Add RegexSearchTool (#26555)

Release Notes:

- N/A

Cargo.lock

@@ -675,6 +675,7 @@ dependencies = [
  "collections",
  "futures 0.3.31",
  "gpui",
+ "language",
  "language_model",
  "project",
  "rand 0.8.5",

crates/assistant_tools/Cargo.toml

@@ -18,6 +18,7 @@ chrono.workspace = true
 collections.workspace = true
 futures.workspace = true
 gpui.workspace = true
+language.workspace = true
 language_model.workspace = true
 project.workspace = true
 schemars.workspace = true
@@ -29,4 +30,5 @@ util.workspace = true
 rand.workspace = true
 collections = { workspace = true, features = ["test-support"] }
 gpui = { workspace = true, features = ["test-support"] }
+language = { workspace = true, features = ["test-support"] }
 project = { workspace = true, features = ["test-support"] }

crates/assistant_tools/src/assistant_tools.rs

@@ -2,6 +2,7 @@ mod edit_files_tool;
 mod list_directory_tool;
 mod now_tool;
 mod read_file_tool;
+mod regex_search;
 
 use assistant_tool::ToolRegistry;
 use gpui::App;
@@ -10,6 +11,7 @@ use crate::edit_files_tool::EditFilesTool;
 use crate::list_directory_tool::ListDirectoryTool;
 use crate::now_tool::NowTool;
 use crate::read_file_tool::ReadFileTool;
+use crate::regex_search::RegexSearchTool;
 
 pub fn init(cx: &mut App) {
     assistant_tool::init(cx);
@@ -19,4 +21,5 @@ pub fn init(cx: &mut App) {
     registry.register_tool(ReadFileTool);
     registry.register_tool(ListDirectoryTool);
     registry.register_tool(EditFilesTool);
+    registry.register_tool(RegexSearchTool);
 }

crates/assistant_tools/src/regex_search.rs

@@ -0,0 +1,119 @@
+use anyhow::{anyhow, Result};
+use assistant_tool::Tool;
+use futures::StreamExt;
+use gpui::{App, Entity, Task};
+use language::OffsetRangeExt;
+use language_model::LanguageModelRequestMessage;
+use project::{search::SearchQuery, Project};
+use schemars::JsonSchema;
+use serde::{Deserialize, Serialize};
+use std::{cmp, fmt::Write, sync::Arc};
+use util::paths::PathMatcher;
+
+#[derive(Debug, Serialize, Deserialize, JsonSchema)]
+pub struct RegexSearchToolInput {
+    /// A regex pattern to search for in the entire project. Note that the regex
+    /// will be parsed by the Rust `regex` crate.
+    pub regex: String,
+}
+
+pub struct RegexSearchTool;
+
+impl Tool for RegexSearchTool {
+    fn name(&self) -> String {
+        "regex-search".into()
+    }
+
+    fn description(&self) -> String {
+        include_str!("./regex_search_tool/description.md").into()
+    }
+
+    fn input_schema(&self) -> serde_json::Value {
+        let schema = schemars::schema_for!(RegexSearchToolInput);
+        serde_json::to_value(&schema).unwrap()
+    }
+
+    fn run(
+        self: Arc<Self>,
+        input: serde_json::Value,
+        _messages: &[LanguageModelRequestMessage],
+        project: Entity<Project>,
+        cx: &mut App,
+    ) -> Task<Result<String>> {
+        const CONTEXT_LINES: u32 = 2;
+
+        let input = match serde_json::from_value::<RegexSearchToolInput>(input) {
+            Ok(input) => input,
+            Err(err) => return Task::ready(Err(anyhow!(err))),
+        };
+
+        let query = match SearchQuery::regex(
+            &input.regex,
+            false,
+            false,
+            false,
+            PathMatcher::default(),
+            PathMatcher::default(),
+            None,
+        ) {
+            Ok(query) => query,
+            Err(error) => return Task::ready(Err(error)),
+        };
+
+        let results = project.update(cx, |project, cx| project.search(query, cx));
+        cx.spawn(|cx| async move {
+            futures::pin_mut!(results);
+
+            let mut output = String::new();
+            while let Some(project::search::SearchResult::Buffer { buffer, ranges }) =
+                results.next().await
+            {
+                if ranges.is_empty() {
+                    continue;
+                }
+
+                buffer.read_with(&cx, |buffer, cx| {
+                    if let Some(path) = buffer.file().map(|file| file.full_path(cx)) {
+                        writeln!(output, "### Found matches in {}:\n", path.display()).unwrap();
+                        let mut ranges = ranges
+                            .into_iter()
+                            .map(|range| {
+                                let mut point_range = range.to_point(buffer);
+                                point_range.start.row =
+                                    point_range.start.row.saturating_sub(CONTEXT_LINES);
+                                point_range.start.column = 0;
+                                point_range.end.row = cmp::min(
+                                    buffer.max_point().row,
+                                    point_range.end.row + CONTEXT_LINES,
+                                );
+                                point_range.end.column = buffer.line_len(point_range.end.row);
+                                point_range
+                            })
+                            .peekable();
+
+                        while let Some(mut range) = ranges.next() {
+                            while let Some(next_range) = ranges.peek() {
+                                if range.end.row >= next_range.start.row {
+                                    range.end = next_range.end;
+                                    ranges.next();
+                                } else {
+                                    break;
+                                }
+                            }
+
+                            writeln!(output, "```").unwrap();
+                            output.extend(buffer.text_for_range(range));
+                            writeln!(output, "\n```\n").unwrap();
+                        }
+                    }
+                })?;
+            }
+
+            if output.is_empty() {
+                Ok("No matches found".into())
+            } else {
+                Ok(output)
+            }
+        })
+    }
+}

crates/assistant_tools/src/regex_search_tool/description.md

@@ -0,0 +1,3 @@
+Searches the entire project for the given regular expression.
+
+Returns a list of paths that matched the query. For each path, it returns a list of excerpts of the matched text.