mirror of
https://github.com/joshuadavidthomas/django-language-server.git
synced 2025-09-04 01:10:42 +00:00
wip
This commit is contained in:
parent
74b6b5b56d
commit
48dacb277c
2 changed files with 290 additions and 32 deletions
|
|
@ -90,24 +90,32 @@ impl TextDocument {
|
|||
#[derive(Clone, Debug)]
|
||||
pub struct LineIndex {
|
||||
pub line_starts: Vec<u32>,
|
||||
pub line_starts_utf16: Vec<u32>,
|
||||
pub length: u32,
|
||||
pub length_utf16: u32,
|
||||
}
|
||||
|
||||
impl LineIndex {
|
||||
pub fn new(text: &str) -> Self {
|
||||
let mut line_starts = vec![0];
|
||||
let mut pos = 0;
|
||||
let mut line_starts_utf16 = vec![0];
|
||||
let mut pos_utf8 = 0;
|
||||
let mut pos_utf16 = 0;
|
||||
|
||||
for c in text.chars() {
|
||||
pos += u32::try_from(c.len_utf8()).unwrap_or(0);
|
||||
pos_utf8 += u32::try_from(c.len_utf8()).unwrap_or(0);
|
||||
pos_utf16 += u32::try_from(c.len_utf16()).unwrap_or(0);
|
||||
if c == '\n' {
|
||||
line_starts.push(pos);
|
||||
line_starts.push(pos_utf8);
|
||||
line_starts_utf16.push(pos_utf16);
|
||||
}
|
||||
}
|
||||
|
||||
Self {
|
||||
line_starts,
|
||||
length: pos,
|
||||
line_starts_utf16,
|
||||
length: pos_utf8,
|
||||
length_utf16: pos_utf16,
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -117,6 +125,38 @@ impl LineIndex {
|
|||
Some(line_start + position.character)
|
||||
}
|
||||
|
||||
/// Convert UTF-16 LSP position to UTF-8 byte offset
|
||||
pub fn offset_utf16(&self, position: Position, text: &str) -> Option<u32> {
|
||||
let line_start_utf8 = self.line_starts.get(position.line as usize)?;
|
||||
let _line_start_utf16 = self.line_starts_utf16.get(position.line as usize)?;
|
||||
|
||||
// If position is at start of line, return UTF-8 line start
|
||||
if position.character == 0 {
|
||||
return Some(*line_start_utf8);
|
||||
}
|
||||
|
||||
// Find the line text
|
||||
let next_line_start = self.line_starts.get(position.line as usize + 1)
|
||||
.copied()
|
||||
.unwrap_or(self.length);
|
||||
|
||||
let line_text = text.get(*line_start_utf8 as usize..next_line_start as usize)?;
|
||||
|
||||
// Convert UTF-16 character offset to UTF-8 byte offset within the line
|
||||
let mut utf16_pos = 0;
|
||||
let mut utf8_pos = 0;
|
||||
|
||||
for c in line_text.chars() {
|
||||
if utf16_pos >= position.character {
|
||||
break;
|
||||
}
|
||||
utf16_pos += u32::try_from(c.len_utf16()).unwrap_or(0);
|
||||
utf8_pos += u32::try_from(c.len_utf8()).unwrap_or(0);
|
||||
}
|
||||
|
||||
Some(line_start_utf8 + utf8_pos)
|
||||
}
|
||||
|
||||
#[allow(dead_code)]
|
||||
pub fn position(&self, offset: u32) -> Position {
|
||||
let line = match self.line_starts.binary_search(&offset) {
|
||||
|
|
|
|||
|
|
@ -12,6 +12,7 @@ use tower_lsp_server::lsp_types::CompletionResponse;
|
|||
use tower_lsp_server::lsp_types::DidChangeTextDocumentParams;
|
||||
use tower_lsp_server::lsp_types::DidCloseTextDocumentParams;
|
||||
use tower_lsp_server::lsp_types::DidOpenTextDocumentParams;
|
||||
use tower_lsp_server::lsp_types::TextDocumentContentChangeEvent;
|
||||
use tower_lsp_server::lsp_types::Documentation;
|
||||
use tower_lsp_server::lsp_types::InsertTextFormat;
|
||||
use tower_lsp_server::lsp_types::MarkupContent;
|
||||
|
|
@ -92,35 +93,14 @@ impl Store {
|
|||
.get_text(file_id)
|
||||
.ok_or_else(|| anyhow!("File content not found: {}", uri_str))?;
|
||||
|
||||
// Apply text changes
|
||||
let mut new_content = current_content.to_string();
|
||||
for change in ¶ms.content_changes {
|
||||
if let Some(range) = change.range {
|
||||
// Get current line index for position calculations
|
||||
let line_index = self
|
||||
.line_indices
|
||||
.get(&file_id)
|
||||
.ok_or_else(|| anyhow!("Line index not found for: {}", uri_str))?;
|
||||
// Get current line index for position calculations
|
||||
let line_index = self
|
||||
.line_indices
|
||||
.get(&file_id)
|
||||
.ok_or_else(|| anyhow!("Line index not found for: {}", uri_str))?;
|
||||
|
||||
if let (Some(start_offset), Some(end_offset)) = (
|
||||
line_index.offset(range.start).map(|o| o as usize),
|
||||
line_index.offset(range.end).map(|o| o as usize),
|
||||
) {
|
||||
let mut updated_content = String::with_capacity(
|
||||
new_content.len() - (end_offset - start_offset) + change.text.len(),
|
||||
);
|
||||
|
||||
updated_content.push_str(&new_content[..start_offset]);
|
||||
updated_content.push_str(&change.text);
|
||||
updated_content.push_str(&new_content[end_offset..]);
|
||||
|
||||
new_content = updated_content;
|
||||
}
|
||||
} else {
|
||||
// Full document update
|
||||
new_content.clone_from(&change.text);
|
||||
}
|
||||
}
|
||||
// Apply text changes using the new function
|
||||
let new_content = apply_text_changes(¤t_content, ¶ms.content_changes, line_index)?;
|
||||
|
||||
// Update TextDocument version
|
||||
if let Some(document) = self.documents.get_mut(&uri_str) {
|
||||
|
|
@ -235,3 +215,241 @@ impl Store {
|
|||
}
|
||||
}
|
||||
}
|
||||
|
||||
/// Apply text changes to content, handling multiple changes correctly
|
||||
fn apply_text_changes(
|
||||
content: &str,
|
||||
changes: &[TextDocumentContentChangeEvent],
|
||||
line_index: &LineIndex,
|
||||
) -> Result<String> {
|
||||
if changes.is_empty() {
|
||||
return Ok(content.to_string());
|
||||
}
|
||||
|
||||
// Check for full document replacement first
|
||||
for change in changes {
|
||||
if change.range.is_none() {
|
||||
return Ok(change.text.clone());
|
||||
}
|
||||
}
|
||||
|
||||
// Sort changes by start position in reverse order (end to start)
|
||||
let mut sorted_changes = changes.to_vec();
|
||||
sorted_changes.sort_by(|a, b| {
|
||||
match (a.range, b.range) {
|
||||
(Some(range_a), Some(range_b)) => {
|
||||
// Primary sort: by line (reverse)
|
||||
let line_cmp = range_b.start.line.cmp(&range_a.start.line);
|
||||
if line_cmp != std::cmp::Ordering::Equal {
|
||||
line_cmp
|
||||
} else {
|
||||
// Secondary sort: by character (reverse)
|
||||
range_b.start.character.cmp(&range_a.start.character)
|
||||
}
|
||||
}
|
||||
_ => std::cmp::Ordering::Equal,
|
||||
}
|
||||
});
|
||||
|
||||
let mut result = content.to_string();
|
||||
|
||||
for change in &sorted_changes {
|
||||
if let Some(range) = change.range {
|
||||
// Convert UTF-16 positions to UTF-8 offsets
|
||||
let start_offset = line_index.offset_utf16(range.start, &result)
|
||||
.ok_or_else(|| anyhow!("Invalid start position: {:?}", range.start))?;
|
||||
let end_offset = line_index.offset_utf16(range.end, &result)
|
||||
.ok_or_else(|| anyhow!("Invalid end position: {:?}", range.end))?;
|
||||
|
||||
if start_offset as usize > result.len() || end_offset as usize > result.len() {
|
||||
return Err(anyhow!("Offset out of bounds: start={}, end={}, len={}",
|
||||
start_offset, end_offset, result.len()));
|
||||
}
|
||||
|
||||
// Apply the change
|
||||
result.replace_range(start_offset as usize..end_offset as usize, &change.text);
|
||||
}
|
||||
}
|
||||
|
||||
Ok(result)
|
||||
}
|
||||
|
||||
#[cfg(test)]
|
||||
mod tests {
|
||||
use super::*;
|
||||
use tower_lsp_server::lsp_types::Range;
|
||||
|
||||
#[test]
|
||||
fn test_apply_single_character_insertion() {
|
||||
let content = "Hello world";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 6), Position::new(0, 6))),
|
||||
range_length: None,
|
||||
text: "beautiful ".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "Hello beautiful world");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_apply_single_character_deletion() {
|
||||
let content = "Hello world";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 5), Position::new(0, 6))),
|
||||
range_length: None,
|
||||
text: "".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "Helloworld");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_apply_multiple_changes_in_reverse_order() {
|
||||
let content = "line 1\nline 2\nline 3";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// Insert "new " at position (1, 0) and "another " at position (0, 0)
|
||||
let changes = vec![
|
||||
TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 0), Position::new(0, 0))),
|
||||
range_length: None,
|
||||
text: "another ".to_string(),
|
||||
},
|
||||
TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(1, 0), Position::new(1, 0))),
|
||||
range_length: None,
|
||||
text: "new ".to_string(),
|
||||
},
|
||||
];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "another line 1\nnew line 2\nline 3");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_apply_multiline_replacement() {
|
||||
let content = "line 1\nline 2\nline 3";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 0), Position::new(2, 6))),
|
||||
range_length: None,
|
||||
text: "completely new content".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "completely new content");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_apply_full_document_replacement() {
|
||||
let content = "old content";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: None,
|
||||
range_length: None,
|
||||
text: "brand new content".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "brand new content");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_utf16_line_index_basic() {
|
||||
let content = "hello world";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// ASCII characters should have 1:1 UTF-8:UTF-16 mapping
|
||||
let pos = Position::new(0, 6);
|
||||
let offset = line_index.offset_utf16(pos, content).unwrap();
|
||||
assert_eq!(offset, 6);
|
||||
assert_eq!(&content[6..7], "w");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_utf16_line_index_with_emoji() {
|
||||
let content = "hello 👋 world";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// 👋 is 2 UTF-16 code units but 4 UTF-8 bytes
|
||||
let pos_after_emoji = Position::new(0, 8); // UTF-16 position after "hello 👋"
|
||||
let offset = line_index.offset_utf16(pos_after_emoji, content).unwrap();
|
||||
|
||||
// Should point to the space before "world"
|
||||
assert_eq!(offset, 10); // UTF-8 byte offset
|
||||
assert_eq!(&content[10..11], " ");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_utf16_line_index_multiline() {
|
||||
let content = "first line\nsecond line";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
let pos = Position::new(1, 7); // Position at 'l' in "line" on second line
|
||||
let offset = line_index.offset_utf16(pos, content).unwrap();
|
||||
assert_eq!(offset, 18); // 11 (first line + \n) + 7
|
||||
assert_eq!(&content[18..19], "l");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_apply_changes_with_emoji() {
|
||||
let content = "hello 👋 world";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// Insert text after the space following the emoji (UTF-16 position 9)
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 9), Position::new(0, 9))),
|
||||
range_length: None,
|
||||
text: "beautiful ".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "hello 👋 beautiful world");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_line_index_utf16_tracking() {
|
||||
let content = "a👋b";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// Check UTF-16 line starts are tracked correctly
|
||||
assert_eq!(line_index.line_starts_utf16, vec![0]);
|
||||
assert_eq!(line_index.length_utf16, 4); // 'a' (1) + 👋 (2) + 'b' (1) = 4 UTF-16 units
|
||||
assert_eq!(line_index.length, 6); // 'a' (1) + 👋 (4) + 'b' (1) = 6 UTF-8 bytes
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_edge_case_changes_at_boundaries() {
|
||||
let content = "abc";
|
||||
let line_index = LineIndex::new(content);
|
||||
|
||||
// Insert at beginning
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 0), Position::new(0, 0))),
|
||||
range_length: None,
|
||||
text: "start".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "startabc");
|
||||
|
||||
// Insert at end
|
||||
let line_index = LineIndex::new(content);
|
||||
let changes = vec![TextDocumentContentChangeEvent {
|
||||
range: Some(Range::new(Position::new(0, 3), Position::new(0, 3))),
|
||||
range_length: None,
|
||||
text: "end".to_string(),
|
||||
}];
|
||||
|
||||
let result = apply_text_changes(content, &changes, &line_index).unwrap();
|
||||
assert_eq!(result, "abcend");
|
||||
}
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue