Optimize text chunk retrieval with batch fetching

- Replace individual chunk fetches with batch get
- Simplify deduplication logic
- Improve error handling for missing data

lightrag/operate.py

@@ -2722,31 +2722,31 @@ async def _find_related_text_unit_from_relationships(
     all_text_units_lookup = {}
 
     # Deduplicate and preserve order | {c_id:order}
-    async def build_text_units_unique(text_units):
+    text_units_unique_flat = {}
-        text_units_flat = {}
+    for index, unit_list in enumerate(text_units):
-        for index, unit_list in enumerate(text_units):
+        for c_id in unit_list:
-            for c_id in unit_list:
+            if (
-                if c_id not in text_units_flat or index < text_units_flat[c_id]:
+                c_id not in text_units_unique_flat
-                    # Keep the smallest order
+                or index < text_units_unique_flat[c_id]
-                    text_units_flat[c_id] = index
+            ):
-        return text_units_flat
+                # Keep the smallest order
+                text_units_unique_flat[c_id] = index
 
-    text_units_unique_flat = build_text_units_unique(text_units)
+    if not text_units_unique_flat:
+        logger.warning("No valid text chunks found")
+        return []
 
-    async def fetch_chunk_data(c_id, index):
+    # Batch get all text chunk data
-        if c_id not in all_text_units_lookup:
+    chunk_ids = list(text_units_unique_flat.keys())
-            chunk_data = await text_chunks_db.get_by_id(c_id)
+    chunk_data_list = await text_chunks_db.get_by_ids(chunk_ids)
-            # Only store valid data
-            if chunk_data is not None and "content" in chunk_data:
-                all_text_units_lookup[c_id] = {
-                    "data": chunk_data,
-                    "order": index,
-                }
 
-    tasks = [
+    # Build lookup table, handling possible missing data
-        fetch_chunk_data(c_id, order) for c_id, order in text_units_unique_flat.items()
+    for chunk_id, chunk_data in zip(chunk_ids, chunk_data_list):
-    ]
+        if chunk_data is not None and "content" in chunk_data:
-    await asyncio.gather(*tasks)
+            all_text_units_lookup[chunk_id] = {
+                "data": chunk_data,
+                "order": text_units_unique_flat[chunk_id],
+            }
 
     if not all_text_units_lookup:
         logger.warning("No valid text chunks found")