add feature to disable checking strings for valid utf8

Cargo.toml

@@ -17,4 +17,7 @@ bitstream_reader_derive = { version = "0.4", path = "bitstream_reader_derive" }
 [dev-dependencies]
 maplit = "1.0.1"
 
+[features]
+unchecked_utf8 = []
+
 [workspace]

src/buffer.rs

@@ -8,10 +8,10 @@ use std::rc::Rc;
 
 use num_traits::{Float, PrimInt};
 
+use crate::{ReadError, Result};
 use crate::endianness::Endianness;
 use crate::is_signed::IsSigned;
 use crate::unchecked_primitive::{UncheckedPrimitiveFloat, UncheckedPrimitiveInt};
-use crate::{ReadError, Result};
 
 const USIZE_SIZE: usize = size_of::<usize>();
 
@@ -35,7 +35,7 @@ const USIZE_SIZE: usize = size_of::<usize>();
 /// # }
 /// ```
 pub struct BitBuffer<E>
-where
+    where
         E: Endianness,
 {
     bytes: Rc<Vec<u8>>,
@@ -45,7 +45,7 @@ where
 }
 
 impl<E> BitBuffer<E>
-where
+    where
         E: Endianness,
 {
     /// Create a new BitBuffer from a byte vector
@@ -73,7 +73,7 @@ where
 }
 
 impl<E> BitBuffer<E>
-where
+    where
         E: Endianness,
 {
     /// The available number of bits in the buffer
@@ -349,6 +349,12 @@ where
     ///
     /// You can either read a fixed number of bytes, or a dynamic length null-terminated string
     ///
+    /// # Features
+    ///
+    /// To disable the overhead of checking if the read bytes are valid you can enable the `unchecked_utf8`
+    /// feature of the crate to use `String::from_utf8_unchecked` instead of `String::from_utf8`
+    /// to create the string from the read bytes.
+    ///
     /// # Errors
     ///
     /// - [`ReadError::NotEnoughData`]: not enough bits available in the buffer
@@ -381,19 +387,32 @@ where
     /// [`ReadError::NotEnoughData`]: enum.ReadError.html#variant.NotEnoughData
     /// [`ReadError::Utf8Error`]: enum.ReadError.html#variant.Utf8Error
     pub fn read_string(&self, position: usize, byte_len: Option<usize>) -> Result<String> {
-        let bytes = self.read_string_bytes(position, byte_len)?;
+        match byte_len {
-        let raw_string = String::from_utf8(bytes)?;
+            Some(byte_len) => {
-        if byte_len.is_some() {
+                let bytes = self.read_bytes(position, byte_len)?;
-            Ok(raw_string.trim_end_matches(char::from(0)).to_owned())
+                let raw_string = if cfg!(feature = "unchecked_utf8") {
+                    unsafe {
+                        String::from_utf8_unchecked(bytes)
+                    }
                 } else {
-            Ok(raw_string)
+                    String::from_utf8(bytes)?
+                };
+                Ok(raw_string.trim_end_matches(char::from(0)).to_owned())
+            },
+            None => {
+                let bytes = self.read_string_bytes(position);
+                if cfg!(feature = "unchecked_utf8") {
+                    unsafe {
+                        Ok(String::from_utf8_unchecked(bytes))
+                    }
+                } else {
+                    String::from_utf8(bytes).map_err(ReadError::from)
+                }
+            }
         }
     }
 
-    fn read_string_bytes(&self, position: usize, byte_len: Option<usize>) -> Result<Vec<u8>> {
+    fn read_string_bytes(&self, position: usize) -> Vec<u8> {
-        match byte_len {
-            Some(len) => return self.read_bytes(position, len),
-            None => {
         let mut acc = Vec::with_capacity(25);
         let mut pos = position;
         loop {
@@ -409,15 +428,13 @@ where
                 let byte = if E::is_le() { bytes[i] } else { bytes[1 + i] };
 
                 if byte == 0 {
-                            return Ok(acc);
+                    return acc;
                 }
                 acc.push(byte);
             }
             pos += read;
         }
     }
-        };
-    }
 
     /// Read a sequence of bits from the buffer as float
     ///