Add Collation-Based String Sorting

Cargo.toml

@@ -34,6 +34,8 @@ biblatex = { git = "https://github.com/typst/biblatex.git", features = [
 ciborium = { version = "0.2.1", optional = true }
 clap = { version = "4", optional = true, features = ["cargo"] }
 strum = { version = "0.26", features = ["derive"], optional = true }
+icu_collator = "2.0.0"
+icu_locale = "2.0.0"
 
 [dev-dependencies]
 heck = "0.5"

src/csl/sort.rs

@@ -5,13 +5,49 @@ use citationberg::{
     DemoteNonDroppingParticle, InheritableNameOptions, LocaleCode, LongShortForm, Sort,
     SortDirection, SortKey,
 };
+use icu_collator::options::CollatorOptions;
+use icu_collator::Collator;
+use icu_locale::Locale as IcuLocale;
 
 use crate::csl::rendering::RenderCsl;
 use crate::csl::BufWriteFormat;
 
 use super::taxonomy::EntryLike;
 use super::{CitationItem, InstanceContext, StyleContext};
 
+trait CollationOrd: Ord {
+    fn collation_cmp(&self, other: &Self, locale: LocaleCode) -> Ordering;
+}
+
+impl CollationOrd for str {
+    fn collation_cmp(&self, other: &Self, locale: LocaleCode) -> Ordering {
+        if let Ok(locale) = locale.0.parse::<IcuLocale>() {
+            let options = CollatorOptions::default();
+            let collator = Collator::try_new(locale.into(), options).unwrap();
+            collator.compare(self, other)
+        } else {
+            self.cmp(other)
+        }
+    }
+}
+
+impl<T: CollationOrd> CollationOrd for Option<T> {
+    fn collation_cmp(&self, other: &Self, locale: LocaleCode) -> Ordering {
+        match (self, other) {
+            (Some(a), Some(b)) => a.collation_cmp(b, locale),
+            (Some(_), None) => Ordering::Greater,
+            (None, Some(_)) => Ordering::Less,
+            (None, None) => Ordering::Equal,
+        }
+    }
+}
+
+impl CollationOrd for String {
+    fn collation_cmp(&self, other: &Self, locale: LocaleCode) -> Ordering {
+        CollationOrd::collation_cmp(self as &str, other, locale)
+    }
+}
+
 impl StyleContext<'_> {
     /// Retrieve the ordering of two entries according to the given sort key.
     fn cmp_entries<T: EntryLike>(
@@ -32,7 +68,7 @@ impl StyleContext<'_> {
                     .resolve_standard_variable(LongShortForm::default(), *s)
                     .map(|s| s.to_string().to_lowercase());
 
-                a.cmp(&b)
+                a.collation_cmp(&b, self.locale())
             }
             SortKey::Variable { variable: Variable::Date(d), .. } => {
                 let a = a.entry.resolve_date_variable(*d);
@@ -133,7 +169,7 @@ impl StyleContext<'_> {
                 let a_rendered = render(a, a_idx);
                 let b_rendered = render(b, b_idx);
 
-                a_rendered.cmp(&b_rendered)
+                a_rendered.collation_cmp(&b_rendered, self.locale())
             }
         };
 

tests/citeproc-pass.txt

@@ -400,6 +400,7 @@ sort_CitationNumberPrimaryAscendingViaMacroBibliography
 sort_CitationNumberPrimaryAscendingViaVariableBibliography
 sort_CitationSecondaryKey
 sort_CiteGroupDelimiter
+sort_ConditionalMacroDates
 sort_DaleDalebout
 sort_DateMacroSortWithSecondFieldAlign
 sort_DateVariable

tests/local/sort_Unicode.txt

@@ -0,0 +1,120 @@
+>>==== MODE ====>>
+citation
+<<==== MODE ====<<
+
+>>==== RESULT ====>>
+Ambroise-Rendu; Ère; Łakomy; Roux; Žižek
+<<==== RESULT ====<<
+
+>>==== CITATION-ITEMS ====>>
+[
+    [
+        {
+            "id": "ITEM-1"
+        },
+        {
+            "id": "ITEM-2"
+        },
+        {
+            "id": "ITEM-3"
+        },
+        {
+            "id": "ITEM-4"
+        },
+        {
+            "id": "ITEM-5"
+        }
+    ]
+]
+<<==== CITATION-ITEMS ====<<
+
+>>==== CSL ====>>
+<style 
+      xmlns="http://purl.org/net/xbiblio/csl"
+      class="note"
+      version="1.0">
+  <info>
+    <title>Test fixture</title>
+    <id>http://citationstyles.org/tests/fixture</id>
+    <link href="http://citationstyles.org/tests/fixture" rel="self"/>
+    <link href="http://citationstyles.org/documentation/text" rel="documentation"/>
+    <category citation-format="author-date"/>
+    <updated>2014-04-30T13:19:38+00:00</updated>
+    <rights license="http://creativecommons.org/licenses/by-sa/3.0/">This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License</rights>
+  </info>
+  <macro name="author">
+    <names variable="author">
+      <name form="short" />
+    </names>
+  </macro>
+  <citation>
+    <sort>
+      <key macro="author" />
+    </sort>
+    <layout delimiter="; ">
+      <text macro="author" />
+    </layout>
+  </citation>
+</style>
+<<==== CSL ====<<
+
+>>==== INPUT ====>>
+[
+  {
+    "author": [
+      {
+        "family": "Ambroise-Rendu",
+        "given": "Marc"
+      }
+    ],
+    "id": "ITEM-1",
+    "type": "book"
+  },
+  {
+    "author": [
+      {
+        "family": "Roux",
+        "given": "Nicolas"
+      }
+    ],
+    "id": "ITEM-2",
+    "type": "book"
+  },
+  {
+    "author": [
+      {
+        "family": "Łakomy",
+        "given": "Henryk"
+      }
+    ],
+    "id": "ITEM-3",
+    "type": "book"
+  },
+  {
+    "author": [
+      {
+        "family": "Ère",
+        "given": "Informatique"
+      }
+    ],
+    "id": "ITEM-4",
+    "type": "book"
+  },
+  {
+    "author": [
+      {
+        "family": "Žižek",
+        "given": "Slavoj"
+      }
+    ],
+    "id": "ITEM-5",
+    "type": "book"
+  }
+]
+<<==== INPUT ====<<
+
+
+
+>>===== VERSION =====>>
+1.0
+<<===== VERSION =====<<