From 1a44f890c85ce51106d79d41b581f296877c9b5b Mon Sep 17 00:00:00 2001
From: allaun <allaun@researchstack.info>
Date: Mon, 22 Jun 2026 02:48:51 -0500
Subject: [PATCH] fix(infra): make lake_build_ingest.py duplicate-safe

Sets content_hash to NULL for build_log packages so the partial unique

index ene_pkg_hash_idx does not prevent multiple ingestions of the same

build output. Each build session now gets its own package row. Also makes

receipt/ingest_event inserts idempotent with ON CONFLICT DO NOTHING.
---
 4-Infrastructure/shim/lake_build_ingest.py | 26 ++++++++++------------
 1 file changed, 12 insertions(+), 14 deletions(-)

diff --git a/4-Infrastructure/shim/lake_build_ingest.py b/4-Infrastructure/shim/lake_build_ingest.py
index a04ce87a..e3bf39f6 100644
--- a/4-Infrastructure/shim/lake_build_ingest.py
+++ b/4-Infrastructure/shim/lake_build_ingest.py
@@ -184,10 +184,13 @@ def _insert_session(target: str, git: dict[str, str], build: dict[str, Any], ela
 
 
 def _insert_package(session_id: str, target: str, git: dict[str, str], build: dict[str, Any]) -> str:
-    """Insert a package row representing the build artifact and return its id."""
-    pkg_id = hashlib.sha256(f"lake-build:{git['commit']}:{target}".encode()).hexdigest()[:32]
+    """Insert a package row representing the build artifact and return its id.
+
+    Build logs get content_hash = NULL so the partial unique index
+    ene_pkg_hash_idx (WHERE content_hash IS NOT NULL) does not fire.
+    Each build session therefore gets its own package row."""
+    pkg_id = hashlib.sha256(f"lake-build:{session_id}:{git['commit']}:{target}".encode()).hexdigest()[:32]
     content = build["raw_output"]
-    content_hash = hashlib.sha256(content.encode()).hexdigest()
     title = f"lake build {target}" if target else "lake build"
     sql = f"""
     INSERT INTO ene.packages
@@ -196,20 +199,13 @@ def _insert_package(session_id: str, target: str, git: dict[str, str], build: di
          verification_status, promotion_state, domain, archetype, notes)
     VALUES
         ('{pkg_id}', 'build_log', '{title}', '{_escape_sql(content)}',
-         '{content_hash}', 'METAL',
+         NULL, 'METAL',
          '["lake", "build", "lean"]'::jsonb,
          '4-Infrastructure/shim/lake_build_ingest.py',
          '{_now()}', '{_now()}', '{session_id}',
          '{{"git_commit": "{git['commit']}", "git_branch": "{git['branch']}", "dirty": {str(git['dirty']).lower()}}}'::jsonb,
          '{build['status']}', 'held', 'core-formalism', 'build_artifact',
-         'Build output captured from lake build {target}')
-    ON CONFLICT (pkg) DO UPDATE SET
-        content = EXCLUDED.content,
-        content_hash = EXCLUDED.content_hash,
-        updated_at = EXCLUDED.updated_at,
-        session_id = EXCLUDED.session_id,
-        verification_status = EXCLUDED.verification_status,
-        notes = EXCLUDED.notes;
+         'Build output captured from lake build {target}');
     """
     _ssh_psql(sql)
     return pkg_id
@@ -230,7 +226,8 @@ def _insert_receipt(pkg_id: str, session_id: str, build: dict[str, Any]) -> None
     VALUES
         ('{pkg_id}', 'lake_build', '{receipt_hash}', '{input_hash}',
          '{receipt_hash}', 'lake build', 'opencode', '{build['status']}',
-         '{residual}'::jsonb);
+         '{residual}'::jsonb)
+    ON CONFLICT (package_id, receipt_type, receipt_hash) DO NOTHING;
     """
     _ssh_psql(sql)
 
@@ -246,7 +243,8 @@ def _insert_ingest_event(pkg_id: str, session_id: str) -> None:
         ('{pkg_id}', 'file://4-Infrastructure/shim/lake_build_ingest.py',
          'build_log_ingest', '', 'lake_build_ingest_v1', '1.0.0',
          '{{"session_id": "{session_id}"}}'::jsonb, '[]'::jsonb,
-         '[]'::jsonb, '[]'::jsonb, '[]'::jsonb);
+         '[]'::jsonb, '[]'::jsonb, '[]'::jsonb)
+    ON CONFLICT DO NOTHING;
     """
     _ssh_psql(sql)