Allow prime in binding names (#813)

* Allow prime in binding names

* Add note about primes in bindings to tutorial

Author: Omnikar

changelog.md

@@ -22,6 +22,8 @@ This version is not yet released. If you are reading this on the website, then t
 - Modules containing data definition variants now contains a `Variants` binding that lists their names
 - Add array pack syntactic sugar. This lets you write code like `[⊃(+|×)]` as `⊃[+|×]`.
 - Subscripts can now be typed with `,` instead of `__`s
+- Binding names can now end with prime characters `′`, `″`, and `‴`
+  - These will format from `'` at the end of a binding name
 - Add numeric subscripts for [`keep ▽`](https://uiua.org/docs/keep) to keep along a number of dimensions
 - Add [`under ⍜`](https://uiua.org/docs/under) capability to [`fork ⊃`](https://uiua.org/docs/fork)s of monadic functions
   - This allows using [`un °`](https://uiua.org/docs/un)[`by ⊸`](https://uiua.org/docs/by)[`fork ⊃`](https://uiua.org/docs/fork) to set multiple properties at once

parser/src/lex.rs

@@ -1323,9 +1323,15 @@ impl<'a> Lexer<'a> {
                     self.end(Subscr(sub), start)
                 }
                 // Identifiers and unformatted glyphs
-                c if is_custom_glyph(c) || c.chars().all(is_ident_char) || "&!‼".contains(c) => {
+                c if is_custom_glyph(c)
+                    || c.chars().all(is_ident_char)
+                    || "&!‼'′″‴".contains(c) =>
+                {
                     // Get ident start
                     let mut ident = self.ident(start, c).to_string();
+                    while let Some(ch) = self.next_char_if(|ch| "'′″‴".contains(ch)) {
+                        ident.push_str(ch);
+                    }
                     let mut exclam_count = match c {
                         "!" => 1,
                         "‼" => 2,
@@ -1766,7 +1772,7 @@ fn is_formatted_subscript(c: &str) -> bool {
 }
 
 pub(crate) fn canonicalize_ident(ident: &str) -> Ident {
-    canonicalize_subscripts(canonicalize_exclams(ident))
+    canonicalize_subscripts(canonicalize_primes(canonicalize_exclams(ident).as_str()))
 }
 
 /// Rewrite the identifier with the same number of exclamation points
@@ -1790,6 +1796,38 @@ fn place_exclams(ident: &str, count: usize) -> Ident {
     new
 }
 
+/// Rewrite the identifier with the same number of primes
+/// using triple, double, and single prime characters as needed
+fn canonicalize_primes(ident: &str) -> Ident {
+    let mut count = 0;
+    for ch in ident.chars() {
+        count += match ch {
+            '\'' => 1,
+            '′' => 1,
+            '″' => 2,
+            '‴' => 3,
+            _ => 0,
+        };
+    }
+    place_primes(ident, count)
+}
+
+fn place_primes(ident: &str, count: usize) -> Ident {
+    let mut new: Ident = ident.trim_end_matches(['\'', '′', '″', '‴']).into();
+    let num_triple = count / 3;
+    let trailing_num = count % 3;
+    for _ in 0..num_triple {
+        new.push('‴');
+    }
+    match trailing_num {
+        0 => {}
+        1 => new.push('′'),
+        2 => new.push('″'),
+        _ => unreachable!(),
+    }
+    new
+}
+
 /// Rewrite the identifier with numerals preceded by `,` replaced with subscript characters
 fn canonicalize_subscripts(ident: Ident) -> Ident {
     if !ident.contains(',') {

parser/src/parse.rs

@@ -558,7 +558,7 @@ impl Parser<'_> {
                     chars.extend(Primitive::non_deprecated().filter_map(|p| p.glyph()));
                     chars.extend(' '..='~');
                     chars.extend(SUBSCRIPT_DIGITS);
-                    chars.extend("←↚‼₋⌞⌟↓".chars());
+                    chars.extend("←↚‼′″‴₋⌞⌟↓".chars());
                     chars.sort_unstable();
                     chars
                 };
@@ -738,7 +738,12 @@ impl Parser<'_> {
             self.errors
                 .push(name.span.clone().sp(ParseError::AmpersandBindingName));
         }
-        if name.value.trim_end_matches(['!', '‼']).chars().count() >= 2
+        if name
+            .value
+            .trim_end_matches(['!', '‼', '\'', '′', '″', '‴'])
+            .chars()
+            .count()
+            >= 2
             && name.value.chars().next().unwrap().is_ascii_lowercase()
         {
             let captialized: String = name

site/src/tutorial.rs

@@ -657,6 +657,8 @@ fn TutorialBindings() -> impl IntoView {
         <p>"Bindings "<em>"can"</em>" contain subscript numbers. These will format from "<code>","</code>" followed by some digits. Try formatting the example below!"</p>
         <Editor example="X,1 = 5\nSha,256 = \"TODO\""/>
         <p>"Subscripts in binding names are only allowed at the end."</p>
+        <p>"Bindings are allowed to end with "<code>"′"</code>" characters to help indicate related bindings. These format from "<code>"'"</code>". Try formatting the following example."</p>
+        <Editor example="X = [1_2 3_4]\nX' = [2_3 4_5]\nX'' = [2_4 6_8]"/>
         <p><strong>"Bindings are case-sensitive."</strong></p>
         <p>"The parser can sometimes mistake all-lowercase binding names for unformatted built-in functions."</p>
         <p>"Here, the parser thinks that "<code>"part"</code>" is "<Prim prim=Partition/>"."</p>

src/run_prim.rs

@@ -2479,7 +2479,7 @@ mod tests {
 	"repository": {{
         "idents": {{
             "name": "variable.parameter.uiua",
-            "match": "\\b[a-zA-Z]+(₋?[₀₁₂₃₄₅₆₇₈₉]|,`?\\d+)*[!‼]*\\b"
+            "match": "\\b[a-zA-Z]+['′″‴]*(₋?[₀₁₂₃₄₅₆₇₈₉]|,`?\\d+)*[!‼]*\\b"
         }},
 		"comments": {{
 			"name": "comment.line.uiua",