Fix more panics

Cargo.toml

@@ -30,13 +30,13 @@ path-absolutize = { version = "3.0.14" }
 proc-macro2 = { version = "1.0.51" }
 quote = { version = "1.0.23" }
 regex = { version = "1.7.1" }
-ruff_text_size = { git = "https://github.com/charliermarsh/RustPython.git", rev = "834a2c356bbb31c33af900d9d02ce0228a0f6145" }
+ruff_text_size = { git = "https://github.com/charliermarsh/RustPython.git", rev = "50b8570bc262098b84dac86092133d3ee26caf23" }
 rustc-hash = { version = "1.1.0" }
-rustpython-common = { git = "https://github.com/charliermarsh/RustPython.git", rev = "834a2c356bbb31c33af900d9d02ce0228a0f6145" }
+rustpython-common = { git = "https://github.com/charliermarsh/RustPython.git", rev = "50b8570bc262098b84dac86092133d3ee26caf23" }
 rustpython-parser = { features = [
   "lalrpop",
   "serde",
-], git = "https://github.com/charliermarsh/RustPython.git", rev = "834a2c356bbb31c33af900d9d02ce0228a0f6145" }
+], git = "https://github.com/charliermarsh/RustPython.git", rev = "50b8570bc262098b84dac86092133d3ee26caf23" }
 schemars = { version = "0.8.12" }
 serde = { version = "1.0.152", features = ["derive"] }
 serde_json = { version = "1.0.93", features = ["preserve_order"] }

crates/ruff/src/jupyter/notebook.rs

@@ -18,7 +18,7 @@ pub const JUPYTER_NOTEBOOK_EXT: &str = "ipynb";
 /// Jupyter Notebook indexing table
 ///
 /// When we lint a jupyter notebook, we have to translate the row/column based on
-/// [`crate::message::Location`]
+/// [`ruff_text_size::TextSize`]
 /// to jupyter notebook cell/row/column.
 #[derive(Debug, Eq, PartialEq)]
 pub struct JupyterIndex {

crates/ruff/src/rules/pycodestyle/rules/doc_line_too_long.rs

@@ -58,7 +58,7 @@ pub(crate) fn doc_line_too_long(line: &Line, settings: &Settings) -> Option<Diag
         let mut start = line.end();
         let mut width = line_width;
 
-        for c in line.chars() {
+        for c in line.chars().rev() {
             width -= c.width().unwrap_or(0);
             start -= c.text_len();
             if width == limit {

crates/ruff/src/rules/pycodestyle/rules/invalid_escape_sequence.rs

@@ -4,7 +4,7 @@ use ruff_text_size::{TextLen, TextRange, TextSize};
 
 use ruff_diagnostics::{AlwaysAutofixableViolation, Diagnostic, Edit};
 use ruff_macros::{derive_message_formats, violation};
-use ruff_python_ast::newlines::StrExt;
+use ruff_python_ast::newlines::UniversalNewlineIterator;
 use ruff_python_ast::source_code::Locator;
 
 /// ## What it does
@@ -73,10 +73,13 @@ pub fn invalid_escape_sequence(
     };
     let quote_pos = text.find(quote).unwrap();
     let prefix = text[..quote_pos].to_lowercase();
-    let body = &text[(quote_pos + quote.len())..(text.len() - quote.len())];
+    let body = &text[quote_pos + quote.len()..text.len() - quote.len()];
 
     if !prefix.contains('r') {
-        for line in body.universal_newlines() {
+        for line in UniversalNewlineIterator::with_offset(
+            &body,
+            start + TextSize::try_from(quote_pos).unwrap() + quote.text_len(),
+        ) {
             let mut chars_iter = line.char_indices().peekable();
 
             while let Some((i, c)) = chars_iter.next() {
@@ -99,11 +102,8 @@ pub fn invalid_escape_sequence(
                     continue;
                 }
 
-                let location = start
-                    + line.start()
-                    + quote.text_len()
-                    + TextSize::try_from(quote_pos + i).unwrap();
-                let range = TextRange::at(location, TextSize::from(2));
+                let location = line.start() + TextSize::try_from(i).unwrap();
+                let range = TextRange::at(location, next_char.text_len() + TextSize::from(1));
                 let mut diagnostic = Diagnostic::new(InvalidEscapeSequence(*next_char), range);
                 if autofix {
                     diagnostic.set_fix(Edit::insertion(

crates/ruff/src/rules/pycodestyle/rules/line_too_long.rs

@@ -51,7 +51,7 @@ pub(crate) fn line_too_long(line: &Line, settings: &Settings) -> Option<Diagnost
         let mut offset = line.end();
         let mut width = line_width;
 
-        for c in line.chars() {
+        for c in line.chars().rev() {
             width -= c.width().unwrap_or(0);
             offset -= c.text_len();
 

crates/ruff/src/rules/pycodestyle/rules/logical_lines/mod.rs

@@ -261,7 +261,7 @@ impl<'a> LogicalLine<'a> {
         }
     }
 
-    /// Returns the [`Location`] of the first token on the line or [`None`].
+    /// Returns the [`TextSize`] of the first token on the line or [`None`].
     pub fn first_token_location(&self) -> Option<TextSize> {
         self.tokens().first().map(|t| t.start())
     }

crates/ruff/src/rules/pycodestyle/snapshots/ruff__rules__pycodestyle__tests__E501_E501.py.snap

@@ -10,11 +10,11 @@ E501.py:5:89: E501 Line too long (123 > 88 characters)
 8 | """
   |
 
-E501.py:16:86: E501 Line too long (95 > 88 characters)
+E501.py:16:85: E501 Line too long (95 > 88 characters)
    |
 16 | _ = "---------------------------------------------------------------------------AAAAAAA"
 17 | _ = "---------------------------------------------------------------------------亜亜亜亜亜亜亜"
-   |                                                                                           ^^^^^ E501
+   |                                                                                         ^^^^^^^ E501
    |
 
 E501.py:25:89: E501 Line too long (127 > 88 characters)

crates/ruff/src/rules/pyflakes/rules/unused_variable.rs

@@ -61,7 +61,7 @@ impl AlwaysAutofixableViolation for UnusedVariable {
     }
 }
 
-/// Return the start and end [`Location`] of the token after the next match of
+/// Return the [`TextRange`] of the token after the next match of
 /// the predicate, skipping over any bracketed expressions.
 fn match_token_after<F, T>(located: &Located<T>, locator: &Locator, f: F) -> TextRange
 where
@@ -124,7 +124,7 @@ where
     unreachable!("No token after matched");
 }
 
-/// Return the start and end [`Location`] of the token matching the predicate,
+/// Return the [`TextRange`] of the token matching the predicate,
 /// skipping over any bracketed expressions.
 fn match_token<F, T>(located: &Located<T>, locator: &Locator, f: F) -> TextRange
 where
@@ -190,7 +190,7 @@ enum DeletionKind {
 }
 
 /// Generate a [`Edit`] to remove an unused variable assignment, given the
-/// enclosing [`Stmt`] and the [`Range`] of the variable binding.
+/// enclosing [`Stmt`] and the [`TextRange`] of the variable binding.
 fn remove_unused_variable(
     stmt: &Stmt,
     range: TextRange,

crates/ruff/src/rules/pyupgrade/fixes.rs

@@ -12,7 +12,7 @@ use ruff_python_ast::source_code::{Locator, Stylist};
 
 use crate::cst::matchers::match_module;
 
-/// Safely adjust the indentation of the indented block at [`Range`].
+/// Safely adjust the indentation of the indented block at [`TextRange`].
 pub fn adjust_indentation(
     range: TextRange,
     indentation: &str,

crates/ruff_cli/src/cache.rs

@@ -108,7 +108,7 @@ impl Serialize for SerializeMessage<'_> {
             noqa_offset: noqa_row,
         } = self.message;
 
-        let mut s = serializer.serialize_struct("Message", 6)?;
+        let mut s = serializer.serialize_struct("Message", 5)?;
 
         s.serialize_field("kind", &kind)?;
         s.serialize_field("range", &range)?;
@@ -222,11 +222,9 @@ pub fn get(
             let source_files: Vec<_> = sources
                 .into_iter()
                 .map(|(filename, text)| {
-                    let mut builder = SourceFileBuilder::from_string(filename);
-
-                    builder.set_source_text_string(text);
-
-                    builder.finish()
+                    SourceFileBuilder::from_string(filename)
+                        .source_text_string(text)
+                        .finish()
                 })
                 .collect();
 

crates/ruff_cli/src/diagnostics.rs

@@ -200,6 +200,7 @@ pub fn lint_path(
     let imports = imports.unwrap_or_default();
 
     if let Some(err) = parse_error {
+        // FIXME micha manually print parse errors to get line and column numbers
         // Notify the user of any parse errors.
         error!(
             "{}{}{} {err}",

crates/ruff_diagnostics/Cargo.toml

@@ -10,8 +10,5 @@ rust-version = { workspace = true }
 [dependencies]
 anyhow = { workspace = true }
 log = { workspace = true }
-ruff_python_ast = { path = "../ruff_python_ast" }
 ruff_text_size = { workspace = true }
-
-rustpython-parser = { workspace = true }
 serde = { workspace = true, optional = true, default_features = false, features = [] }

crates/ruff_diagnostics/src/edit.rs

@@ -46,7 +46,7 @@ impl Edit {
         }
     }
 
-    /// Creates an edit that inserts `content` at the [`Location`] `at`.
+    /// Creates an edit that inserts `content` at the [`TextSize`] `at`.
     pub fn insertion(content: String, at: TextSize) -> Self {
         debug_assert!(!content.is_empty(), "Insert content is empty");
 

crates/ruff_python_ast/src/helpers.rs

@@ -626,7 +626,7 @@ pub fn has_comments<T>(located: &Located<T>, locator: &Locator) -> bool {
     has_comments_in(TextRange::new(start, end), locator)
 }
 
-/// Returns `true` if a [`Range`] includes at least one comment.
+/// Returns `true` if a [`TextRange`] includes at least one comment.
 pub fn has_comments_in(range: TextRange, locator: &Locator) -> bool {
     let source = &locator.contents()[range];
 
@@ -887,7 +887,7 @@ pub fn trailing_lines_end(stmt: &Stmt, locator: &Locator) -> TextSize {
         .map_or(line_end, |l| l.full_end())
 }
 
-/// Return the range of the first parenthesis pair after a given [`Location`].
+/// Return the range of the first parenthesis pair after a given [`TextSize`].
 pub fn match_parens(start: TextSize, locator: &Locator) -> Option<TextRange> {
     let contents = &locator.contents()[usize::from(start)..];
 

crates/ruff_python_ast/src/source_code/line_index.rs

@@ -8,7 +8,7 @@ use std::num::NonZeroUsize;
 use std::ops::Deref;
 use std::sync::Arc;
 
-/// Index for fast [`Location`] to [byte offset](TextSize) conversions.
+/// Index for fast [byte offset](TextSize) to [`SourceLocation`] conversions.
 ///
 /// Cloning a [`LineIndex`] is cheap because it only requires bumping a reference count.
 #[derive(Clone)]
@@ -105,7 +105,7 @@ impl LineIndex {
                 } else {
                     // Don't count the BOM character as a column.
                     if line_start == TextSize::from(0) && content.starts_with('\u{feff}') {
-                        line_start += '\u{feff}'.text_len();
+                        line_start = '\u{feff}'.text_len();
                     }
 
                     content[TextRange::new(line_start, offset)].chars().count()

crates/ruff_python_ast/src/source_code/locator.rs

@@ -375,23 +375,22 @@ impl<'a> Locator<'a> {
         &self.contents[self.full_lines_range(range)]
     }
 
-    // TODO remove
-    /// Take the source code up to the given [`Location`].
+    /// Take the source code up to the given [`TextSize`].
     #[inline]
     pub fn up_to(&self, offset: TextSize) -> &'a str {
         &self.contents[TextRange::up_to(offset)]
     }
 
-    /// Take the source code after the given [`Location`].
+    /// Take the source code after the given [`TextSize`].
     #[inline]
     pub fn after(&self, offset: TextSize) -> &'a str {
         &self.contents[usize::from(offset)..]
     }
 
-    /// Take the source code between the given [`Range`].
+    /// Take the source code between the given [`TextRange`].
     #[inline]
-    pub fn slice<R: Into<TextRange>>(&self, range: R) -> &'a str {
-        &self.contents[range.into()]
+    pub fn slice(&self, range: TextRange) -> &'a str {
+        &self.contents[range]
     }
 
     /// Return the underlying source code.

crates/ruff_python_ast/src/source_code/mod.rs

@@ -29,7 +29,7 @@ pub fn round_trip(code: &str, source_path: &str) -> Result<String, ParseError> {
     Ok(generator.generate())
 }
 
-/// Gives access to the source code of a file and allows mapping between [`Location`] and byte offsets.
+/// Gives access to the source code of a file and allows mapping between [`TextSize`] and [`SourceLocation`].
 #[derive(Debug)]
 pub struct SourceCode<'src, 'index> {
     text: &'src str,
@@ -45,28 +45,31 @@ impl<'src, 'index> SourceCode<'src, 'index> {
     }
 
     /// Computes the one indexed row and column numbers for `offset`.
+    #[inline]
     pub fn source_location(&self, offset: TextSize) -> SourceLocation {
         self.index.source_location(offset, self.text)
     }
 
+    #[inline]
     pub fn line_index(&self, offset: TextSize) -> OneIndexed {
         self.index.line_index(offset)
     }
 
-    // TODO inline
-    /// Take the source code up to the given [`Location`].
+    /// Take the source code up to the given [`TextSize`].
+    #[inline]
     pub fn up_to(&self, offset: TextSize) -> &'src str {
         &self.text[TextRange::up_to(offset)]
     }
 
-    /// Take the source code after the given [`Location`].
+    /// Take the source code after the given [`TextSize`].
+    #[inline]
     pub fn after(&self, offset: TextSize) -> &'src str {
         &self.text[usize::from(offset)..]
     }
 
-    /// Take the source code between the given [`Range`].
-    pub fn slice<R: Into<TextRange>>(&self, range: R) -> &'src str {
-        &self.text[range.into()]
+    /// Take the source code between the given [`TextRange`].
+    pub fn slice(&self, range: TextRange) -> &'src str {
+        &self.text[range]
     }
 
     pub fn line_start(&self, line: OneIndexed) -> TextSize {

crates/ruff_wasm/tests/api.rs

@@ -5,6 +5,7 @@ use js_sys;
 use wasm_bindgen_test::*;
 
 use ruff::registry::Rule;
+use ruff_python_ast::source_code::{OneIndexed, SourceLocation};
 use ruff_wasm::*;
 
 macro_rules! check {