mirror of
https://github.com/QwenLM/qwen-code.git
synced 2026-05-13 15:32:19 +00:00
fix(attribution): dedup snapshot writes, cap excludedGenerated, doc commit toggle scope
rsf- (Copilot): recordAttributionSnapshot wrote a full snapshot to the JSONL on every non-retry turn, even when the tracked state was unchanged. Long-running sessions accumulated thousands of identical snapshot copies, inflating session size and slowing /resume hydrate. Dedup by JSON-equality with the prior write — first write always goes through, identical successors are no-ops. rsgo (Copilot): excludedGenerated path list was unbounded. A commit churning thousands of generated artifacts (large dist/ rebuild) could push the JSON note past MAX_NOTE_BYTES (30KB) and lose attribution for the real source files in the same commit. Cap the serialized sample at MAX_EXCLUDED_GENERATED_SAMPLE (50) and add excludedGeneratedCount for the true total. rsg9 + rshM (Copilot): the gitCoAuthor.commit description claimed the toggle only controlled the Co-authored-by trailer, but attachCommitAttribution also gates the per-file git-notes payload on the same flag. Update both the schema description and the settings.md table to mention both effects so disabling the option isn't a silent surprise.
This commit is contained in:
wenshao
parent
090758c5b1
commit
3c0e3293be
7 changed files with 101 additions and 17 deletions
|
|
@ -83,17 +83,17 @@ Settings are organized into categories. Most settings should be placed within th
|
|||
|
||||
#### general
|
||||
|
||||
| Setting | Type | Description | Default |
|
||||
| ------------------------------------------ | ------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------- |
|
||||
| `general.preferredEditor` | string | The preferred editor to open files in. | `undefined` |
|
||||
| `general.vimMode` | boolean | Enable Vim keybindings. | `false` |
|
||||
| `general.enableAutoUpdate` | boolean | Enable automatic update checks and installations on startup. | `true` |
|
||||
| `general.showSessionRecap` | boolean | Auto-show a one-line "where you left off" recap when returning to the terminal after being away. Off by default. Use `/recap` to trigger manually regardless of this setting. | `false` |
|
||||
| `general.sessionRecapAwayThresholdMinutes` | number | Minutes the terminal must be blurred before an auto-recap fires on focus-in. Only used when `showSessionRecap` is enabled. | `5` |
|
||||
| `general.gitCoAuthor.commit` | boolean | Automatically add a Co-authored-by trailer to git commit messages when commits are made through Qwen Code. | `true` |
|
||||
| `general.gitCoAuthor.pr` | boolean | Append a Qwen Code attribution line to pull request descriptions when running `gh pr create`. | `true` |
|
||||
| `general.checkpointing.enabled` | boolean | Enable session checkpointing for recovery. | `false` |
|
||||
| `general.defaultFileEncoding` | string | Default encoding for new files. Use `"utf-8"` (default) for UTF-8 without BOM, or `"utf-8-bom"` for UTF-8 with BOM. Only change this if your project specifically requires BOM. | `"utf-8"` |
|
||||
| Setting | Type | Description | Default |
|
||||
| ------------------------------------------ | ------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------- |
|
||||
| `general.preferredEditor` | string | The preferred editor to open files in. | `undefined` |
|
||||
| `general.vimMode` | boolean | Enable Vim keybindings. | `false` |
|
||||
| `general.enableAutoUpdate` | boolean | Enable automatic update checks and installations on startup. | `true` |
|
||||
| `general.showSessionRecap` | boolean | Auto-show a one-line "where you left off" recap when returning to the terminal after being away. Off by default. Use `/recap` to trigger manually regardless of this setting. | `false` |
|
||||
| `general.sessionRecapAwayThresholdMinutes` | number | Minutes the terminal must be blurred before an auto-recap fires on focus-in. Only used when `showSessionRecap` is enabled. | `5` |
|
||||
| `general.gitCoAuthor.commit` | boolean | Add a Co-authored-by trailer to git commit messages AND attach a per-file AI-attribution git note (`refs/notes/ai-attribution`) for commits made through Qwen Code. Disabling skips both. | `true` |
|
||||
| `general.gitCoAuthor.pr` | boolean | Append a Qwen Code attribution line to pull request descriptions when running `gh pr create`. | `true` |
|
||||
| `general.checkpointing.enabled` | boolean | Enable session checkpointing for recovery. | `false` |
|
||||
| `general.defaultFileEncoding` | string | Default encoding for new files. Use `"utf-8"` (default) for UTF-8 without BOM, or `"utf-8-bom"` for UTF-8 with BOM. Only change this if your project specifically requires BOM. | `"utf-8"` |
|
||||
|
||||
#### output
|
||||
|
||||
|
|
|
|||
|
|
@ -395,7 +395,7 @@ const SETTINGS_SCHEMA = {
|
|||
requiresRestart: false,
|
||||
default: true,
|
||||
description:
|
||||
'Automatically add a Co-authored-by trailer to git commit messages when commits are made through Qwen Code.',
|
||||
'Add a Co-authored-by trailer to git commit messages AND attach a per-file AI-attribution git note (`refs/notes/ai-attribution`) for commits made through Qwen Code. Disabling skips both.',
|
||||
showInDialog: true,
|
||||
},
|
||||
pr: {
|
||||
|
|
|
|||
|
|
@ -28,6 +28,7 @@ const sampleNote: CommitAttributionNote = {
|
|||
},
|
||||
surfaceBreakdown: { cli: { aiChars: 150, percent: 38 } },
|
||||
excludedGenerated: ['package-lock.json'],
|
||||
excludedGeneratedCount: 1,
|
||||
promptCount: 3,
|
||||
};
|
||||
|
||||
|
|
@ -68,6 +69,7 @@ describe('attributionTrailer', () => {
|
|||
...sampleNote,
|
||||
files: {},
|
||||
excludedGenerated: [],
|
||||
excludedGeneratedCount: 0,
|
||||
};
|
||||
for (let i = 0; i < 2000; i++) {
|
||||
hugeNote.files[
|
||||
|
|
|
|||
|
|
@ -91,10 +91,8 @@ export function formatAttributionSummary(note: CommitAttributionNote): string {
|
|||
);
|
||||
}
|
||||
|
||||
if (note.excludedGenerated.length > 0) {
|
||||
lines.push(
|
||||
` Excluded generated: ${note.excludedGenerated.length} file(s)`,
|
||||
);
|
||||
if (note.excludedGeneratedCount > 0) {
|
||||
lines.push(` Excluded generated: ${note.excludedGeneratedCount} file(s)`);
|
||||
}
|
||||
|
||||
return lines.join('\n');
|
||||
|
|
|
|||
|
|
@ -429,6 +429,45 @@ describe('ChatRecordingService', () => {
|
|||
});
|
||||
});
|
||||
|
||||
describe('recordAttributionSnapshot', () => {
|
||||
const baseSnapshot = {
|
||||
type: 'attribution-snapshot' as const,
|
||||
version: 1,
|
||||
surface: 'cli',
|
||||
fileStates: {},
|
||||
baselines: {},
|
||||
promptCount: 0,
|
||||
promptCountAtLastCommit: 0,
|
||||
};
|
||||
|
||||
it('should write each distinct snapshot', async () => {
|
||||
chatRecordingService.recordAttributionSnapshot(baseSnapshot);
|
||||
chatRecordingService.recordAttributionSnapshot({
|
||||
...baseSnapshot,
|
||||
promptCount: 1,
|
||||
});
|
||||
chatRecordingService.recordAttributionSnapshot({
|
||||
...baseSnapshot,
|
||||
promptCount: 2,
|
||||
});
|
||||
await chatRecordingService.flush();
|
||||
expect(jsonl.writeLine).toHaveBeenCalledTimes(3);
|
||||
});
|
||||
|
||||
// Sessions that touch many files emit a non-retry turn snapshot
|
||||
// every prompt cycle. Without dedup, repeated identical snapshots
|
||||
// (no edits, no prompt-counter change) would re-serialize the entire
|
||||
// attribution state into the JSONL on every turn, inflating session
|
||||
// size and slowing /resume.
|
||||
it('should skip a snapshot identical to the previous write', async () => {
|
||||
chatRecordingService.recordAttributionSnapshot(baseSnapshot);
|
||||
chatRecordingService.recordAttributionSnapshot(baseSnapshot);
|
||||
chatRecordingService.recordAttributionSnapshot(baseSnapshot);
|
||||
await chatRecordingService.flush();
|
||||
expect(jsonl.writeLine).toHaveBeenCalledTimes(1);
|
||||
});
|
||||
});
|
||||
|
||||
// Note: Session management tests (listSessions, loadSession, deleteSession, etc.)
|
||||
// have been moved to sessionService.test.ts
|
||||
// Session resume integration tests should test via SessionService mock
|
||||
|
|
|
|||
|
|
@ -351,6 +351,16 @@ export class ChatRecordingService {
|
|||
*/
|
||||
private autoTitleController: AbortController | undefined;
|
||||
|
||||
/**
|
||||
* JSON-serialized form of the most recent attribution snapshot we
|
||||
* wrote, used to deduplicate identical writes on every non-retry
|
||||
* turn. Without this, sessions that touch many files would write a
|
||||
* full duplicate of the entire snapshot to the JSONL on every turn,
|
||||
* inflating the on-disk session and making `/resume` slower to
|
||||
* hydrate.
|
||||
*/
|
||||
private lastAttributionSnapshotJson: string | undefined;
|
||||
|
||||
constructor(config: Config) {
|
||||
this.config = config;
|
||||
this.lastRecordUuid =
|
||||
|
|
@ -969,9 +979,19 @@ export class ChatRecordingService {
|
|||
* Records an attribution state snapshot for session persistence.
|
||||
* Called at the start of every non-retry turn so that a resumed session
|
||||
* sees the most recent state including edits made during the prior turn.
|
||||
*
|
||||
* Deduplicates identical successive writes: if the snapshot's JSON
|
||||
* form is byte-identical to the last one we wrote, skip the append.
|
||||
* Without this, sessions that touch many files would write a full
|
||||
* duplicate of the entire snapshot to the JSONL on every turn, even
|
||||
* when nothing changed — inflating session size and slowing /resume.
|
||||
*/
|
||||
recordAttributionSnapshot(snapshot: AttributionSnapshot): void {
|
||||
try {
|
||||
const json = JSON.stringify(snapshot);
|
||||
if (json === this.lastAttributionSnapshotJson) {
|
||||
return;
|
||||
}
|
||||
const record: ChatRecord = {
|
||||
...this.createBaseRecord('system'),
|
||||
type: 'system',
|
||||
|
|
@ -980,6 +1000,7 @@ export class ChatRecordingService {
|
|||
};
|
||||
|
||||
this.appendRecord(record);
|
||||
this.lastAttributionSnapshotJson = json;
|
||||
} catch (error) {
|
||||
debugLogger.error('Error saving attribution snapshot:', error);
|
||||
}
|
||||
|
|
|
|||
|
|
@ -83,10 +83,27 @@ export interface CommitAttributionNote {
|
|||
surfaces: string[];
|
||||
};
|
||||
surfaceBreakdown: Record<string, { aiChars: number; percent: number }>;
|
||||
/**
|
||||
* Sample of generated/vendored files that were excluded from
|
||||
* attribution. Capped at `MAX_EXCLUDED_GENERATED_SAMPLE` paths so a
|
||||
* commit churning thousands of `dist/` artifacts can't blow past the
|
||||
* 30 KB note budget and silently drop attribution for the real
|
||||
* source files in the same commit. Use `excludedGeneratedCount` for
|
||||
* the true total.
|
||||
*/
|
||||
excludedGenerated: string[];
|
||||
/** Total count of excluded files (≥ excludedGenerated.length). */
|
||||
excludedGeneratedCount: number;
|
||||
promptCount: number;
|
||||
}
|
||||
|
||||
/**
|
||||
* Upper bound on the number of excluded-generated paths we serialize
|
||||
* into the git note. Keeps the JSON payload bounded for commits with
|
||||
* lots of generated artifacts.
|
||||
*/
|
||||
export const MAX_EXCLUDED_GENERATED_SAMPLE = 50;
|
||||
|
||||
/** Result of running git commands to get staged file info. */
|
||||
export interface StagedFileInfo {
|
||||
files: string[];
|
||||
|
|
@ -439,6 +456,7 @@ export class CommitAttributionService {
|
|||
|
||||
const files: Record<string, FileAttributionDetail> = {};
|
||||
const excludedGenerated: string[] = [];
|
||||
let excludedGeneratedCount = 0;
|
||||
const surfaceCounts: Record<string, number> = {};
|
||||
let totalAiChars = 0;
|
||||
let totalHumanChars = 0;
|
||||
|
|
@ -463,7 +481,12 @@ export class CommitAttributionService {
|
|||
|
||||
for (const relFile of stagedInfo.files) {
|
||||
if (isGeneratedFile(relFile)) {
|
||||
excludedGenerated.push(relFile);
|
||||
excludedGeneratedCount++;
|
||||
// Cap the sample so a commit churning thousands of `dist/`
|
||||
// artifacts can't blow past the 30 KB note budget.
|
||||
if (excludedGenerated.length < MAX_EXCLUDED_GENERATED_SAMPLE) {
|
||||
excludedGenerated.push(relFile);
|
||||
}
|
||||
continue;
|
||||
}
|
||||
|
||||
|
|
@ -535,6 +558,7 @@ export class CommitAttributionService {
|
|||
},
|
||||
surfaceBreakdown,
|
||||
excludedGenerated,
|
||||
excludedGeneratedCount,
|
||||
promptCount: this.getPromptsSinceLastCommit(),
|
||||
};
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue