mirror of
https://github.com/Dicklesworthstone/coding_agent_session_search.git
synced 2026-09-22 11:59:25 +08:00
Dicklesworthstone
7306e692b8
Large-scale restructuring across the codebase (43 files, +3307/-7223 lines — net reduction of ~3900 lines through aggressive cleanup and deduplication): Connector layer: - Refactor connectors/mod.rs: consolidate and simplify connector dispatch, removing ~1650 lines of redundant code paths - Streamline connectors/vibe.rs connector implementation - Add dedicated per-connector integration tests: connector_chatgpt, connector_clawdbot, connector_copilot, connector_cursor, connector_openclaw, connector_vibe, and connector_factory Search pipeline: - Rewrite canonicalize.rs search text normalization (~839 lines refactored) - Refactor embedder registry and reranker registry for cleaner trait-based plugin architecture - Update fastembed embedder and reranker implementations - Improve hash_embedder, query processing, and two_tier_search pipeline - Enhance daemon_client for better search service communication UI and theming: - Overhaul app.rs (~1085 lines refactored) for cleaner component lifecycle and event handling - Redesign theme system: new style_system.rs and theme.rs with comprehensive token-based theming - Update components/theme.rs with new palette definitions - Add keyboard shortcuts module improvements Testing and quality: - Add golden test fixtures (tests/fixtures/golden/) for regression testing - Add serialization compatibility and golden tests for format stability - Update all UI snapshots to match new theme system - Expand bakeoff harness, cross-workstream, and daemon client tests - Add beads migration baseline Infrastructure: - Update Cargo.toml and Cargo.lock for dependency changes - Update beads issue tracking metadata Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
249 lines
7.9 KiB
Rust
249 lines
7.9 KiB
Rust
use coding_agent_search::connectors::clawdbot::ClawdbotConnector;
|
|
use coding_agent_search::connectors::{Connector, ScanContext};
|
|
use std::fs;
|
|
use std::path::Path;
|
|
use tempfile::TempDir;
|
|
|
|
// ============================================================================
|
|
// Helper
|
|
// ============================================================================
|
|
|
|
fn write_session(root: &Path, name: &str, lines: &[&str]) -> std::path::PathBuf {
|
|
let path = root.join(name);
|
|
fs::write(&path, lines.join("\n")).unwrap();
|
|
path
|
|
}
|
|
|
|
// ============================================================================
|
|
// Detection tests
|
|
// ============================================================================
|
|
|
|
#[test]
|
|
fn detect_does_not_panic() {
|
|
let connector = ClawdbotConnector::new();
|
|
let result = connector.detect();
|
|
let _ = result.detected;
|
|
}
|
|
|
|
// ============================================================================
|
|
// Scan — happy path
|
|
// ============================================================================
|
|
|
|
#[test]
|
|
fn scan_parses_basic_conversation() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"session.jsonl",
|
|
&[
|
|
r#"{"role":"user","content":"Hello Clawdbot","timestamp":"2025-06-15T10:00:00.000Z"}"#,
|
|
r#"{"role":"assistant","content":"Hi there!","timestamp":"2025-06-15T10:00:05.000Z"}"#,
|
|
],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs.len(), 1);
|
|
assert_eq!(convs[0].agent_slug, "clawdbot");
|
|
assert_eq!(convs[0].messages.len(), 2);
|
|
assert_eq!(convs[0].messages[0].role, "user");
|
|
assert_eq!(convs[0].messages[0].content, "Hello Clawdbot");
|
|
assert_eq!(convs[0].messages[1].role, "assistant");
|
|
assert!(convs[0].started_at.is_some());
|
|
assert!(convs[0].ended_at.is_some());
|
|
assert_eq!(convs[0].title, Some("Hello Clawdbot".to_string()));
|
|
}
|
|
|
|
#[test]
|
|
fn scan_multiple_sessions() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"a.jsonl",
|
|
&[r#"{"role":"user","content":"Session A","timestamp":"2025-06-15T09:00:00.000Z"}"#],
|
|
);
|
|
write_session(
|
|
&sessions,
|
|
"b.jsonl",
|
|
&[r#"{"role":"user","content":"Session B","timestamp":"2025-06-15T10:00:00.000Z"}"#],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs.len(), 2);
|
|
let contents: Vec<&str> = convs
|
|
.iter()
|
|
.map(|c| c.messages[0].content.as_str())
|
|
.collect();
|
|
assert!(contents.contains(&"Session A"));
|
|
assert!(contents.contains(&"Session B"));
|
|
}
|
|
|
|
// ============================================================================
|
|
// Scan — edge cases
|
|
// ============================================================================
|
|
|
|
#[test]
|
|
fn scan_skips_invalid_json_and_empty_lines() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"mixed.jsonl",
|
|
&[
|
|
"",
|
|
"not-json-at-all",
|
|
r#"{"role":"user","content":"Valid","timestamp":"2025-06-15T10:00:00.000Z"}"#,
|
|
r#"{"role":"assistant","content":"","timestamp":"2025-06-15T10:00:01.000Z"}"#,
|
|
],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs.len(), 1);
|
|
assert_eq!(
|
|
convs[0].messages.len(),
|
|
1,
|
|
"empty content should be skipped"
|
|
);
|
|
assert_eq!(convs[0].messages[0].content, "Valid");
|
|
}
|
|
|
|
#[test]
|
|
fn scan_empty_directory_returns_empty() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert!(convs.is_empty());
|
|
}
|
|
|
|
#[test]
|
|
fn scan_handles_malformed_json() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"bad.jsonl",
|
|
&[
|
|
r#"{"broken json ..."#,
|
|
r#"{"role":"user","content":"After malformed","timestamp":"2025-06-15T10:00:00.000Z"}"#,
|
|
],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs.len(), 1);
|
|
assert_eq!(convs[0].messages[0].content, "After malformed");
|
|
}
|
|
|
|
#[test]
|
|
fn scan_preserves_message_ordering() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"ordered.jsonl",
|
|
&[
|
|
r#"{"role":"user","content":"First","timestamp":"2025-06-15T10:00:00.000Z"}"#,
|
|
r#"{"role":"assistant","content":"Second","timestamp":"2025-06-15T10:00:01.000Z"}"#,
|
|
r#"{"role":"user","content":"Third","timestamp":"2025-06-15T10:00:02.000Z"}"#,
|
|
],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs[0].messages[0].idx, 0);
|
|
assert_eq!(convs[0].messages[0].content, "First");
|
|
assert_eq!(convs[0].messages[1].idx, 1);
|
|
assert_eq!(convs[0].messages[1].content, "Second");
|
|
assert_eq!(convs[0].messages[2].idx, 2);
|
|
assert_eq!(convs[0].messages[2].content, "Third");
|
|
}
|
|
|
|
// ============================================================================
|
|
// Incremental scanning (since_ts)
|
|
// ============================================================================
|
|
|
|
#[test]
|
|
fn scan_incremental_since_ts() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"recent.jsonl",
|
|
&[r#"{"role":"user","content":"Recent","timestamp":"2025-06-15T10:00:00.000Z"}"#],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
|
|
// Far-future since_ts should filter out everything.
|
|
let far_future = chrono::Utc::now().timestamp_millis() + 86_400_000;
|
|
let ctx = ScanContext::local_default(sessions.clone(), Some(far_future));
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert!(
|
|
convs.is_empty(),
|
|
"far-future since_ts should filter out old files"
|
|
);
|
|
}
|
|
|
|
// ============================================================================
|
|
// External ID extraction
|
|
// ============================================================================
|
|
|
|
#[test]
|
|
fn scan_extracts_external_id_from_filename() {
|
|
let tmp = TempDir::new().unwrap();
|
|
let sessions = tmp.path().join(".clawdbot/sessions");
|
|
fs::create_dir_all(&sessions).unwrap();
|
|
|
|
write_session(
|
|
&sessions,
|
|
"my-session-42.jsonl",
|
|
&[r#"{"role":"user","content":"Test","timestamp":"2025-06-15T10:00:00.000Z"}"#],
|
|
);
|
|
|
|
let connector = ClawdbotConnector::new();
|
|
let ctx = ScanContext::local_default(sessions.clone(), None);
|
|
let convs = connector.scan(&ctx).unwrap();
|
|
|
|
assert_eq!(convs.len(), 1);
|
|
assert!(
|
|
convs[0]
|
|
.external_id
|
|
.as_ref()
|
|
.unwrap()
|
|
.contains("my-session-42"),
|
|
"external_id should be derived from filename stem"
|
|
);
|
|
}
|