458 lines
14 KiB
Rust
458 lines
14 KiB
Rust
use std::path::Path;
|
|
use std::process::Stdio;
|
|
|
|
use async_trait::async_trait;
|
|
use serde_json::json;
|
|
use tokio::process::Command;
|
|
use tokio::time::timeout;
|
|
|
|
use crate::tools::traits::{Tool, ToolResult};
|
|
|
|
const MAX_RESULTS: usize = 100;
|
|
const MAX_OUTPUT_CHARS: usize = 50_000;
|
|
const TIMEOUT_SECS: u64 = 60;
|
|
|
|
pub struct ContentSearchTool;
|
|
|
|
impl ContentSearchTool {
|
|
pub fn new() -> Self {
|
|
Self
|
|
}
|
|
|
|
fn resolve_dir(&self, dir: Option<&str>) -> String {
|
|
match dir {
|
|
Some(d) if !d.is_empty() => d.to_string(),
|
|
_ => ".".to_string(),
|
|
}
|
|
}
|
|
|
|
fn truncate_output(&self, lines: &[String]) -> String {
|
|
let mut output = String::new();
|
|
for (i, line) in lines.iter().enumerate() {
|
|
if output.len() + line.len() + 1 > MAX_OUTPUT_CHARS {
|
|
let omitted = lines.len() - i;
|
|
output.push_str(&format!(
|
|
"\n... ({} matches omitted) ...",
|
|
omitted
|
|
));
|
|
break;
|
|
}
|
|
if !output.is_empty() {
|
|
output.push('\n');
|
|
}
|
|
output.push_str(line);
|
|
}
|
|
output
|
|
}
|
|
}
|
|
|
|
impl Default for ContentSearchTool {
|
|
fn default() -> Self {
|
|
Self::new()
|
|
}
|
|
}
|
|
|
|
#[async_trait]
|
|
impl Tool for ContentSearchTool {
|
|
fn name(&self) -> &str {
|
|
"content_search"
|
|
}
|
|
|
|
fn description(&self) -> &str {
|
|
"Search file contents by regex or text pattern. Internally uses ripgrep (rg) for fast searching — if rg is not available, falls back to grep, then pure Rust. Supports context lines, file filtering, and case-sensitivity options."
|
|
}
|
|
|
|
fn parameters_schema(&self) -> serde_json::Value {
|
|
json!({
|
|
"type": "object",
|
|
"properties": {
|
|
"pattern": {
|
|
"type": "string",
|
|
"description": "Regex or text pattern to search for in file contents"
|
|
},
|
|
"dir": {
|
|
"type": "string",
|
|
"description": "Directory to search in (default: current working directory)"
|
|
},
|
|
"file_pattern": {
|
|
"type": "string",
|
|
"description": "Optional glob to restrict which files to search (e.g. '*.rs', '*.{rs,toml}')"
|
|
},
|
|
"case_sensitive": {
|
|
"type": "boolean",
|
|
"description": "Whether to match case-sensitively (default: false)"
|
|
},
|
|
"context_lines": {
|
|
"type": "integer",
|
|
"description": "Number of context lines to show before and after each match (default: 0)"
|
|
},
|
|
"max_results": {
|
|
"type": "integer",
|
|
"description": "Maximum number of matching lines to return (default: 100)"
|
|
}
|
|
},
|
|
"required": ["pattern"]
|
|
})
|
|
}
|
|
|
|
fn read_only(&self) -> bool {
|
|
true
|
|
}
|
|
|
|
async fn execute(&self, args: serde_json::Value) -> anyhow::Result<ToolResult> {
|
|
let pattern = match args.get("pattern").and_then(|v| v.as_str()) {
|
|
Some(p) if !p.is_empty() => p,
|
|
_ => {
|
|
return Ok(ToolResult {
|
|
success: false,
|
|
output: String::new(),
|
|
error: Some("Missing required parameter: pattern".to_string()),
|
|
});
|
|
}
|
|
};
|
|
|
|
let dir = self.resolve_dir(args.get("dir").and_then(|v| v.as_str()));
|
|
let file_pattern = args.get("file_pattern").and_then(|v| v.as_str());
|
|
let case_sensitive = args.get("case_sensitive").and_then(|v| v.as_bool()).unwrap_or(false);
|
|
let context_lines = args.get("context_lines").and_then(|v| v.as_u64()).unwrap_or(0) as usize;
|
|
let max_results = args.get("max_results").and_then(|v| v.as_u64()).unwrap_or(MAX_RESULTS as u64) as usize;
|
|
|
|
let result = self.run_search(pattern, &dir, file_pattern, case_sensitive, context_lines, max_results).await;
|
|
|
|
match result {
|
|
Ok(lines) => {
|
|
let count = lines.len();
|
|
let mut output = self.truncate_output(&lines);
|
|
output.push_str(&format!("\n\n---\n共 {} 条匹配", count));
|
|
Ok(ToolResult { success: true, output, error: None })
|
|
}
|
|
Err(e) => Ok(ToolResult {
|
|
success: false,
|
|
output: String::new(),
|
|
error: Some(e.to_string()),
|
|
}),
|
|
}
|
|
}
|
|
}
|
|
|
|
impl ContentSearchTool {
|
|
async fn run_search(
|
|
&self,
|
|
pattern: &str,
|
|
dir: &str,
|
|
file_pattern: Option<&str>,
|
|
case_sensitive: bool,
|
|
context_lines: usize,
|
|
max_results: usize,
|
|
) -> anyhow::Result<Vec<String>> {
|
|
if which::which("rg").is_ok() {
|
|
match self.search_with_rg(pattern, dir, file_pattern, case_sensitive, context_lines, max_results).await {
|
|
Ok(lines) => return Ok(lines),
|
|
Err(e) => tracing::warn!("rg failed: {}, falling back", e),
|
|
}
|
|
}
|
|
|
|
if which::which("grep").is_ok() {
|
|
match self.search_with_grep(pattern, dir, file_pattern, case_sensitive, context_lines, max_results).await {
|
|
Ok(lines) if !lines.is_empty() => return Ok(lines),
|
|
Ok(_) => {},
|
|
Err(e) => tracing::warn!("grep failed: {}, falling back", e),
|
|
}
|
|
}
|
|
|
|
tracing::warn!("No rg/grep available, using built-in content search (much slower). Install ripgrep for better performance.");
|
|
self.search_with_rust(pattern, dir, file_pattern, case_sensitive, context_lines, max_results).await
|
|
}
|
|
|
|
async fn search_with_rg(
|
|
&self,
|
|
pattern: &str,
|
|
dir: &str,
|
|
file_pattern: Option<&str>,
|
|
case_sensitive: bool,
|
|
context_lines: usize,
|
|
max_results: usize,
|
|
) -> anyhow::Result<Vec<String>> {
|
|
let mut cmd = Command::new("rg");
|
|
cmd.arg("-n")
|
|
.arg("--no-heading")
|
|
.arg("--color").arg("never")
|
|
.arg("--max-count").arg(max_results.to_string())
|
|
.arg(pattern)
|
|
.arg(dir)
|
|
.stdout(Stdio::piped())
|
|
.stderr(Stdio::piped());
|
|
|
|
if !case_sensitive {
|
|
cmd.arg("-i");
|
|
}
|
|
if context_lines > 0 {
|
|
cmd.arg("-C").arg(context_lines.to_string());
|
|
}
|
|
if let Some(fp) = file_pattern {
|
|
cmd.arg("--glob").arg(fp);
|
|
}
|
|
|
|
let output = timeout(
|
|
std::time::Duration::from_secs(TIMEOUT_SECS),
|
|
cmd.output(),
|
|
)
|
|
.await
|
|
.map_err(|_| anyhow::anyhow!("rg timed out after {}s", TIMEOUT_SECS))??;
|
|
|
|
if !output.status.success() && output.status.code() != Some(1) {
|
|
let stderr = String::from_utf8_lossy(&output.stderr);
|
|
return Err(anyhow::anyhow!("rg error: {}", stderr.trim()));
|
|
}
|
|
|
|
let text = String::from_utf8_lossy(&output.stdout);
|
|
let lines: Vec<String> = text.lines()
|
|
.take(max_results)
|
|
.map(|l| l.to_string())
|
|
.collect();
|
|
Ok(lines)
|
|
}
|
|
|
|
async fn search_with_grep(
|
|
&self,
|
|
pattern: &str,
|
|
dir: &str,
|
|
file_pattern: Option<&str>,
|
|
case_sensitive: bool,
|
|
context_lines: usize,
|
|
max_results: usize,
|
|
) -> anyhow::Result<Vec<String>> {
|
|
let mut cmd = Command::new("grep");
|
|
cmd.arg("-rn")
|
|
.arg("-E")
|
|
.arg("--color=never")
|
|
.arg("--binary-files=without-match")
|
|
.arg(pattern)
|
|
.arg(dir)
|
|
.stdout(Stdio::piped())
|
|
.stderr(Stdio::piped());
|
|
|
|
if !case_sensitive {
|
|
cmd.arg("-i");
|
|
}
|
|
if context_lines > 0 {
|
|
cmd.arg("-C").arg(context_lines.to_string());
|
|
}
|
|
if let Some(fp) = file_pattern {
|
|
cmd.arg("--include").arg(fp);
|
|
}
|
|
|
|
let output = timeout(
|
|
std::time::Duration::from_secs(TIMEOUT_SECS),
|
|
cmd.output(),
|
|
)
|
|
.await
|
|
.map_err(|_| anyhow::anyhow!("grep timed out after {}s", TIMEOUT_SECS))??;
|
|
|
|
let text = String::from_utf8_lossy(&output.stdout);
|
|
let lines: Vec<String> = text.lines()
|
|
.take(max_results)
|
|
.map(|l| l.to_string())
|
|
.collect();
|
|
Ok(lines)
|
|
}
|
|
|
|
async fn search_with_rust(
|
|
&self,
|
|
pattern: &str,
|
|
dir: &str,
|
|
file_pattern: Option<&str>,
|
|
case_sensitive: bool,
|
|
_context_lines: usize,
|
|
max_results: usize,
|
|
) -> anyhow::Result<Vec<String>> {
|
|
let re = if case_sensitive {
|
|
regex::Regex::new(pattern)
|
|
} else {
|
|
regex::RegexBuilder::new(pattern)
|
|
.case_insensitive(true)
|
|
.build()
|
|
}
|
|
.map_err(|e| anyhow::anyhow!("Invalid regex pattern '{}': {}", pattern, e))?;
|
|
|
|
let file_re = file_pattern.map(|fp| {
|
|
let re_str = glob_to_regex(fp);
|
|
if case_sensitive {
|
|
regex::Regex::new(&re_str)
|
|
} else {
|
|
regex::RegexBuilder::new(&re_str).case_insensitive(true).build()
|
|
}
|
|
});
|
|
|
|
let file_re = match file_re {
|
|
Some(Ok(r)) => Some(r),
|
|
Some(Err(e)) => return Err(anyhow::anyhow!("Invalid file pattern: {}", e)),
|
|
None => None,
|
|
};
|
|
|
|
let mut results = Vec::new();
|
|
grep_dir(Path::new(dir), Path::new(dir), &re, file_re.as_ref(), &mut results, max_results)?;
|
|
|
|
Ok(results)
|
|
}
|
|
}
|
|
|
|
fn glob_to_regex(glob: &str) -> String {
|
|
let mut regex = String::from("^");
|
|
let chars: Vec<char> = glob.chars().collect();
|
|
let mut i = 0;
|
|
while i < chars.len() {
|
|
match chars[i] {
|
|
'*' => {
|
|
if i + 1 < chars.len() && chars[i + 1] == '*' {
|
|
regex.push_str(".*");
|
|
i += 1;
|
|
} else {
|
|
regex.push_str("[^/]*");
|
|
}
|
|
}
|
|
'?' => regex.push_str("[^/]"),
|
|
'.' | '+' | '(' | ')' | '[' | ']' | '{' | '}' | '^' | '$' | '|' | '\\' => {
|
|
regex.push('\\');
|
|
regex.push(chars[i]);
|
|
}
|
|
c => regex.push(c),
|
|
}
|
|
i += 1;
|
|
}
|
|
regex.push('$');
|
|
regex
|
|
}
|
|
|
|
fn grep_dir(
|
|
base: &Path,
|
|
current: &Path,
|
|
re: ®ex::Regex,
|
|
file_re: Option<®ex::Regex>,
|
|
results: &mut Vec<String>,
|
|
max: usize,
|
|
) -> anyhow::Result<()> {
|
|
if results.len() >= max {
|
|
return Ok(());
|
|
}
|
|
|
|
let entries = match std::fs::read_dir(current) {
|
|
Ok(e) => e,
|
|
Err(_) => return Ok(()),
|
|
};
|
|
|
|
for entry in entries.flatten() {
|
|
let path = entry.path();
|
|
let rel = match path.strip_prefix(base) {
|
|
Ok(r) => r,
|
|
Err(_) => continue,
|
|
};
|
|
|
|
if path.is_dir() {
|
|
if let Some(name) = rel.file_name().and_then(|n| n.to_str())
|
|
&& name.starts_with('.') && name.len() > 1 {
|
|
continue;
|
|
}
|
|
grep_dir(base, &path, re, file_re, results, max)?;
|
|
} else if path.is_file() {
|
|
if let Some(file_re) = file_re
|
|
&& let Some(name) = rel.file_name().and_then(|n| n.to_str())
|
|
&& !file_re.is_match(name) {
|
|
continue;
|
|
}
|
|
|
|
if let Ok(content) = std::fs::read_to_string(&path) {
|
|
for (line_num, line) in content.lines().enumerate() {
|
|
if re.is_match(line) {
|
|
results.push(format!(
|
|
"{}:{}:{}",
|
|
rel.to_string_lossy(),
|
|
line_num + 1,
|
|
line
|
|
));
|
|
if results.len() >= max {
|
|
return Ok(());
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
Ok(())
|
|
}
|
|
|
|
#[cfg(test)]
|
|
mod tests {
|
|
use super::*;
|
|
use std::fs;
|
|
use tempfile::TempDir;
|
|
|
|
#[tokio::test]
|
|
async fn test_content_search_rust_fallback() {
|
|
let dir = TempDir::new().unwrap();
|
|
fs::write(dir.path().join("main.rs"), "fn main() {\n let x = 42;\n println!(\"hello\");\n}").unwrap();
|
|
fs::write(dir.path().join("lib.rs"), "pub fn foo() -> u32 {\n let y = 42;\n y\n}").unwrap();
|
|
fs::write(dir.path().join("README.md"), "# Project\nHello world").unwrap();
|
|
|
|
let tool = ContentSearchTool::new();
|
|
let result = tool
|
|
.execute(json!({
|
|
"pattern": "let.*=.*42",
|
|
"dir": dir.path().to_str().unwrap()
|
|
}))
|
|
.await
|
|
.unwrap();
|
|
|
|
assert!(result.success);
|
|
assert!(result.output.contains("main.rs"));
|
|
assert!(result.output.contains("lib.rs"));
|
|
assert!(!result.output.contains("README.md"));
|
|
assert!(result.output.contains("共 2 条匹配"));
|
|
}
|
|
|
|
#[tokio::test]
|
|
async fn test_content_search_file_filter() {
|
|
let dir = TempDir::new().unwrap();
|
|
fs::write(dir.path().join("main.rs"), "fn main() {}").unwrap();
|
|
fs::write(dir.path().join("config.toml"), "name = \"test\"").unwrap();
|
|
|
|
let tool = ContentSearchTool::new();
|
|
let result = tool
|
|
.execute(json!({
|
|
"pattern": "test",
|
|
"dir": dir.path().to_str().unwrap(),
|
|
"file_pattern": "*.toml"
|
|
}))
|
|
.await
|
|
.unwrap();
|
|
|
|
assert!(result.success);
|
|
assert!(result.output.contains("config.toml"));
|
|
assert!(!result.output.contains("main.rs"));
|
|
}
|
|
|
|
#[tokio::test]
|
|
async fn test_content_search_max_results() {
|
|
let dir = TempDir::new().unwrap();
|
|
let mut content = String::new();
|
|
for i in 0..10 {
|
|
content.push_str(&format!("match line {}\n", i));
|
|
}
|
|
fs::write(dir.path().join("data.txt"), &content).unwrap();
|
|
|
|
let tool = ContentSearchTool::new();
|
|
let result = tool
|
|
.execute(json!({
|
|
"pattern": "match line",
|
|
"dir": dir.path().to_str().unwrap(),
|
|
"max_results": 3
|
|
}))
|
|
.await
|
|
.unwrap();
|
|
|
|
assert!(result.success);
|
|
assert!(result.output.contains("共 3 条匹配"));
|
|
}
|
|
}
|