From f4c6a76651cd21fad871e916807e847d611abe4e Mon Sep 17 00:00:00 2001
From: Classic298 <27028174+Classic298@users.noreply.github.com>
Date: Fri, 31 Jul 2026 23:24:57 +0200
Subject: [PATCH] perf: use the orjson codec for Valkey vector metadata
 (#27805)

The Valkey backend serializes chunk metadata on every insert and parses it back on every result row in `get` and `query`. Both directions now go through `JSONCodec`, which selects orjson when `ENABLE_ORJSON` is set.

The stored `metadata_json` field is never matched against as text. `_build_filter_expression` only emits TAG predicates, and the TAG fields are `id`, `hash`, `file_id`, `source` and `knowledge_base_id`; `metadata_json` appears only as a return field that is immediately re-parsed. So rows written with escaped non-ASCII and rows written raw are indistinguishable to every reader, and no migration is needed.

`process_metadata` already stringifies datetimes and strips null bytes and lone surrogates before the write, so the two backends cannot disagree about what is serializable here.

Both read `except` clauses widen from `(json.JSONDecodeError, TypeError)` to `(ValueError, TypeError)`. The codec falls back to engineio's codec, which installs `parse_int=_safe_int` and raises a bare `ValueError` for integer literals longer than 100 characters; the narrower clause would have let that escape and abort a search instead of yielding empty metadata. `json.JSONDecodeError` is a `ValueError` subclass, so this is a strict superset. That removes the module's last use of stdlib `json`, so the import goes with it.

With `ENABLE_ORJSON` unset, which is the default, `JSONCodec` is stdlib `json` and this call site behaves exactly as before.
---
 backend/open_webui/retrieval/vector/dbs/valkey.py | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/backend/open_webui/retrieval/vector/dbs/valkey.py b/backend/open_webui/retrieval/vector/dbs/valkey.py
index 1db10281a4..950ccf969f 100644
--- a/backend/open_webui/retrieval/vector/dbs/valkey.py
+++ b/backend/open_webui/retrieval/vector/dbs/valkey.py
@@ -2,7 +2,6 @@
 # Requires Valkey core >= 9.0.1 with the valkey-search module >= 1.2.0 loaded.
 
 import atexit
-import json
 import logging
 import re
 import struct
@@ -24,6 +23,7 @@ from open_webui.retrieval.vector.main import (
     VectorItem,
 )
 from open_webui.retrieval.vector.utils import process_metadata
+from open_webui.utils.json_codec import JSONCodec
 
 log = logging.getLogger(__name__)
 
@@ -482,7 +482,7 @@ class ValkeyClient(VectorDBBase):
                 'id': item['id'],
                 'vector': _vector_to_bytes(item['vector']),
                 'text': item['text'],
-                'metadata_json': json.dumps(metadata),
+                'metadata_json': JSONCodec.dumps(metadata),
                 # `or ''` prevents indexing literal 'None' as a TAG value, which would
                 # poison $ne / equality queries.
                 'hash': str(metadata.get('hash') or ''),
@@ -588,8 +588,8 @@ class ValkeyClient(VectorDBBase):
                     ids.append(_decode(fields.get(b'id', b'')))
                     documents.append(_decode(fields.get(b'text', b'')))
                     try:
-                        metadatas.append(json.loads(_decode(fields.get(b'metadata_json', b'{}'))))
-                    except (json.JSONDecodeError, TypeError):
+                        metadatas.append(JSONCodec.loads(_decode(fields.get(b'metadata_json', b'{}'))))
+                    except (ValueError, TypeError):
                         metadatas.append({})
                     if limit is not None and limit > 0 and len(ids) >= limit:
                         return GetResult(ids=[ids], documents=[documents], metadatas=[metadatas])
@@ -734,8 +734,8 @@ class ValkeyClient(VectorDBBase):
             ids.append(_decode(fields.get(b'id', b'')))
             documents.append(_decode(fields.get(b'text', b'')))
             try:
-                metadatas.append(json.loads(_decode(fields.get(b'metadata_json', b'{}'))))
-            except (json.JSONDecodeError, TypeError):
+                metadatas.append(JSONCodec.loads(_decode(fields.get(b'metadata_json', b'{}'))))
+            except (ValueError, TypeError):
                 metadatas.append({})
 
             if include_score: