feat: 添加客户端 ID 过滤机制，避免接收自己发送的消息 #1
Conversation
- 为每个客户端生成唯一 ID - ClipboardMessage 添加 client_id 字段 - 客户端接收消息时检查并跳过自己发送的消息 - 统一客户端剪贴板管理，使用单一 ClipboardMonitor 实例 - 优化网络传输，减少不必要的处理 这解决了设备复制内容后收到自己消息的问题
WalkthroughThis pull request introduces per-client identification in the clipboard synchronization protocol. A unique client_id is generated at startup and embedded in ClipboardMessage structures. The server and clients now distinguish messages by originator, and clients filter out self-originated messages to prevent feedback loops. Changes
Sequence Diagram(s)sequenceDiagram
participant Client1
participant Server
participant Client2
Note over Client1: Startup<br/>Generate unique client_id
Client1->>Server: Connect with client_id
rect rgb(200, 220, 255)
Note over Client1, Server: Scenario A: Client1 sends clipboard update
Client1->>Server: ClipboardMessage {<br/>content, timestamp,<br/>client_id: Some("client_1")<br/>}
Server->>Server: Broadcast to all clients
Server->>Client1: ClipboardMessage {<br/>client_id: Some("client_1")<br/>}
Note over Client1: Filter: Ignore<br/>(self-originated)
Server->>Client2: ClipboardMessage {<br/>client_id: Some("client_1")<br/>}
Note over Client2: Accept & apply<br/>(from other client)
end
rect rgb(220, 255, 220)
Note over Server: Scenario B: Server-local clipboard change
Server->>Server: Local clipboard update
Server->>Client1: ClipboardMessage {<br/>client_id: None<br/>}
Note over Client1: Accept<br/>(server-originated)
Server->>Client2: ClipboardMessage {<br/>client_id: None<br/>}
Note over Client2: Accept<br/>(server-originated)
end
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes
Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (3)
src/main.rs (2)
172-180: Client ID generation may not be unique across fast restarts or clock adjustments.Using
as_micros()fromSystemTime::now()could produce collisions if:
- Two clients start within the same microsecond
- System clock is adjusted backward
- Client restarts very quickly
Consider using a UUID or random component for stronger uniqueness guarantees.
- let client_id = format!( - "{}-{}", - std::env::consts::OS, - std::time::SystemTime::now() - .duration_since(std::time::UNIX_EPOCH) - .unwrap() - .as_micros() - ); + let client_id = format!( + "{}-{}-{}", + std::env::consts::OS, + std::process::id(), + std::time::SystemTime::now() + .duration_since(std::time::UNIX_EPOCH) + .unwrap() + .as_micros() + );Alternatively, consider adding the
uuidcrate for guaranteed uniqueness.
222-237: Clipboard monitoring uses polling instead of callback pattern.The monitoring spawns a separate task that sends a unit signal every 500ms, which then triggers
get_clipboard_content(). This is less efficient than the callback-basedmonitor()used on the server side (line 104). Consider using the same pattern for consistency.However, the current approach does provide unified control in the
select!loop, which is necessary for coordinating local changes with server updates.src/modules/sync.rs (1)
222-225:unwrap()onduration_sincecould panic if system time is before UNIX epoch.While extremely unlikely on modern systems, consider using
unwrap_or(Duration::ZERO)or similar defensive handling.timestamp: std::time::SystemTime::now() .duration_since(std::time::UNIX_EPOCH) - .unwrap() + .unwrap_or_default() .as_secs(),
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (2)
src/main.rs(7 hunks)src/modules/sync.rs(8 hunks)
🧰 Additional context used
🧬 Code graph analysis (2)
src/main.rs (2)
src/modules/sync.rs (2)
new(84-94)new(178-180)src/modules/clipboard.rs (3)
new(25-58)get_clipboard_content(86-141)set_clipboard_content(222-269)
src/modules/sync.rs (1)
src/main.rs (2)
broadcast(62-62)broadcast(184-184)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Cursor Bugbot
🔇 Additional comments (13)
src/main.rs (6)
75-88: Relay mode correctly preserves client_id for filtering.The relay mode now broadcasts the full
ClipboardMessage(includingclient_id), enabling downstream clients to filter out their own messages. This aligns well with the PR objective.
113-119: Server-originated messages correctly marked withclient_id: None.This ensures server clipboard changes are delivered to all clients without being filtered.
264-268: Self-message filtering logic is correct.The comparison
message.client_id.as_ref() == Some(&client_id_for_clipboard)properly handles theOption<String>and correctly skips messages originating from this client.
292-293: Monitor handle abort on loop exit is good cleanup.Properly aborting the spawned polling task prevents resource leaks when the main loop exits.
139-157: Server receive path correctly updates and broadcasts clipboard content.The server properly accesses
message.contentwhen setting the clipboard. Client-sent content updates the server clipboard throughset_clipboard_content(), which then triggers the clipboard monitor to broadcast the change to other clients. This design correctly establishes the server as the clipboard authority and ensures consistency across all connected clients.
284-287: No race condition exists in this code.Both
get_clipboard_content()andset_clipboard_content()execute within the same async task in atokio::select!block (lines 240-300). Theselect!macro ensures that only one branch executes at a time, providing mutual exclusion on the sharedclipboardreference. Whenset_clipboard_content()is called from the server message branch, it completes fully—including thelast_hashupdate on line 267 of clipboard.rs—before the next 500ms polling cycle can callget_clipboard_content(). The last_hash will always be synchronized with the clipboard state before any subsequent poll reads it.src/modules/sync.rs (7)
20-26: ClipboardMessage structure with#[serde(default)]for backward compatibility.Using
#[serde(default)]onclient_idallows deserialization of messages from older clients that don't include this field—good for rolling upgrades.
77-81: Type change fromClipboardContenttoClipboardMessageis correct.This change propagates the full message (including
client_id) through the server's internal channel, enabling proper filtering downstream.
124-129: Server forwards fullClipboardMessagepreserving client_id.The server now sends the complete message to its internal channel, which is essential for the relay mode to preserve origin information.
143-159: Server broadcasts to all connected clients including the sender.The broadcast task sends messages to all clients without filtering. This means the originating client will receive its own message back. The filtering correctly happens client-side (in main.rs lines 266-268). This is a valid design choice—server stays simple, clients filter.
177-180: SyncClient constructor correctly stores client_id.The client_id is stored and later used when constructing outbound messages.
182-186: Type mismatch betweentxandrxparameters.
txsendsClipboardMessagebutrxreceivesClipboardContent. This is intentional—the client receives raw content from clipboard monitoring and wraps it in aClipboardMessagewith itsclient_idbefore sending. The naming could be clearer, but the logic is sound.
218-227: Outbound messages correctly include client_id.The message construction properly attaches the client's ID, enabling other clients to filter appropriately.
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
This PR is being reviewed by Cursor Bugbot
Details
You are on the Bugbot Free tier. On this plan, Bugbot will review limited PRs each billing cycle.
To receive Bugbot reviews on all of your PRs, visit the Cursor dashboard to activate Pro and start your 14-day free trial.
| } | ||
| if let Err(e) = clipboard.set_clipboard_content(&content) { | ||
| eprintln!("Failed to set clipboard: {}", e); | ||
| else => break, |
There was a problem hiding this comment.
Bug: Unreachable else branch due to kept channel sender
The else => break branch in the tokio::select! loop is unreachable. The original local_tx sender created at line 222 remains alive in the same async block as the loop, so local_rx.recv() will never return None. Since the else branch requires ALL futures' patterns to not match simultaneously, and local_rx will always either return Some(()) (from the spawned monitor) or block (never None), this branch can never execute. The clone at line 227 should be removed and the original local_tx should be moved directly into the spawned task instead.
这解决了设备复制内容后收到自己消息的问题
Description
Linked Issues
Additional context
Note
Add client IDs to clipboard messages, skip self-originated updates, and consolidate client clipboard monitoring/receiving.
client_id: Option<String>toClipboardMessage(with serde default).ClipboardMessageinstead ofClipboardContent.src/main.rs,src/modules/sync.rs)ClipboardMessageto preserveclient_id.client_id: None.client_idand pass toSyncClient.client_idin outgoing messages; skip processing messages wheremessage.client_idmatches own ID.tokio::select!(poll local changes and handle server messages), removing the separate receive task.ClipboardMessageend-to-end.Written by Cursor Bugbot for commit 24dbe58. This will update automatically on new commits. Configure here.
Summary by CodeRabbit
Release Notes
New Features
Improvements
✏️ Tip: You can customize this high-level summary in your review settings.