From c1e14347eca8139e7fee45094ff5328c9892f057 Mon Sep 17 00:00:00 2001
From: Yaroslav Bolyukin <iam@lach.pw>
Date: Fri, 22 Apr 2022 19:51:09 +0000
Subject: [PATCH] fix: decode non-bmp characters


---
--- a/crates/jrsonnet-parser/src/unescape.rs
+++ b/crates/jrsonnet-parser/src/unescape.rs
@@ -1,3 +1,11 @@
+use std::str::Chars;
+
+fn decode_unicode(chars: &mut Chars) -> Option<u16> {
+	IntoIterator::into_iter([chars.next()?, chars.next()?, chars.next()?, chars.next()?])
+		.map(|c| c.to_digit(16).map(|f| f as u16))
+		.try_fold(0u16, |acc, v| Some((acc << 4) | (v?)))
+}
+
 pub fn unescape(s: &str) -> Option<String> {
 	let mut chars = s.chars();
 	let mut out = String::with_capacity(s.len());
@@ -14,17 +22,26 @@
 			'n' => out.push('\n'),
 			'r' => out.push('\r'),
 			't' => out.push('\t'),
-			'u' => {
-				let c = IntoIterator::into_iter([
-					chars.next()?,
-					chars.next()?,
-					chars.next()?,
-					chars.next()?,
-				])
-				.map(|c| c.to_digit(16))
-				.try_fold(0u32, |acc, v| Some((acc << 8) | (v?)))?;
-				out.push(char::from_u32(c)?)
-			}
+			'u' => match decode_unicode(&mut chars)? {
+				// May only be second byte
+				0xDC00..=0xDFFF => return None,
+				// Surrogate pair
+				n1 @ 0xD800..=0xDBFF => {
+					if chars.next() != Some('\\') {
+						return None;
+					}
+					if chars.next() != Some('u') {
+						return None;
+					}
+					let n2 = decode_unicode(&mut chars)?;
+					if !matches!(n2, 0xDC00..=0xDFFF) {
+						return None;
+					}
+					let n = (((n1 - 0xD800) as u32) << 10 | (n2 - 0xDC00) as u32) + 0x1_0000;
+					out.push(char::from_u32(n)?);
+				}
+				n => out.push(char::from_u32(n as u32)?),
+			},
 			'x' => {
 				let c = IntoIterator::into_iter([chars.next()?, chars.next()?])
 					.map(|c| c.to_digit(16))

-- 
gitstuff