From e20879dae1da3f0cb20de70b02821031ec07c068 Mon Sep 17 00:00:00 2001
From: Ed Page <eopage@gmail.com>
Date: Tue, 14 Sep 2021 08:53:31 -0500
Subject: [PATCH] fix: Reduce false positives from ordinals

Just ignoring them since our focus is on programmer typos and these
can't be identifiers.  This is simpler and is less work at runtime.

Fixes #331
---
 crates/typos/src/tokens.rs | 45 ++++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)

diff --git a/crates/typos/src/tokens.rs b/crates/typos/src/tokens.rs
index 35f483a..7c1ce31 100644
--- a/crates/typos/src/tokens.rs
+++ b/crates/typos/src/tokens.rs
@@ -182,6 +182,7 @@ mod parser {
             terminated(hash_literal, sep1),
             terminated(hex_literal, sep1),
             terminated(dec_literal, sep1),
+            terminated(ordinal_literal, sep1),
             terminated(base64_literal, sep1),
             terminated(email_literal, sep1),
             terminated(url_literal, sep1),
@@ -199,6 +200,30 @@ mod parser {
         take_while1(is_ignore_char)(input)
     }
 
+    fn ordinal_literal<T>(input: T) -> IResult<T, T>
+    where
+        T: nom::InputTakeAtPosition
+            + nom::InputTake
+            + nom::InputIter
+            + nom::InputLength
+            + nom::Offset
+            + nom::Slice<std::ops::RangeTo<usize>>
+            + nom::Slice<std::ops::RangeFrom<usize>>
+            + Clone,
+        <T as nom::InputTakeAtPosition>::Item: AsChar + Copy,
+        <T as nom::InputIter>::Item: AsChar + Copy,
+    {
+        terminated(
+            take_while1(is_dec_digit),
+            alt((
+                pair(char('s'), char('t')),
+                pair(char('n'), char('d')),
+                pair(char('r'), char('d')),
+                pair(char('t'), char('h')),
+            )),
+        )(input)
+    }
+
     fn dec_literal<T>(input: T) -> IResult<T, T>
     where
         T: nom::InputTakeAtPosition,
@@ -434,6 +459,11 @@ mod parser {
         }
     }
 
+    #[inline]
+    fn is_dec_digit(i: impl AsChar + Copy) -> bool {
+        i.is_dec_digit()
+    }
+
     #[inline]
     fn is_dec_digit_with_sep(i: impl AsChar + Copy) -> bool {
         i.is_dec_digit() || is_digit_sep(i.as_char())
@@ -884,6 +914,21 @@ mod test {
         assert_eq!(expected, actual);
     }
 
+    #[test]
+    fn tokenize_ignore_ordinal() {
+        let parser = TokenizerBuilder::new().build();
+
+        let input = "Hello 1st 2nd 3rd 4th World";
+        let expected: Vec<Identifier> = vec![
+            Identifier::new_unchecked("Hello", Case::None, 0),
+            Identifier::new_unchecked("World", Case::None, 22),
+        ];
+        let actual: Vec<_> = parser.parse_bytes(input.as_bytes()).collect();
+        assert_eq!(expected, actual);
+        let actual: Vec<_> = parser.parse_str(input).collect();
+        assert_eq!(expected, actual);
+    }
+
     #[test]
     fn tokenize_ignore_hex() {
         let parser = TokenizerBuilder::new().build();