Merge PR #8598: fix(session): repair torn JSONL appends (@roboomp)

# Conflicts:
#	packages/coding-agent/src/session/session-loader.ts
#	packages/coding-agent/src/session/session-manager.ts
#	packages/coding-agent/test/session-loader-stream.test.ts
This commit is contained in:
can1357
2026-08-16 02:48:12 +02:00
11 changed files with 232 additions and 104 deletions
@@ -2,14 +2,7 @@ import type { AgentMessage } from "@oh-my-pi/pi-agent-core";
import { getBlobsDir, isEnoent, parseJsonlLenient } from "@oh-my-pi/pi-utils";
import { BlobStore, isBlobRef, resolveImageData, resolveImageDataUrl } from "./blob-store";
import { buildSessionContext } from "./session-context";
import {
type FileEntry,
type RawFileEntry,
SESSION_TITLE_SLOT_BYTES,
type SessionEntry,
type SessionHeader,
type SessionTitleSlotEntry,
} from "./session-entries";
import type { FileEntry, RawFileEntry, SessionEntry, SessionHeader } from "./session-entries";
import { migrateToCurrentVersion } from "./session-migrations";
import { isImageBlock, isImageDataPayload } from "./session-persistence";
import { FileSessionStorage, type SessionStorage } from "./session-storage";
@@ -33,6 +26,15 @@ export interface VisitEntriesFromFileStreamOptions {
yieldEveryBytes?: number;
/** Yield to the macrotask queue after this many entries have been visited. */
yieldEveryEntries?: number;
/** Called once for every malformed JSONL record skipped by the stream. */
onMalformedRecord?: () => void;
}
/** Parsed session entries plus corruption metadata needed by writable loaders. */
export interface SessionLoadResult {
entries: FileEntry[];
titleSlot: SessionTitleUpdate | undefined;
malformedRecords: number;
}
function splitTitleSlot(content: string): { body: string; slot: SessionTitleUpdate | undefined } {
@@ -61,14 +63,16 @@ function applyTitleSlot(entry: FileEntry | undefined, slot: SessionTitleUpdate |
}
/** Parse session JSONL while stripping and folding the optional fixed title slot. */
export function parseSessionContent(content: string): {
entries: FileEntry[];
titleSlot: SessionTitleUpdate | undefined;
} {
export function parseSessionContent(content: string): SessionLoadResult {
const { body, slot } = splitTitleSlot(content);
const entries = parseJsonlLenient<RawFileEntry>(body) as FileEntry[];
let malformedRecords = 0;
const entries = parseJsonlLenient<RawFileEntry>(body, {
onMalformedRecord: () => {
malformedRecords++;
},
}) as FileEntry[];
applyTitleSlot(entries[0], slot);
return { entries, titleSlot: slot };
return { entries, titleSlot: slot, malformedRecords };
}
/** Parse session JSONL and visit each entry without retaining prior entries. */
@@ -151,6 +155,15 @@ export async function visitEntriesFromFileStream(
// Malformed record: skip past the next newline and continue.
const nextNewline = buffer.indexOf(0x0a, read);
if (nextNewline === -1) break; // rest of the bad line not yet received
let nonWhitespace = false;
for (let index = read; index < nextNewline; index++) {
const byte = buffer[index];
if (byte !== 0x09 && byte !== 0x0d && byte !== 0x20) {
nonWhitespace = true;
break;
}
}
if (nonWhitespace) options.onMalformedRecord?.();
recordsSeen++;
buffer = buffer.subarray(nextNewline + 1);
if (recordsSeen >= maxRecords) {
@@ -211,33 +224,23 @@ export async function visitEntriesFromFileStream(
}
/** Exported for testing — the ≥8MiB streaming path (works on any file size). */
export async function loadEntriesFromFileStream(filePath: string): Promise<{
entries: FileEntry[];
titleSlot: SessionTitleUpdate | undefined;
}> {
export async function loadEntriesFromFileStream(filePath: string): Promise<SessionLoadResult> {
const entries: FileEntry[] = [];
const titleSlot = await visitEntriesFromFileStream(filePath, entry => {
entries.push(entry);
});
return { entries, titleSlot };
let malformedRecords = 0;
const titleSlot = await visitEntriesFromFileStream(
filePath,
entry => {
entries.push(entry);
},
{
onMalformedRecord: () => {
malformedRecords++;
},
},
);
return { entries, titleSlot, malformedRecords };
}
/** Read only the fixed-size head window to detect a physical title slot. */
export async function readTitleSlotFromFile(
filePath: string,
storage: SessionStorage = new FileSessionStorage(),
): Promise<SessionTitleSlotEntry | undefined> {
let head: string;
try {
[head] = await storage.readTextSlices(filePath, SESSION_TITLE_SLOT_BYTES, 0);
} catch (err) {
if (isEnoent(err)) return undefined;
throw err;
}
const newlineIndex = head.indexOf("\n");
if (newlineIndex < 0) return undefined;
return parseTitleSlotLine(head.slice(0, newlineIndex));
}
/** Exported for compaction.test.ts */
export function parseSessionEntries(content: string): FileEntry[] {
return parseSessionContent(content).entries;
@@ -247,25 +250,32 @@ function shouldStreamEntries(storage: SessionStorage, size: number): boolean {
return storage instanceof FileSessionStorage && size >= STREAM_LOAD_THRESHOLD_BYTES;
}
async function loadEntriesWithKnownSize(filePath: string, storage: SessionStorage, size: number): Promise<FileEntry[]> {
async function loadWithKnownSize(filePath: string, storage: SessionStorage, size: number): Promise<SessionLoadResult> {
const loaded = shouldStreamEntries(storage, size)
? await loadEntriesFromFileStream(filePath)
: parseSessionContent(await storage.readText(filePath));
const { entries } = loaded;
return isValidSessionHeader(entries[0]) ? entries : [];
return isValidSessionHeader(loaded.entries[0]) ? loaded : { ...loaded, entries: [] };
}
/** Exported for testing */
/** Load and validate a session while retaining malformed-record diagnostics. */
export async function loadSessionFile(
filePath: string,
storage: SessionStorage = new FileSessionStorage(),
): Promise<SessionLoadResult> {
try {
return await loadWithKnownSize(filePath, storage, storage.statSync(filePath).size);
} catch (err) {
if (isEnoent(err)) return { entries: [], titleSlot: undefined, malformedRecords: 0 };
throw err;
}
}
/** Load the valid entries from a session file, skipping malformed records. */
export async function loadEntriesFromFile(
filePath: string,
storage: SessionStorage = new FileSessionStorage(),
): Promise<FileEntry[]> {
try {
return await loadEntriesWithKnownSize(filePath, storage, storage.statSync(filePath).size);
} catch (err) {
if (isEnoent(err)) return [];
throw err;
}
return (await loadSessionFile(filePath, storage)).entries;
}
/**
@@ -290,7 +300,7 @@ export async function visitEntriesFromFile(
return;
}
for (const entry of await loadEntriesWithKnownSize(filePath, storage, size)) {
for (const entry of (await loadWithKnownSize(filePath, storage, size)).entries) {
if (visit(entry) === false) return;
}
}