aboutsummaryrefslogtreecommitdiff
path: root/src
diff options
context:
space:
mode:
authorzousar <[email protected]>2025-04-03 17:14:57 -0600
committerzousar <[email protected]>2025-04-03 17:14:57 -0600
commit708735d15da5960ff3bccc1ca22351d70d54880f (patch)
treeca6842ae4d7032305bf302eb6bdf3b36acd90243 /src
parentbuild store save access times (#341) (diff)
downloadzen-708735d15da5960ff3bccc1ca22351d70d54880f.tar.xz
zen-708735d15da5960ff3bccc1ca22351d70d54880f.zip
Don't duplicate ID bytes when more than one pkg_data
ID was getting extended and left shifted if we encountered multiple package data items in a single entry. So instead of the ID being 0x0c6500b7fb8dbe2e, it was 0x0C6500B7FB8DBE2E0C6500B7FB8DBE2E. When we went to look up an imported package by ID, it would not be found and the import would be presented as a blank string. Addressing this by making the first package data the only referenceable one. Second package datas are currently used for optional data blobs, and will not be imported or referenced. They are sidecar data.
Diffstat (limited to 'src')
-rw-r--r--src/zenserver/frontend/html/indexer/worker.js1
1 files changed, 1 insertions, 0 deletions
diff --git a/src/zenserver/frontend/html/indexer/worker.js b/src/zenserver/frontend/html/indexer/worker.js
index 25c8d7671..dd0f8bb17 100644
--- a/src/zenserver/frontend/html/indexer/worker.js
+++ b/src/zenserver/frontend/html/indexer/worker.js
@@ -101,6 +101,7 @@ async function map_id_to_key(project_id, oplog, start, end, page_size, stride)
id <<= 8n;
id |= BigInt(pkg_id[i]);
}
+ break;
}
if (bulk_data)