From 0c94d2b34e77edbd116c9f7be591ca710363b844 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?S=C3=A9bastien=20d=27Herbais=20de=20Thun?=
 <sebastien.d.herbais@gmail.com>
Date: Tue, 18 Jul 2023 15:23:56 +0200
Subject: [PATCH] Adding `dedup` to `array` (#1738)

---
 crates/typst/src/eval/array.rs   | 32 ++++++++++++++++++++++++++++++++
 crates/typst/src/eval/methods.rs |  1 +
 docs/reference/types.md          | 15 +++++++++++++++
 tests/typ/compiler/array.typ     | 15 +++++++++++++++
 4 files changed, 63 insertions(+)

diff --git a/crates/typst/src/eval/array.rs b/crates/typst/src/eval/array.rs
index a7a1387b7..5d5fdcdbd 100644
--- a/crates/typst/src/eval/array.rs
+++ b/crates/typst/src/eval/array.rs
@@ -398,6 +398,38 @@ impl Array {
             .map(|(i, value)| array![i, value.clone()].into_value())
             .collect()
     }
+
+    /// Deduplicates all items in the array.
+    pub fn dedup(&self, vm: &mut Vm, key: Option<Func>) -> SourceResult<Self> {
+        let mut out = EcoVec::with_capacity(self.0.len());
+        let mut key_of = |x: Value| match &key {
+            // NOTE: We are relying on `comemo`'s memoization of function
+            // evaluation to not excessively reevaluate the `key`.
+            Some(f) => f.call_vm(vm, Args::new(f.span(), [x])),
+            None => Ok(x),
+        };
+
+        // This algorithm is O(N^2) because we cannot rely on `HashSet` since:
+        // 1. We would like to preserve the order of the elements.
+        // 2. We cannot hash arbitrary `Value`.
+        'outer: for value in self.iter() {
+            let key = key_of(value.clone())?;
+            if out.is_empty() {
+                out.push(value.clone());
+                continue;
+            }
+
+            for second in out.iter() {
+                if typst::eval::ops::equal(&key, &key_of(second.clone())?) {
+                    continue 'outer;
+                }
+            }
+
+            out.push(value.clone());
+        }
+
+        Ok(Self(out))
+    }
 }
 
 impl Debug for Array {
diff --git a/crates/typst/src/eval/methods.rs b/crates/typst/src/eval/methods.rs
index a73684268..bf47b7aa5 100644
--- a/crates/typst/src/eval/methods.rs
+++ b/crates/typst/src/eval/methods.rs
@@ -147,6 +147,7 @@ pub fn call(
             "sorted" => array.sorted(vm, span, args.named("key")?)?.into_value(),
             "zip" => array.zip(args.expect("other")?).into_value(),
             "enumerate" => array.enumerate().into_value(),
+            "dedup" => array.dedup(vm, args.named("key")?)?.into_value(),
             _ => return missing(),
         },
 
diff --git a/docs/reference/types.md b/docs/reference/types.md
index 28646cd20..d602bc3e6 100644
--- a/docs/reference/types.md
+++ b/docs/reference/types.md
@@ -966,6 +966,21 @@ Return a new array with the same items, but sorted.
   If given, applies this function to the elements in the array to determine the keys to sort by.
 - returns: array
 
+### dedup()
+Returns a new array with all duplicate items removed.
+
+Only the first element of each duplicate is kept.
+
+```example
+#{
+  (1, 1, 2, 3, 1).dedup() == (1, 2, 3)
+}
+```
+
+- key: function (named)
+  If given, applies this function to the elements in the array to determine the keys to deduplicate by.
+- returns: array
+
 # Dictionary
 A map from string keys to values.
 
diff --git a/tests/typ/compiler/array.typ b/tests/typ/compiler/array.typ
index a96a800f3..4bc027efc 100644
--- a/tests/typ/compiler/array.typ
+++ b/tests/typ/compiler/array.typ
@@ -238,6 +238,21 @@
 #test(((1, 2), 3).zip((4, 5)), (((1, 2), 4), (3, 5)))
 #test((1, "hi").zip((true, false)), ((1, true), ("hi", false)))
 
+---
+// Test the `dedup` method.
+#test(().dedup(), ())
+#test((1,).dedup(), (1,))
+#test((1, 1).dedup(), (1,))
+#test((1, 2, 1).dedup(), (1, 2))
+#test(("Jane", "John", "Eric").dedup(), ("Jane", "John", "Eric"))
+#test(("Jane", "John", "Eric", "John").dedup(), ("Jane", "John", "Eric"))
+
+---
+// Test the `dedup` with the `key` argument.
+#test((1, 2, 3, 4, 5, 6).dedup(key: x => calc.rem(x, 2)), (1, 2))
+#test((1, 2, 3, 4, 5, 6).dedup(key: x => calc.rem(x, 3)), (1, 2, 3))
+#test(("Hello", "World", "Hi", "There").dedup(key: x => x.len()), ("Hello", "Hi"))
+#test(("Hello", "World", "Hi", "There").dedup(key: x => x.at(0)), ("Hello", "World", "There"))
 
 ---
 // Error: 32-37 cannot divide by zero