further improve text object movement performance

Author: pascalkuthe

helix-core/src/movement.rs

@@ -405,7 +405,13 @@ pub fn goto_treesitter_object(
 
         let cap_name = |t: TextObject| format!("{}.{}", object_name, t);
         let mut cursor = QueryCursor::new();
-        let nodes = lang_config.textobject_query()?.capture_nodes_any(
+        let range = match dir {
+            Direction::Forward => (byte_pos + 1)..slice.len_bytes(),
+            Direction::Backward => 0..byte_pos,
+        };
+        // only search within the relevant range
+        cursor.set_byte_range(range);
+        let mut nodes = lang_config.textobject_query()?.capture_nodes_any(
             &[
                 &cap_name(TextObject::Movement),
                 &cap_name(TextObject::Around),
@@ -417,12 +423,12 @@ pub fn goto_treesitter_object(
         )?;
 
         let node = match dir {
-            Direction::Forward => nodes
-                .filter(|n| n.start_byte() > byte_pos)
-                .min_by_key(|n| n.start_byte())?,
-            Direction::Backward => nodes
-                .filter(|n| n.end_byte() < byte_pos)
-                .max_by_key(|n| n.end_byte())?,
+            // catpure_nodes_any always returns nodes in the correct order
+            // so we don't need to iterate trough all of them which helps performance
+            // tramendously
+            Direction::Forward => nodes.next()?,
+            // captures can not be iterated in reverse so we are stuck with this
+            Direction::Backward => nodes.last()?,
         };
 
         let len = slice.len_bytes();

helix-core/src/syntax.rs

@@ -356,19 +356,22 @@ impl<'a> CapturedNode<'a> {
 
 /// The number of matches a TS cursor can at once to avoid performance problems for medium to large files.
 /// Set with `set_match_limit`.
-/// Using such a limit means that we loose valid captures in so there is fundementally a tradeoff here.
+/// Using such a limit means that we loose valid captures in, so there is fundamentally a tradeoff here.
+///
 ///
 /// Old tree sitter versions used a limit of 32 by default until this limit was removed in version `0.19.5` (must now best set manually).
-/// However this causes performance issues for medium to large files.
-/// In helix this problem caused textobjt motions to take multiple seconds to complete in medium sized rust files (3k loc).
+/// However, this causes performance issues for medium to large files.
+/// In helix, this problem caused treesitter motions to take multiple seconds to complete in medium-sized rust files (3k loc).
 /// Neovim also encountered this problem and reintroduced this limit after it was removed upstream
-/// (see https://github.com/neovim/neovim/issues/14897 and https://github.com/neovim/neovim/pull/14915).
-/// The number used here is fundementally a tradeoff between breaking some obscure edgecases and performance.
+/// (see <https://github.com/neovim/neovim/issues/14897> and <https://github.com/neovim/neovim/pull/14915>).
+/// The number used here is fundamentally a tradeoff between breaking some obscure edge cases and performance.
+///
 ///
-/// 64 was choosen because neovim uses that value.
-/// It has been choosen somewhat arbitrarly (https://github.com/neovim/neovim/pull/18397) mostly based
-/// upon how many issues occur in practice.
-/// So this could be changed if we ran into problems.
+/// A value of 64 was chosen because neovim uses that value.
+/// Neovim chose this value somewhat arbitrarily (<https://github.com/neovim/neovim/pull/18397>) adjusting it whenever issues occur in practice.
+/// However this value has been in use for a long time and due to the large userbase of neovim it is probably a good choice.
+/// If this limit causes problems for a grammar in the future, it could be increased.
+
 const TREE_SITTER_MATCH_LIMIT: u32 = 64;
 
 impl TextObjectQuery {