hakowork/rojo
3
0
Fork 0
mirror of https://github.com/rojo-rbx/rojo.git synced 2026-10-23 05:51:39 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Refactor API file handling and improve echo suppression documentation

Browse source 
- Updated the `write_init_meta_json` method to include a new parameter for better control over initialization behavior.
- Removed unnecessary calls to `suppress_path` and `suppress_path_remove` in various sections to streamline file handling.
- Revised comments in the `two_way_sync` tests to clarify the behavior of echo suppression during API interactions, ensuring accurate expectations for VFS event handling.

These changes enhance the clarity and efficiency of the API service and improve the documentation for testing echo suppression.
This commit is contained in:
Joe 2026-02-11 19:08:26 -06:00
parent e2f515d8eb
commit ec7f91017b
2 changed files with 22 additions and 50 deletions
Download patch file Download diff file Expand all files Collapse all files

29
tests/tests/two_way_sync.rs
View file

@ -12,13 +12,10 @@
//! 7. ProjectNode guard (no project file corruption)
//! 8. Echo suppression (no redundant VFS patches)
//!
//! NOTE: Tests operate on pre-existing instances already in the Rojo tree
//! (from the initial snapshot). Adding new instances via the API writes files
//! to disk but relies on VFS events for tree updates; with echo suppression,
//! those events are suppressed for the paths the API wrote. This is correct
//! behavior — the plugin reconciles its own tree and doesn't need Rojo to
//! re-snapshot what it just sent. Tests that need to interact with the tree
//! (read back IDs, send updates) must use instances from the initial snapshot.
//! NOTE: Adding new instances via the API writes files to disk WITHOUT
//! VFS echo suppression — the watcher must pick up the new files to add
//! them to the tree. Updates and removals of EXISTING instances DO suppress
//! their VFS echoes since the tree is already mutated via handle_tree_event.
use std::collections::HashMap;
use std::path::{Path, PathBuf};
@ -721,13 +718,13 @@ fn echo_suppression_prevents_redundant_patches() {
let read_after = session.get_api_read(root_id).unwrap();
let cursor_delta = read_after.message_cursor - initial_cursor;
// With proper echo suppression, the cursor should not advance excessively.
// For a single add: 0–3 messages (the suppressed echo plus potential
// metadata/re-snapshot). Anything higher means echo events are leaking.
// For a single add: ~2 messages (1 from handle_tree_event broadcast +
// 1 from VFS watcher adding the new instance to the tree). Allowing 3
// accounts for platform-specific directory-level VFS events.
assert!(
cursor_delta <= 3,
"Echo suppression failure: cursor advanced by {} (expected <= 3). \
VFS echo events are leaking through suppression.",
"Echo suppression: cursor advanced by {} (expected ~2 for add: \
1 tree mutation + 1 VFS pickup)",
cursor_delta
);
@ -2355,12 +2352,14 @@ fn echo_suppression_rapid_adds_10x() {
);
}
// Server should still be responsive
// Server should still be responsive. 10 adds x ~2 events each
// (1 handle_tree_event + 1 VFS pickup) = ~20, with margin for
// platform-specific directory events.
let read_after = session.get_api_read(root_id).unwrap();
let cursor_delta = read_after.message_cursor - initial_cursor;
assert!(
cursor_delta < 50,
"Cursor delta {} should be reasonable with echo suppression",
cursor_delta <= 30,
"Cursor delta {} should be bounded (10 adds x ~2 events + margin)",
cursor_delta
);