From 0b098ced2f243b5fd7297422fb5e5bc28db5a475 Mon Sep 17 00:00:00 2001
From: snsmac <snsmac@users.noreply.github.com>
Date: Mon, 1 Aug 2022 20:34:32 +0200
Subject: [PATCH] Syntax: Make \p{Sc} work

'sc' refers to the 'Currency_Symbol' general category,
 but is also the abbreviation for the 'Script' property.

 Fixes #835
 Related #719 b1489c8
---
 regex-syntax/src/unicode.rs | 68 +++++++++++++++++++------------------
 tests/unicode.rs            |  1 +
 2 files changed, 36 insertions(+), 33 deletions(-)

diff --git a/regex-syntax/src/unicode.rs b/regex-syntax/src/unicode.rs
index 70d5954b7..040532289 100644
--- a/regex-syntax/src/unicode.rs
+++ b/regex-syntax/src/unicode.rs
@@ -80,11 +80,11 @@ impl fmt::Display for UnicodeWordError {
 /// This returns an error if the Unicode case folding tables are not available.
 pub fn simple_fold(
     c: char,
-) -> FoldResult<result::Result<impl Iterator<Item = char>, Option<char>>> {
+) -> FoldResult<result::Result<impl Iterator<Item=char>, Option<char>>> {
     #[cfg(not(feature = "unicode-case"))]
     fn imp(
         _: char,
-    ) -> FoldResult<result::Result<impl Iterator<Item = char>, Option<char>>>
+    ) -> FoldResult<result::Result<impl Iterator<Item=char>, Option<char>>>
     {
         use std::option::IntoIter;
         Err::<result::Result<IntoIter<char>, _>, _>(CaseFoldError(()))
@@ -93,7 +93,7 @@ pub fn simple_fold(
     #[cfg(feature = "unicode-case")]
     fn imp(
         c: char,
-    ) -> FoldResult<result::Result<impl Iterator<Item = char>, Option<char>>>
+    ) -> FoldResult<result::Result<impl Iterator<Item=char>, Option<char>>>
     {
         use crate::unicode_tables::case_folding_simple::CASE_FOLDING_SIMPLE;
 
@@ -220,7 +220,7 @@ impl<'a> ClassQuery<'a> {
                         let canon_val =
                             match canonical_value(vals, &property_value) {
                                 None => {
-                                    return Err(Error::PropertyValueNotFound)
+                                    return Err(Error::PropertyValueNotFound);
                                 }
                                 Some(canon_val) => canon_val,
                             };
@@ -243,7 +243,9 @@ impl<'a> ClassQuery<'a> {
         // a general category. (Currently, we don't even support the
         // 'Case_Folding' property. But if we do in the future, users will be
         // required to spell it out.)
-        if norm != "cf" {
+        // 'sc' refers to the 'Currency_Symbol' general category, but is also
+        // the abbreviation for the 'Script' property.
+        if norm != "cf" && norm != "sc" {
             if let Some(canon) = canonical_prop(&norm)? {
                 return Ok(CanonicalClassQuery::Binary(canon));
             }
@@ -462,24 +464,24 @@ fn canonical_script(normalized_value: &str) -> Result<Option<&'static str>> {
 /// If the property names data is not available, then an error is returned.
 fn canonical_prop(normalized_name: &str) -> Result<Option<&'static str>> {
     #[cfg(not(any(
-        feature = "unicode-age",
-        feature = "unicode-bool",
-        feature = "unicode-gencat",
-        feature = "unicode-perl",
-        feature = "unicode-script",
-        feature = "unicode-segment",
+    feature = "unicode-age",
+    feature = "unicode-bool",
+    feature = "unicode-gencat",
+    feature = "unicode-perl",
+    feature = "unicode-script",
+    feature = "unicode-segment",
     )))]
     fn imp(_: &str) -> Result<Option<&'static str>> {
         Err(Error::PropertyNotFound)
     }
 
     #[cfg(any(
-        feature = "unicode-age",
-        feature = "unicode-bool",
-        feature = "unicode-gencat",
-        feature = "unicode-perl",
-        feature = "unicode-script",
-        feature = "unicode-segment",
+    feature = "unicode-age",
+    feature = "unicode-bool",
+    feature = "unicode-gencat",
+    feature = "unicode-perl",
+    feature = "unicode-script",
+    feature = "unicode-segment",
     ))]
     fn imp(name: &str) -> Result<Option<&'static str>> {
         use crate::unicode_tables::property_names::PROPERTY_NAMES;
@@ -519,24 +521,24 @@ fn property_values(
     canonical_property_name: &'static str,
 ) -> Result<Option<PropertyValues>> {
     #[cfg(not(any(
-        feature = "unicode-age",
-        feature = "unicode-bool",
-        feature = "unicode-gencat",
-        feature = "unicode-perl",
-        feature = "unicode-script",
-        feature = "unicode-segment",
+    feature = "unicode-age",
+    feature = "unicode-bool",
+    feature = "unicode-gencat",
+    feature = "unicode-perl",
+    feature = "unicode-script",
+    feature = "unicode-segment",
     )))]
     fn imp(_: &'static str) -> Result<Option<PropertyValues>> {
         Err(Error::PropertyValueNotFound)
     }
 
     #[cfg(any(
-        feature = "unicode-age",
-        feature = "unicode-bool",
-        feature = "unicode-gencat",
-        feature = "unicode-perl",
-        feature = "unicode-script",
-        feature = "unicode-segment",
+    feature = "unicode-age",
+    feature = "unicode-bool",
+    feature = "unicode-gencat",
+    feature = "unicode-perl",
+    feature = "unicode-script",
+    feature = "unicode-segment",
     ))]
     fn imp(name: &'static str) -> Result<Option<PropertyValues>> {
         use crate::unicode_tables::property_values::PROPERTY_VALUES;
@@ -569,15 +571,15 @@ fn property_set(
 ///
 /// If the given age value isn't valid or if the data isn't available, then an
 /// error is returned instead.
-fn ages(canonical_age: &str) -> Result<impl Iterator<Item = Range>> {
+fn ages(canonical_age: &str) -> Result<impl Iterator<Item=Range>> {
     #[cfg(not(feature = "unicode-age"))]
-    fn imp(_: &str) -> Result<impl Iterator<Item = Range>> {
+    fn imp(_: &str) -> Result<impl Iterator<Item=Range>> {
         use std::option::IntoIter;
         Err::<IntoIter<Range>, _>(Error::PropertyNotFound)
     }
 
     #[cfg(feature = "unicode-age")]
-    fn imp(canonical_age: &str) -> Result<impl Iterator<Item = Range>> {
+    fn imp(canonical_age: &str) -> Result<impl Iterator<Item=Range>> {
         use crate::unicode_tables::age;
 
         const AGES: &[(&str, Range)] = &[
@@ -878,7 +880,7 @@ mod tests {
     };
 
     #[cfg(feature = "unicode-case")]
-    fn simple_fold_ok(c: char) -> impl Iterator<Item = char> {
+    fn simple_fold_ok(c: char) -> impl Iterator<Item=char> {
         simple_fold(c).unwrap().unwrap()
     }
 
diff --git a/tests/unicode.rs b/tests/unicode.rs
index 9b3228624..748bbb79c 100644
--- a/tests/unicode.rs
+++ b/tests/unicode.rs
@@ -77,6 +77,7 @@ mat!(uni_class_gencat_format, r"\p{Format}", "\u{E007F}", Some((0, 4)));
 // See: https://github.com/rust-lang/regex/issues/719
 mat!(uni_class_gencat_format_abbrev1, r"\p{cf}", "\u{E007F}", Some((0, 4)));
 mat!(uni_class_gencat_format_abbrev2, r"\p{gc=cf}", "\u{E007F}", Some((0, 4)));
+mat!(uni_class_gencat_format_abbrev3, r"\p{Sc}", "$", Some((0, 1)));
 mat!(
     uni_class_gencat_initial_punctuation,
     r"\p{Initial_Punctuation}",