-body) = {
+  // Default: just render it like normal text
+  term-body
 }
 
 // Initializes the glossary system and sets up term reference handling.
@@ -407,12 +441,13 @@
 // Returns:
 //   content: Processed document content with glossary functionality enabled
 //
-// Throws:
+// Panics:
 //   - If entries parameter is not a dictionary
 //
 #let init-glossary(
   entries,
-  show-term: (term-body) => term-body,
+  format-term: __default-format-term,
+  show-term: __default-show-term,
   term-links: true,
   body
 ) = context {
@@ -423,6 +458,7 @@
     let checked-entry = z.parse(entry, dict-schema, scope: ("dictionary entry",))
     checked-entries.insert(checked-key, checked-entry)
   }
+  let checked-format-term = z.parse(show-term, z.function(), scope: ("format-term",))
   let checked-show-term = z.parse(show-term, z.function(), scope: ("show-term",))
   let checked-body = z.parse(body, z.content(), scope: ("body",))
 
@@ -457,7 +493,7 @@
     // Now see if this is an actual glossary term key
     if __has_entry(key) {
       // Found in dictionary, render via __gls()
-      __gls(key, modifiers: modifiers.map(lower), show-term: show-term, term-links: term-links)
+      __gls(key, modifiers: modifiers.map(lower), format-term: format-term, show-term: show-term, term-links: term-links)
     } else {
       // Not one of ours, so just pass it through
       r
@@ -499,12 +535,13 @@
 // Returns:
 //   content: Formatted glossary content
 //
-// Throws:
+// Panics:
 //   - If a requested group doesn't exist
 //
 #let glossary(
   title: "Glossary",
-  theme: theme-twocol,
+  theme: theme-academic,
+  sort: true,
   ignore-case: false,
   groups: (),
 ) = context {
@@ -559,20 +596,24 @@
         ))
       }
     }
-  
+
     // Add non-empty groups to output
     if current_entries.len() > 0 {
       group = if group == none { "" } else { group }
-  
+
       // sort entries by case insensitivity if requested
-      let sorted_entries = current_entries
+      let sorted_entries = if sort {
-        // 1. create array of tuples with (lower [if ignore-case], entry)
+        current_entries
-        .map(e => { if ignore-case { (lower(e.short), e) } else { (e.short, e) } })
+          // 1. create array of tuples with (lower [if ignore-case], entry)
-        // 2. sort the tuples (by first element then second)
+          .map(e => { if ignore-case { (lower(e.short), e) } else { (e.short, e) } })
-        .sorted(key: t => t.first())
+          // 2. sort the tuples (by first element then second)
-        // 3. strip away the tuple's first element, leaving an array of entries
+          .sorted() // NOTE: sorted() is NOT language-aware
-        .map(t => t.last())
+          // 3. strip away the tuple's first element, leaving an array of entries
-  
+          .map(t => t.last())
+      } else {
+        current_entries
+      }
+
       // add entries to this group's output map
       output.insert(group, sorted_entries)
     }

glossy/src/themes/academic.typ

@@ -40,7 +40,7 @@
           grid(
             columns: (1fr, auto),
             gutter: 0.75em,
-            [#short-display#long-display#description],
+            [#short-display#long-display#description#entry.label],
             text(fill: rgb("#666666"), entry.pages)
           )
         }

glossy/src/themes/chicago-index.typ

@@ -37,7 +37,7 @@
           grid(
             columns: (1fr, auto),
             gutter: 1em,
-            [#short-display#entry.label#long-display#description],
+            [#short-display#entry.label#long-display#description#entry.label],
             [#entry.pages]
           )
         }

glossy/src/themes/table.typ

@@ -27,6 +27,6 @@
   },
 
   entry: (entry, index, total) => {
-    (entry.short, entry.long, entry.description, entry.pages)
+    (entry.short + entry.label, entry.long, entry.description, entry.pages)
   }
 )

glossy/src/utils.typ

@@ -203,7 +203,7 @@
 
   if is_acronym {
     // Use acronym rules
-    let first_char = upper_word.slice(0,1)
+    let first_char = upper_word.codepoints().slice(0,1).first()
     if acronym_vowels.contains(first_char) {
       "an"
     } else {
@@ -211,7 +211,7 @@
     }
   } else {
     // Non-acronym rules
-    let first_char = lower_word.slice(0,1)
+    let first_char = lower_word.codepoints().slice(0,1).first()
 
     // Special cases for words matching a set of known 'silent h' prefixes
     let silent_h_words = (

glossy/tests/format-term/test.typ

@@ -0,0 +1,208 @@
+#import "/lib.typ": *
+
+#let myGlossary = (
+  html: (
+    short: "HTML",
+    long: "Hypertext Markup Language",
+    description: "A standard language for creating web pages",
+    group: "Web"),
+  css: (
+    short: "CSS",
+    long: "Cascading Style Sheets",
+    description: "A language used for describing the presentation of a document",
+    group: "Web"),
+  tps: (
+    short: "TPS",
+    long: "test procedure specification",
+    description: "A document on how to run all the test procedures"),
+  concise: "Concise syntax!",
+  shortonly: (
+    short: "shortonly!!!"),
+  unused: "Unused term, which shouldn't print in the glossary."
+)
+
+#set page(width: 6.5in, height: auto, margin: 1em)
+
+#let test-content = {
+  table(
+    columns: 3,
+    table.header(
+      [`short`], [`long`], [`both`],
+    ),
+
+    [@html:short],      [@html:long],      [@html:both],
+    [@css:short],       [@css:long],       [@css:both],
+    [@tps:short],       [@tps:long],       [@tps:both],
+    [@concise:short],   [@concise:long],   [@concise:both],
+    [@shortonly:short], [@shortonly:long], [@shortonly:both],
+
+    table.hline(stroke: 2pt),
+
+    [@html:short:cap],        [@html:long:cap],        [@html:both:cap],
+    [@a:css:short],           [@a:css:long],           [@a:css:both],
+    [@tps:short:pl],          [@tps:long:pl],          [@tps:both:pl],
+    [@concise:a:cap:short],   [@concise:a:cap:long],   [@concise:a:cap:both],
+    [@an:shortonly:short],    [@an:shortonly:long],    [@an:shortonly:both],
+  )
+
+  [@a:tps:cap is @tps:def.]
+}
+
+// ---------------------------------------------------------------------------------
+
+#let format-basic(mode, short-form, long-form) = {
+  if mode == "short" {
+    short-form
+  } else if mode == "long" {
+  long-form
+  } else {
+    long-form + " (" + short-form + ")"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-basic)[#test-content]
+
+// ---------------------------------------------------------------------------------
+
+#let format-reverse(mode, short-form, long-form) = {
+  if mode == "short" {
+    short-form
+  } else if mode == "long" {
+    long-form
+  } else {
+    short-form + " [" + long-form + "]"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-reverse)[#test-content]
+
+// ---------------------------------------------------------------------------------
+
+#let format-uppercase-short(mode, short-form, long-form) = {
+  let short-up = upper(short-form);
+  if mode == "short" {
+    short-up
+  } else if mode == "long" {
+    long-form
+  } else {
+    long-form + " (" + short-up + ")"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-uppercase-short)[#test-content]
+
+// ---------------------------------------------------------------------------------
+
+#let format-prefix-suffix(mode, short-form, long-form) = {
+  if mode == "short" {
+    "Term: " + short-form + "."
+  } else if mode == "long" {
+    "Term: " + long-form + "."
+  } else {
+    "Term: " + long-form + " (" + short-form + ")."
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-prefix-suffix)[#test-content]
+
+// ---------------------------------------------------------------------------------
+
+#let format-conditional-separator(mode, short-form, long-form) = {
+  let separator = if short-form.len() > 4 { " / " } else { ", " };
+  let has_long = long-form != none and long-form != "";
+
+  if mode == "short" {
+    short-form
+  } else if mode == "long" {
+    if has_long { long-form } else { short-form }
+  } else {
+    if has_long {
+      long-form + separator + short-form
+    } else {
+      short-form
+    }
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-conditional-separator)[#test-content]
+
+// ---------------------------------------------------------------------------------
+// 1. Braces and Labels
+// Idea: Use different bracket styles and labels based on the mode.
+#let format-braces(mode, short-form, long-form) = {
+  if mode == "short" {
+    "{" + short-form + "}"
+  } else if mode == "long" {
+    "[" + long-form + "]"
+  } else {
+    long-form + " <" + short-form + ">"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-braces)[#test-content]
+
+// ---------------------------------------------------------------------------------
+// 2. Mode-Based Prefix
+// Idea: Add a prefix indicating the mode, and then show the term(s).
+#let format-mode-prefix(mode, short-form, long-form) = {
+  let prefix = if mode == "short" { "S:" } else if mode == "long" { "L:" } else { "B:" };
+  if mode == "short" {
+    prefix + short-form
+  } else if mode == "long" {
+    prefix + long-form
+  } else {
+    prefix + long-form + " & " + short-form
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-mode-prefix)[#test-content]
+
+// ---------------------------------------------------------------------------------
+// 3. Conditional Fallback for Missing Long
+// Idea: If no long-form is provided, use the short-form in all cases.
+#let format-fallback-long(mode, short-form, long-form) = {
+  let has_long = long-form != none and long-form != "";
+  if mode == "short" {
+    short-form
+  } else if mode == "long" {
+    if has_long { long-form } else { short-form }
+  } else {
+    if has_long { long-form + " (" + short-form + ")" } else { short-form }
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-fallback-long)[#test-content]
+
+// ---------------------------------------------------------------------------------
+// 4. Uppercase Both Terms in Both Mode
+// Idea: Always uppercase for both mode, normal otherwise.
+#let format-uppercase-both(mode, short-form, long-form) = {
+  if mode == "short" {
+    short-form
+  } else if mode == "long" {
+    long-form
+  } else {
+    upper(long-form) + " (" + upper(short-form) + ")"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-uppercase-both)[#test-content]
+
+// ---------------------------------------------------------------------------------
+// 5. Append Length Info
+// Idea: In both mode, append the length of each string.
+#let format-length-info(mode, short-form, long-form) = {
+  let s_len = str(short-form.len())
+  let l_len = if long-form != none and long-form != "" { str(long-form.len()) } else { "0" }
+
+  if mode == "short" {
+    short-form + " [len:" + s_len + "]"
+  } else if mode == "long" {
+    long-form + " [len:" + l_len + "]"
+  } else {
+    long-form + " (" + short-form + ") [L:" + l_len + ", S:" + s_len + "]"
+  }
+}
+
+#init-glossary(myGlossary, format-term: format-length-info)[#test-content]
+

glossy/tests/forward-links/test.typ

@@ -0,0 +1,45 @@
+#import "/lib.typ": *
+#import "@preview/glossarium:0.5.1": *
+
+#set page(height: auto, width: 4.5in)
+#show link: set text(fill: red)
+
+// -----------------------------------------------------------------------------
+
+= `glossy`:
+
+#box(
+  stroke: 1pt,
+  inset: 1em,
+  [
+    #show: init-glossary.with((WWW: "World Wide Web"))
+
+    This should be a link: @WWW
+
+    This should be a link, too: @WWW
+
+    #glossary()
+  ]
+)
+
+// -----------------------------------------------------------------------------
+
+= `glossarium`:
+
+#box(
+  stroke: 1pt,
+  inset: 1em,
+  [
+    #show: make-glossary
+    #let entry-list = (
+      (key: "ML", short: "ML", long: "Machine Learning"),
+    )
+    #register-glossary(entry-list)
+
+    This should be a link: @ML
+
+    This should be a link, too: @ML
+
+    #print-glossary(entry-list)
+  ]
+)

glossy/tests/no-sort/glossary.yml

@@ -0,0 +1,23 @@
+first:
+  short: "DAL"
+  long: "first"
+
+second:
+  short: "cab"
+  long: "second"
+
+third:
+  short: "delta"
+  long: "third"
+
+fourth:
+  short: "COO"
+  long: "fourth"
+
+fifth:
+  short: "cuckoo"
+  long: "fifth"
+
+sixth:
+  short: "ČR"
+  long: "sixth"

glossy/tests/no-sort/test.typ

@@ -0,0 +1,13 @@
+#import "/lib.typ":*
+
+#show: init-glossary.with(yaml("glossary.yml"))
+
+@first, @second, @third, @fourth, @fifth, @sixth
+
+#glossary(ignore-case: false, sort: true)
+#line(length: 20em)
+#glossary(ignore-case: true, sort: true)
+#line(length: 20em)
+#glossary(ignore-case: false, sort: false)
+#line(length: 20em)
+#glossary(ignore-case: true, sort: false)

glossy/tests/utf-handling/glossary.yml

@@ -0,0 +1,23 @@
+cab:
+  short: "cab"
+  long: "taxi cab"
+
+coo:
+  short: "COO"
+  long: "Chief Operations Officer"
+
+cuckoo:
+  short: "cuckoo"
+  long: "really crazy"
+
+delta:
+  short: "delta"
+  long: "difference between two values"
+
+dal:
+  short: "DAL"
+  long: "Delta Airlines"
+
+čr:
+  short: "ČR"
+  long: "Česká Republika"

glossy/tests/utf-handling/test.typ

@@ -0,0 +1,20 @@
+#import "/lib.typ":*
+
+#set text(lang: "cs", region: "CZ")
+
+#show: init-glossary.with(yaml("glossary.yml"))
+#glossary(ignore-case: true, theme: theme-twocol)
+
+#line(length: 20em)
+
+@čr
+
+@cab
+
+@coo
+
+@cuckoo
+
+@delta
+
+@dal

glossy/typst.toml

@@ -1,6 +1,6 @@
 [package]
 name = "glossy"
-version = "0.4.0"
+version = "0.5.1"
 entrypoint = "lib.typ"
 authors = ["Stephen Waits <steve@waits.net>"]
 license = "MIT"