Merge pull request #2998 from eth-p/improve-preprocessor

Update bat's tab expansion preprocessor to use bat's ANSI escape sequence iterator.

CHANGELOG.md

@@ -38,6 +38,7 @@
 - Display which theme is the default one in colored output, see #2838 (@sblondon)
 - Add aarch64-apple-darwin ("Apple Silicon") binary tarballs to releases, see #2967 (@someposer)
 - Update the Lisp syntax, see #2970 (@ccqpein)
+- Use bat's ANSI iterator during tab expansion, see #2998 (@eth-p)
 
 ## Syntaxes
 

src/preprocessor.rs

@@ -1,17 +1,18 @@
 use std::fmt::Write;
 
-use console::AnsiCodeIterator;
-
-use crate::nonprintable_notation::NonprintableNotation;
+use crate::{
+    nonprintable_notation::NonprintableNotation,
+    vscreen::{EscapeSequenceOffsets, EscapeSequenceOffsetsIterator},
+};
 
 /// Expand tabs like an ANSI-enabled expand(1).
 pub fn expand_tabs(line: &str, width: usize, cursor: &mut usize) -> String {
     let mut buffer = String::with_capacity(line.len() * 2);
 
-    for chunk in AnsiCodeIterator::new(line) {
-        match chunk {
-            (text, true) => buffer.push_str(text),
-            (mut text, false) => {
+    for seq in EscapeSequenceOffsetsIterator::new(line) {
+        match seq {
+            EscapeSequenceOffsets::Text { .. } => {
+                let mut text = &line[seq.index_of_start()..seq.index_past_end()];
                 while let Some(index) = text.find('\t') {
                     // Add previous text.
                     if index > 0 {
@@ -31,6 +32,10 @@ pub fn expand_tabs(line: &str, width: usize, cursor: &mut usize) -> String {
                 *cursor += text.len();
                 buffer.push_str(text);
             }
+            _ => {
+                // Copy the ANSI escape sequence.
+                buffer.push_str(&line[seq.index_of_start()..seq.index_past_end()])
+            }
         }
     }
 

src/vscreen.rs

@@ -285,7 +285,7 @@ fn join(
 
 /// A range of indices for a raw ANSI escape sequence.
 #[derive(Debug, PartialEq)]
-enum EscapeSequenceOffsets {
+pub enum EscapeSequenceOffsets {
     Text {
         start: usize,
         end: usize,
@@ -320,14 +320,40 @@ enum EscapeSequenceOffsets {
     },
 }
 
+impl EscapeSequenceOffsets {
+    /// Returns the byte-index of the first character in the escape sequence.
+    pub fn index_of_start(&self) -> usize {
+        use EscapeSequenceOffsets::*;
+        match self {
+            Text { start, .. } => *start,
+            Unknown { start, .. } => *start,
+            NF { start_sequence, .. } => *start_sequence,
+            OSC { start_sequence, .. } => *start_sequence,
+            CSI { start_sequence, .. } => *start_sequence,
+        }
+    }
+
+    /// Returns the byte-index past the last character in the escape sequence.
+    pub fn index_past_end(&self) -> usize {
+        use EscapeSequenceOffsets::*;
+        match self {
+            Text { end, .. } => *end,
+            Unknown { end, .. } => *end,
+            NF { end, .. } => *end,
+            OSC { end, .. } => *end,
+            CSI { end, .. } => *end,
+        }
+    }
+}
+
 /// An iterator over the offests of ANSI/VT escape sequences within a string.
 ///
 /// ## Example
 ///
 /// ```ignore
 /// let iter = EscapeSequenceOffsetsIterator::new("\x1B[33mThis is yellow text.\x1B[m");
 /// ```
-struct EscapeSequenceOffsetsIterator<'a> {
+pub struct EscapeSequenceOffsetsIterator<'a> {
     text: &'a str,
     chars: Peekable<CharIndices<'a>>,
 }