diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 32bc82f..ce27de2 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -45,13 +45,6 @@ jobs:
       - name: Run cargo test
         run: cargo test --workspace
 
-      - name: Install nightly toolchain
-        uses: dtolnay/rust-toolchain@nightly
-
-      - name: Run cargo miri test
-        run: cargo miri test --workspace
-
-
   miri:
     name: Miri
     runs-on: ubuntu-latest
diff --git a/src/intern.rs b/src/intern.rs
index d955910..cb1ebf4 100644
--- a/src/intern.rs
+++ b/src/intern.rs
@@ -12,7 +12,7 @@ use hashbrown::raw::RawTable;
 /// of using the token data directly.
 /// This allows for much better performance by amortizing the cost hashing/equality.
 ///
-/// While you can intern tokens yourself it is strongly recommended to use [`InternedInput`](crate::intern::InternedInput) module.
+/// While you can intern tokens yourself it is strongly recommended to use [`InternedInput`] module.
 #[derive(PartialEq, Eq, Hash, Clone, Copy, Debug)]
 #[repr(transparent)]
 pub struct Token(pub u32);
@@ -44,7 +44,7 @@ pub trait TokenSource {
 /// of using the token data directly.
 /// This allows for much better performance by amortizing the cost hashing/equality.
 ///
-/// While you can intern tokens yourself it is strongly recommended to use [`InternedInput`](crate::intern::InternedInput) module.
+/// While you can intern tokens yourself it is strongly recommended to use [`InternedInput`] module.
 #[derive(Default)]
 pub struct InternedInput<T: Eq + Hash> {
     pub before: Vec<Token>,
diff --git a/src/sink.rs b/src/sink.rs
index 950bd59..e561def 100644
--- a/src/sink.rs
+++ b/src/sink.rs
@@ -35,7 +35,7 @@ pub trait Sink: Sized {
     /// to obtain the final diff result
     fn finish(self) -> Self::Out;
 
-    /// Utility method that constructs a [`Counter`](crate::sink::Counter) that tracks the total number
+    /// Utility method that constructs a [`Counter`] that tracks the total number
     /// of inserted and removed tokens in the changes passed to [`process_change`](crate::Sink::process_change).
     fn with_counter(self) -> Counter<Self> {
         Counter::new(self)
@@ -58,7 +58,7 @@ impl Sink for () {
     fn finish(self) -> Self::Out {}
 }
 
-/// A [`Sink`](crate::Sink) which wraps a different sink
+/// A [`Sink`] which wraps a different sink
 /// and counts the number of `removed` and `inserted` [tokens](crate::intern::Token).
 pub struct Counter<T> {
     /// Total number of recorded inserted [`tokens`](crate::intern::Token).
@@ -67,10 +67,10 @@ pub struct Counter<T> {
     /// Total number of recorded inserted [`tokens`](crate::intern::Token).
     /// Computed by summing the lengths of the `after` subsequences pass to [`process_change`](crate::Sink::process_change).
     pub insertions: u32,
-    /// The [`Sink`](crate::Sink) for which the counter records [`tokens`](crate::intern::Token).
+    /// The [`Sink`] for which the counter records [`tokens`](crate::intern::Token).
     /// All calls to [`process_change`](crate::Sink::process_change) are forwarded to the `sink` by the counter.
     /// After [`finish`](crate::Sink::finish) is called, this field contains the output returned by the [`finish`](crate::Sink::finish)
-    /// method of the wrapped [`Sink`](crate::Sink)
+    /// method of the wrapped [`Sink`].
     pub wrapped: T,
 }
 
diff --git a/src/sources.rs b/src/sources.rs
index 1c75bd0..5f5cb02 100644
--- a/src/sources.rs
+++ b/src/sources.rs
@@ -3,7 +3,7 @@ use std::str::from_utf8_unchecked;
 
 use crate::TokenSource;
 
-/// Returns a [`TokenSource`](crate::intern::TokenSource) that uses
+/// Returns a [`TokenSource`] that uses
 /// the lines in `data` as Tokens. The newline seperator (`\r\n` or `\n`) is
 /// not included in the emitted tokens.
 /// This means that changing the newline seperator from `\r\n` to `\n`
@@ -12,7 +12,7 @@ pub fn lines(data: &str) -> Lines<'_, false> {
     Lines(ByteLines(data.as_bytes()))
 }
 
-/// Returns a [`TokenSource`](crate::intern::TokenSource) that uses
+/// Returns a [`TokenSource`] that uses
 /// the lines in `data` as Tokens. The newline seperator (`\r\n` or `\n`) is
 /// included in the emitted tokens.
 /// This means that changing the newline seperator from `\r\n` to `\n`
@@ -21,7 +21,7 @@ pub fn lines_with_terminator(data: &str) -> Lines<'_, true> {
     Lines(ByteLines(data.as_bytes()))
 }
 
-/// Returns a [`TokenSource`](crate::intern::TokenSource) that uses
+/// Returns a [`TokenSource`] that uses
 /// the lines in `data` as Tokens. A lines is a continous subslice of
 /// `data` which does not contain `\n` (or `\r\n`).
 /// The newline seperator (`\r\n` or `\n`) is not included in the emitted tokens.
@@ -31,7 +31,7 @@ pub fn byte_lines_with_terminator(data: &[u8]) -> ByteLines<'_, true> {
     ByteLines(data)
 }
 
-/// Returns a [`TokenSource`](crate::intern::TokenSource) that uses
+/// Returns a [`TokenSource`] that uses
 /// the lines in `data` as Tokens. The newline seperator (`\r\n` or `\n`) is
 /// included in the emitted tokens.
 /// This means that changing the newline seperator from `\r\n` to `\n`
@@ -69,8 +69,8 @@ impl<'a> TokenSource for &'a [u8] {
     }
 }
 
-/// A [`TokenSource`](crate::intern::TokenSource) that returns the lines of a `str` as tokens.
-/// See [`lines`](crate::sources::lines) and [`lines_with_terminator`](crate::sources::lines_with_terminator) for details
+/// A [`TokenSource`] that returns the lines of a `str` as tokens.
+/// See [`lines`] and [`lines_with_terminator`] for details
 #[derive(Clone, Copy, PartialEq, Eq)]
 pub struct Lines<'a, const INCLUDE_LINE_TERMINATOR: bool>(ByteLines<'a, INCLUDE_LINE_TERMINATOR>);
 
@@ -99,8 +99,8 @@ impl<'a, const INCLUDE_LINE_TERMINATOR: bool> TokenSource for Lines<'a, INCLUDE_
     }
 }
 
-/// A [`TokenSource`](crate::intern::TokenSource) that returns the lines of a byte slice as tokens.
-/// See [`byte_lines`](crate::sources::lines) and [`byte_lines_with_terminator`](crate::sources::byte_lines_with_terminator) for details
+/// A [`TokenSource`] that returns the lines of a byte slice as tokens.
+/// See [`byte_lines`] and [`byte_lines_with_terminator`] for details
 #[derive(Clone, Copy, PartialEq, Eq)]
 pub struct ByteLines<'a, const INCLUDE_LINE_TERMINATOR: bool>(&'a [u8]);
 
diff --git a/src/unified_diff.rs b/src/unified_diff.rs
index 8df8d72..50116fe 100644
--- a/src/unified_diff.rs
+++ b/src/unified_diff.rs
@@ -5,7 +5,7 @@ use std::ops::Range;
 use crate::intern::{InternedInput, Interner, Token};
 use crate::Sink;
 
-/// A [`Sink`](crate::sink::Sink) that creates a textual diff
+/// A [`Sink`] that creates a textual diff
 /// in the format typically output by git or gnu-diff if the `-u` option is used
 pub struct UnifiedDiffBuilder<'a, W, T>
 where
@@ -31,7 +31,7 @@ where
     T: Hash + Eq + Display,
 {
     /// Create a new `UnifiedDiffBuilder` for the given `input`,
-    /// that will return a [`String`](std::string::String).
+    /// that will return a [`String`].
     pub fn new(input: &'a InternedInput<T>) -> Self {
         Self {
             before_hunk_start: 0,
@@ -54,7 +54,7 @@ where
     T: Hash + Eq + Display,
 {
     /// Create a new `UnifiedDiffBuilder` for the given `input`,
-    /// that will writes it output to the provided implementation of [`Write`](std::fmt::Write).
+    /// that will writes it output to the provided implementation of [`Write`].
     pub fn with_writer(input: &'a InternedInput<T>, writer: W) -> Self {
         Self {
             before_hunk_start: 0,