mirror of
https://github.com/garrytan/gstack.git
synced 2026-06-19 00:00:13 +02:00
fix(gbrain-sync): centralize gbrain spawn surface + seed DATABASE_URL
Cherry-picked from #1508 by jasshultz, restructured per codex review #4 and #7 to widen scope and centralize the spawn surface. The bug: gbrain auto-loads .env.local from cwd via dotenv. When /sync-gbrain runs inside a Next.js / Prisma / Rails project whose .env.local defines its own DATABASE_URL (pointing at the app's local DB), gbrain reads that value instead of its own ~/.gbrain/config.json — auth fails, code + memory stages crash. This commit: - Adds lib/gbrain-exec.ts: buildGbrainEnv, spawnGbrain, execGbrainJson, execGbrainText, spawnGbrainAsync (the last one for memory-ingest's streaming gbrain import call). buildGbrainEnv seeds DATABASE_URL from ${GBRAIN_HOME:-$HOME/.gbrain}/config.json, returns a fresh env object (never the caller's by identity — codex review #11), and honors the GSTACK_RESPECT_ENV_DATABASE_URL=1 escape hatch. - Routes every gbrain spawn in bin/gstack-gbrain-sync.ts and bin/gstack-memory-ingest.ts through the helpers. Both files now own zero direct spawnSync("gbrain"|spawn("gbrain"|execFileSync("gbrain" call sites. - Threads buildGbrainEnv into the spawnSync("bun", [memory-ingest], ...) grandchild in runMemoryIngest (codex review #7). Without this, the parent fix is half-baked — the bun child inherits a clean env but needs DATABASE_URL pre-seeded too. spawnGbrainAsync inside memory-ingest provides defense in depth for standalone invocations. - Adds GBRAIN_HOME support — aligns with detectEngineTier (already honors GBRAIN_HOME) so all gstack-side gbrain calls agree on which config file matters. Resolves baseEnv.HOME first, then homedir(), so test injection works without process-wide HOME mutation. - Adds test/build-gbrain-env.test.ts: 10 unit tests covering all five env-seeding branches (seed from config / override caller / GSTACK_RESPECT escape hatch / missing config / unparseable config / no database_url field / GBRAIN_HOME path / object-identity guard / unrelated-vars preservation / idempotent-when-matches). - Adds test/gbrain-exec-invariant.test.ts: static-source check that greps both bin/gstack-gbrain-sync.ts and bin/gstack-memory-ingest.ts for direct spawnSync("gbrain"|spawn("gbrain"|execFileSync("gbrain"| execSync(...gbrain matches and fails the build if any are found. Refactor-proof against future contributors adding a new gbrain spawn without env threading. The invariant is intentionally narrow — only the two files where the DATABASE_URL bug actually hurts users are guarded. Migrating the spawn sites in lib/gbrain-local-status.ts, lib/gstack-memory-helpers.ts, and bin/gstack-brain-context-load.ts is a follow-up. Co-Authored-By: Jason Shultz <jasshultz@gmail.com> Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
+32
-41
@@ -39,6 +39,7 @@ import "../lib/conductor-env-shim";
|
||||
import { detectEngineTier, withErrorContext, canonicalizeRemote } from "../lib/gstack-memory-helpers";
|
||||
import { ensureSourceRegistered, sourcePageCount } from "../lib/gbrain-sources";
|
||||
import { localEngineStatus, type LocalEngineStatus } from "../lib/gbrain-local-status";
|
||||
import { buildGbrainEnv, spawnGbrain, execGbrainJson } from "../lib/gbrain-exec";
|
||||
|
||||
// ── Types ──────────────────────────────────────────────────────────────────
|
||||
|
||||
@@ -262,11 +263,9 @@ export function _resetGbrainSupportsRenameCache(): void {
|
||||
function gbrainSupportsSourcesRename(env?: NodeJS.ProcessEnv): boolean {
|
||||
if (_gbrainSupportsRenameCache !== null) return _gbrainSupportsRenameCache;
|
||||
try {
|
||||
const r = spawnSync("gbrain", ["sources", "rename", "--help"], {
|
||||
encoding: "utf-8",
|
||||
const r = spawnGbrain(["sources", "rename", "--help"], {
|
||||
timeout: 5_000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
env: env || process.env,
|
||||
baseEnv: env,
|
||||
});
|
||||
const out = `${r.stdout || ""}\n${r.stderr || ""}`;
|
||||
// Match the exact argument shape: `rename <old> <new>` (with literal
|
||||
@@ -290,20 +289,13 @@ function gbrainSupportsSourcesRename(env?: NodeJS.ProcessEnv): boolean {
|
||||
* helper can be exercised without a real gbrain CLI.
|
||||
*/
|
||||
export function sourceLocalPath(sourceId: string, env?: NodeJS.ProcessEnv): string | null {
|
||||
try {
|
||||
const r = spawnSync("gbrain", ["sources", "list", "--json"], {
|
||||
encoding: "utf-8",
|
||||
timeout: 30_000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
env: env || process.env,
|
||||
});
|
||||
if (r.status !== 0) return null;
|
||||
const list = JSON.parse(r.stdout || "[]") as Array<{ id: string; local_path?: string }>;
|
||||
const found = list.find((s) => s.id === sourceId);
|
||||
return found?.local_path ?? null;
|
||||
} catch {
|
||||
return null;
|
||||
}
|
||||
const list = execGbrainJson<Array<{ id: string; local_path?: string }>>(
|
||||
["sources", "list", "--json"],
|
||||
{ baseEnv: env },
|
||||
);
|
||||
if (!list) return null;
|
||||
const found = list.find((s) => s.id === sourceId);
|
||||
return found?.local_path ?? null;
|
||||
}
|
||||
|
||||
/** Result of `planHostnameFoldMigration` — informs `runCodeImport` of next steps. */
|
||||
@@ -352,12 +344,7 @@ export function planHostnameFoldMigration(
|
||||
};
|
||||
}
|
||||
if (gbrainSupportsSourcesRename(env)) {
|
||||
const r = spawnSync("gbrain", ["sources", "rename", legacyPathHashId, newSourceId], {
|
||||
encoding: "utf-8",
|
||||
timeout: 30_000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
env: env || process.env,
|
||||
});
|
||||
const r = spawnGbrain(["sources", "rename", legacyPathHashId, newSourceId], { baseEnv: env });
|
||||
if (r.status === 0) {
|
||||
return { kind: "renamed", oldId: legacyPathHashId, newId: newSourceId };
|
||||
}
|
||||
@@ -377,12 +364,8 @@ export function planHostnameFoldMigration(
|
||||
* flag-helper centralization in commit 4 (lib/gbrain-exec.ts) will resolve
|
||||
* the inconsistency across the codebase.
|
||||
*/
|
||||
export function removeOrphanedSource(oldId: string): boolean {
|
||||
const r = spawnSync("gbrain", ["sources", "remove", oldId, "--confirm-destructive"], {
|
||||
encoding: "utf-8",
|
||||
timeout: 30_000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
});
|
||||
export function removeOrphanedSource(oldId: string, env?: NodeJS.ProcessEnv): boolean {
|
||||
const r = spawnGbrain(["sources", "remove", oldId, "--confirm-destructive"], { baseEnv: env });
|
||||
return r.status === 0;
|
||||
}
|
||||
|
||||
@@ -555,13 +538,16 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
// hits forever if we left the orphan in place. Remove the legacy id once
|
||||
// here so users don't accumulate orphans.
|
||||
// Failure is non-fatal — we still register the new id below.
|
||||
// gbrainEnv seeds DATABASE_URL from gbrain's config so this stage works
|
||||
// inside Next.js / Prisma / Rails projects with their own .env.local
|
||||
// (codex review #7 — bug fix is wider than #1508 as filed).
|
||||
const gbrainEnv = buildGbrainEnv({ announce: !args.quiet });
|
||||
const legacyId = deriveLegacyCodeSourceId(root);
|
||||
let legacyRemoved = false;
|
||||
if (legacyId !== sourceId) {
|
||||
const rm = spawnSync("gbrain", ["sources", "remove", legacyId, "--confirm-destructive"], {
|
||||
encoding: "utf-8",
|
||||
const rm = spawnGbrain(["sources", "remove", legacyId, "--confirm-destructive"], {
|
||||
timeout: 30_000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
baseEnv: gbrainEnv,
|
||||
});
|
||||
// Treat absent-source as success (clean state). gbrain emits "not found" on
|
||||
// missing id; treat any non-zero exit without "not found" as a soft fail.
|
||||
@@ -575,7 +561,7 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
// pages); fall back to register-new → sync-OK → remove-old. Path-drift
|
||||
// (user moved the repo, etc.) skips migration with a warning.
|
||||
const pathOnlyHashLegacyId = derivePathOnlyHashLegacyId(root);
|
||||
const migration = planHostnameFoldMigration(root, sourceId, pathOnlyHashLegacyId);
|
||||
const migration = planHostnameFoldMigration(root, sourceId, pathOnlyHashLegacyId, gbrainEnv);
|
||||
if (migration.kind === "skipped-path-drift" && !args.quiet) {
|
||||
console.error(
|
||||
`[sync:code] hostname-fold migration skipped: legacy source ${migration.oldId} `
|
||||
@@ -590,7 +576,7 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
// no synchronous duplicate here (per /codex review #12).
|
||||
let registered = false;
|
||||
try {
|
||||
const result = await ensureSourceRegistered(sourceId, root, { federated: true });
|
||||
const result = await ensureSourceRegistered(sourceId, root, { federated: true, env: gbrainEnv });
|
||||
registered = result.changed;
|
||||
} catch (err) {
|
||||
return {
|
||||
@@ -608,9 +594,10 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
? ["reindex-code", "--source", sourceId, "--yes"]
|
||||
: ["sync", "--strategy", "code", "--source", sourceId];
|
||||
|
||||
const syncResult = spawnSync("gbrain", syncArgs, {
|
||||
const syncResult = spawnGbrain(syncArgs, {
|
||||
stdio: args.quiet ? ["ignore", "ignore", "ignore"] : ["ignore", "inherit", "inherit"],
|
||||
timeout: 35 * 60 * 1000,
|
||||
baseEnv: gbrainEnv,
|
||||
});
|
||||
|
||||
if (syncResult.status !== 0) {
|
||||
@@ -633,13 +620,12 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
// the wrong/default source. Treat it as a stage failure (ok=false) so the
|
||||
// verdict block surfaces ERR and the user knows to retry rather than
|
||||
// trusting stale results.
|
||||
const attach = spawnSync("gbrain", ["sources", "attach", sourceId], {
|
||||
encoding: "utf-8",
|
||||
const attach = spawnGbrain(["sources", "attach", sourceId], {
|
||||
timeout: 10_000,
|
||||
cwd: root,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
baseEnv: gbrainEnv,
|
||||
});
|
||||
const pageCount = sourcePageCount(sourceId);
|
||||
const pageCount = sourcePageCount(sourceId, gbrainEnv);
|
||||
|
||||
// Step 4: Deferred hostname-fold cleanup.
|
||||
// Only remove the pre-#1468 path-only-hash source NOW that the new source
|
||||
@@ -649,7 +635,7 @@ async function runCodeImport(args: CliArgs): Promise<StageResult> {
|
||||
// safety: register → sync → verify → THEN delete.
|
||||
let hostnameLegacyRemoved = false;
|
||||
if (migration.kind === "pending-cleanup" && pageCount !== null && pageCount > 0) {
|
||||
hostnameLegacyRemoved = removeOrphanedSource(migration.oldId);
|
||||
hostnameLegacyRemoved = removeOrphanedSource(migration.oldId, gbrainEnv);
|
||||
if (hostnameLegacyRemoved && !args.quiet) {
|
||||
console.error(`[sync:code] hostname-fold migration: removed legacy ${migration.oldId} after new source sync verified (page_count=${pageCount})`);
|
||||
}
|
||||
@@ -718,9 +704,14 @@ function runMemoryIngest(args: CliArgs): StageResult {
|
||||
else ingestArgs.push("--incremental");
|
||||
if (args.quiet) ingestArgs.push("--quiet");
|
||||
|
||||
// Thread the seeded env into the bun grandchild (codex review #7 — the
|
||||
// .env.local footgun affects gstack-memory-ingest.ts too, not just the
|
||||
// direct gbrain spawns in this file). The grandchild calls gbrain import
|
||||
// internally and must see the DATABASE_URL from gbrain's own config.
|
||||
const result = spawnSync("bun", ingestArgs, {
|
||||
encoding: "utf-8",
|
||||
timeout: 35 * 60 * 1000,
|
||||
env: buildGbrainEnv({ announce: false }),
|
||||
});
|
||||
|
||||
// D6: parse [memory-ingest] lines from the child's stderr. ERR-prefixed
|
||||
|
||||
+10
-10
@@ -64,6 +64,7 @@ import {
|
||||
detectEngineTier,
|
||||
withErrorContext,
|
||||
} from "../lib/gstack-memory-helpers";
|
||||
import { execGbrainText, spawnGbrainAsync } from "../lib/gbrain-exec";
|
||||
|
||||
// ── Types ──────────────────────────────────────────────────────────────────
|
||||
|
||||
@@ -813,11 +814,10 @@ function gbrainAvailable(): boolean {
|
||||
// `import <dir>` (batch markdown import via path-authoritative slugs).
|
||||
// If absent, we surface a single clean error here rather than failing
|
||||
// the whole stage with a confusing usage message from gbrain itself.
|
||||
const help = execFileSync("gbrain", ["--help"], {
|
||||
encoding: "utf-8",
|
||||
timeout: 5000,
|
||||
stdio: ["ignore", "pipe", "pipe"],
|
||||
});
|
||||
// `gbrain --help` probes only CLI availability, not DB connectivity, so
|
||||
// it doesn't strictly need DATABASE_URL. But routing through the helper
|
||||
// keeps the invariant test from chasing exceptions per call site.
|
||||
const help = execGbrainText(["--help"], { timeout: 5000 });
|
||||
_gbrainAvailability = /^\s+import\s/m.test(help);
|
||||
} catch {
|
||||
_gbrainAvailability = false;
|
||||
@@ -1316,11 +1316,11 @@ function runGbrainImport(
|
||||
): Promise<{ status: number | null; stdout: string; stderr: string }> {
|
||||
installSignalForwarder();
|
||||
return new Promise((resolve) => {
|
||||
const child = spawn(
|
||||
"gbrain",
|
||||
["import", stagingDir, "--no-embed", "--json"],
|
||||
{ stdio: ["ignore", "pipe", "pipe"] },
|
||||
);
|
||||
// Seed DATABASE_URL from gbrain's own config so this stage works
|
||||
// inside Next.js / Prisma / Rails projects with their own
|
||||
// .env.local (codex review #7 — defense in depth on top of the
|
||||
// parent gstack-gbrain-sync seeding the bun grandchild's env).
|
||||
const child = spawnGbrainAsync(["import", stagingDir, "--no-embed", "--json"]);
|
||||
_activeImportChild = child;
|
||||
let stdout = "";
|
||||
let stderr = "";
|
||||
|
||||
Reference in New Issue
Block a user