single and double precision float parsers

the f64 parser was implemented during the GAT migration, but not the f32
one

src/number/mod.rs

@@ -1322,7 +1322,7 @@ where
   ))
 }
 
-///todo
+/// float number text parser that also recognizes "nan", "infinity" and "inf" (case insensitive)
 pub fn recognize_float_or_exceptions<T, E: ParseError<T>>() -> impl Parser<T, Output = T, Error = E>
 where
   T: Clone + Offset,
@@ -1346,7 +1346,22 @@ where
   ))
 }
 
-/// TODO
+/// single precision floating point number parser from text
+pub fn float<T, E: ParseError<T>>() -> impl Parser<T, Output = f32, Error = E>
+where
+  T: Clone + Offset,
+  T: Input + crate::traits::ParseTo<f32> + Compare<&'static str>,
+  <T as Input>::Item: AsChar + Clone,
+  T: AsBytes,
+  T: for<'a> Compare<&'a [u8]>,
+{
+  Float {
+    o: PhantomData,
+    e: PhantomData,
+  }
+}
+
+/// double precision floating point number parser from text
 pub fn double<T, E: ParseError<T>>() -> impl Parser<T, Output = f64, Error = E>
 where
   T: Clone + Offset,
@@ -1355,23 +1370,27 @@ where
   T: AsBytes,
   T: for<'a> Compare<&'a [u8]>,
 {
-  Double { e: PhantomData }
+  Float {
+    o: PhantomData,
+    e: PhantomData,
+  }
 }
 
-/// TODO
-struct Double<E> {
+/// f64 parser from text
+struct Float<O, E> {
+  o: PhantomData<O>,
   e: PhantomData<E>,
 }
 
-impl<I, E: ParseError<I>> Parser<I> for Double<E>
+impl<I, O, E: ParseError<I>> Parser<I> for Float<O, E>
 where
   I: Clone + Offset,
-  I: Input + crate::traits::ParseTo<f64> + Compare<&'static str>,
+  I: Input + crate::traits::ParseTo<O> + Compare<&'static str>,
   <I as Input>::Item: AsChar + Clone,
   I: AsBytes,
   I: for<'a> Compare<&'a [u8]>,
 {
-  type Output = f64;
+  type Output = O;
   type Error = E;
 
   fn process<OM: crate::OutputMode>(