CalvinBackup/gstack
1
0
Fork 0
mirror of https://github.com/garrytan/gstack.git synced 2026-09-12 16:08:59 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix(memory-ingest): parse the current Codex response_item rollout shape

Browse Source 
Fixes #2105. Codex rollout JSONL moved to
{ type: 'response_item', payload: { type: 'message', role, content: [...] } };
the parser's legacy payload.message branch never fired on it, so every Codex
session imported as an empty shell (message_count: 0 — 243/243 sessions on
the reporting machine). Both shapes now parse; non-message response_items
(reasoning etc.) are ignored. parseTranscriptJsonl exported for direct unit
tests (CLI path unchanged — import.meta.main guard).

Note: #2104's staging-in-gitignored-tree half is already defended on main
(--include-gitignored + GIT_CEILING_DIRECTORIES, #2144, plus the #2486
reconcile guard) — verified, no change needed; it moves to the close-only
roster.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Garry Tan
2026-08-16 09:07:27 -07:00
co-authored by Claude Fable 5
parent c2cdf65176
commit 00d0115ac7
2 changed files with 57 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
bin/gstack-memory-ingest.ts
+14 -2
View File
@@ -543,7 +543,7 @@ interface ParsedSession {
partial: boolean;
}
function parseTranscriptJsonl(path: string): ParsedSession | null {
export function parseTranscriptJsonl(path: string): ParsedSession | null {
// Best-effort tolerant parser. Handles truncated last lines (D10 partial-flag).
let raw: string;
try {
@@ -619,7 +619,7 @@ function parseTranscriptJsonl(path: string): ParsedSession | null {
const tool = rec?.name || rec?.tool || rec?.tool_call?.name || "tool";
bodyParts.push(`### Tool call: ${tool}`);
} else if (isCodex && rec?.payload?.message) {
// Codex shape: each record has payload.message
// Legacy Codex shape: each record has payload.message
const msg = rec.payload.message;
const role = msg.role || "user";
const content = extractContentText(msg);
@@ -627,6 +627,18 @@ function parseTranscriptJsonl(path: string): ParsedSession | null {
bodyParts.push(`## ${role.charAt(0).toUpperCase() + role.slice(1)}\n\n${content}`);
messageCount++;
}
} else if (isCodex && rec?.type === "response_item" && rec?.payload?.type === "message") {
// Current Codex rollout shape (#2105): records are
// { type: 'response_item', payload: { type: 'message', role, content: [...] } }.
// The legacy payload.message branch never fires on these, which rendered
// every Codex session as an empty shell (message_count: 0, 243/243 on
// the reporting machine). Flatten payload.content like the Claude branch.
const role = rec.payload.role || "user";
const content = extractContentText(rec.payload);
if (content) {
bodyParts.push(`## ${role.charAt(0).toUpperCase() + role.slice(1)}\n\n${content}`);
messageCount++;
}
}
}