From 8bb252e941ac7a1777221ae6ddfc597fafd103dc Mon Sep 17 00:00:00 2001
From: "J. Hinchliffe" <jjh413@student.bham.ac.uk>
Date: Sat, 28 Jun 2025 23:35:55 +0100
Subject: [PATCH] tokeniser: return TokeniserErrors where relevant.

The UnexpectedEndOfInput case is a little vague.
---
 assembler/src/source/tokeniser.rs       | 50 +++++++++++++++++--------
 assembler/src/source/tokeniser/error.rs | 18 ++++++---
 2 files changed, 47 insertions(+), 21 deletions(-)

diff --git a/assembler/src/source/tokeniser.rs b/assembler/src/source/tokeniser.rs
index 49cb4b5..261562d 100644
--- a/assembler/src/source/tokeniser.rs
+++ b/assembler/src/source/tokeniser.rs
@@ -9,7 +9,7 @@ use common::prelude::*;
 
 use crate::{
     context::AssemblerContext,
-    error::{AssembleError, AssembleErrorKind, IoError, IoErrorKind},
+    error::{AssembleError, AssembleErrorKind},
     model::module::Module,
     source::{
         lines::{LineSpan, lines_with_spans},
@@ -18,6 +18,7 @@ use crate::{
         source_info::SourceInfo,
         token::{Token, TokenType},
         token_info::{DirectiveToken, LabelToken, RegisterToken, SymbolToken},
+        tokeniser::error::TokeniserError,
     },
 };
 
@@ -161,11 +162,11 @@ impl Tokeniser {
         while !remaining.is_empty() {
             let start_column = column;
 
-            // Try to match a token
+            // Try to match a token.
             let (token_type, consumed) =
                 self.match_token(remaining, line_span.line_number, column)?;
 
-            // Filter out string continuation tokens and comments
+            // Filter out string continuation tokens and comments.
             match token_type {
                 TokenType::StringContinuation => {
                     // Don't add to token stream, just consume input
@@ -186,11 +187,11 @@ impl Tokeniser {
                 }
             }
 
-            // Advance position
+            // Advance position.
             remaining = &remaining[consumed..];
             column += consumed;
 
-            // Skip whitespace
+            // Skip whitespace.
             let before_trim = remaining.len();
             remaining = remaining.trim_start();
             column += before_trim - remaining.len();
@@ -389,26 +390,43 @@ impl Tokeniser {
             return Ok(m);
         }
 
+        let mut idx_iter = 0..;
+
         // Handle miscellaneous characters.
-        match input.chars().enumerate().next() {
-            Some((_, ',')) => Ok((TokenType::Comma, 1)),
-            Some((idx, c)) => {
+        match input.chars().next() {
+            Some(',') => {
+                _ = idx_iter.next();
+
+                Ok((TokenType::Comma, 1))
+            }
+            Some(c) => {
+                let Some(idx) = idx_iter.next() else {
+                    unreachable!()
+                };
+
                 let source =
                     SourceInfo::new(line_number, self.module.clone(), idx..idx + 1);
 
                 Err(AssembleError::new_source_error(
                     source,
-                    AssembleErrorKind::Tokeniser(error::TokeniserError::UnexpectedChar(
-                        c,
+                    AssembleErrorKind::Tokeniser(TokeniserError::UnexpectedChar(c)),
+                ))
+            }
+            None => {
+                let Some(idx) = idx_iter.next() else {
+                    unreachable!()
+                };
+
+                let source =
+                    SourceInfo::new(line_number, self.module.clone(), idx..idx + 1);
+
+                Err(AssembleError::new_source_error(
+                    source,
+                    AssembleErrorKind::Tokeniser(TokeniserError::UnexpectedEndOfInput(
+                        input.len(),
                     )),
                 ))
             }
-            None => Err(AssembleError::new_other_error(AssembleErrorKind::Io(
-                IoError::new(
-                    IoErrorKind::InvalidData,
-                    Some("Unexpected end of input".to_string()),
-                ),
-            ))),
         }
     }
 }
diff --git a/assembler/src/source/tokeniser/error.rs b/assembler/src/source/tokeniser/error.rs
index ae8eddf..ed3f683 100644
--- a/assembler/src/source/tokeniser/error.rs
+++ b/assembler/src/source/tokeniser/error.rs
@@ -13,19 +13,27 @@ pub enum TokeniserError {
     InvalidNumber(&'static str),
     /// An unrecognized token was encountered.
     UnrecognisedToken,
+    /// Returned if the consumed count was lower than the length of the input file.
+    /// This is a sign you will need to debug some [`Tokeniser`] code to ensure that
+    /// [`Tokeniser::match_token`] is working as intended.
+    ///
+    /// First field is length of the line.
+    UnexpectedEndOfInput(usize),
 }
 
 impl TokeniserError {}
 
 impl std::fmt::Display for TokeniserError {
+    #[rustfmt::skip]
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         match self {
             Self::UnexpectedChar(c) => write!(f, "unexpected char '{c}' found in input")?,
-            Self::InvalidNumber(lit) => {
-                write!(f, "invalid integer literal \"{lit}\" found in input")?;
-            }
-            &Self::UnrecognisedToken => write!(f, "unrecognised token found in input")?,
-            &Self::UnterminatedString => write!(f, "unterminated string literal")?,
+            Self::InvalidNumber(lit) => write!(f, "invalid integer literal \"{lit}\" found in input")?,
+            Self::UnrecognisedToken => write!(f, "unrecognised token found in input")?,
+            Self::UnterminatedString => write!(f, "unterminated string literal")?,
+            Self::UnexpectedEndOfInput(line_length) => write!(
+                f, "unexpected end of input, input length: {line_length}"
+            )?,
         }
 
         Ok(())