Fix prompt template bugs: build template ignored and runtime override not wired (#173)

* Fix prompt template bugs in build and search

Bug 1: Build template ignored in new format
- Updated compute_embeddings_openai() to read build_prompt_template or prompt_template
- Updated compute_embeddings_ollama() with same fix
- Maintains backward compatibility with old single-template format

Bug 2: Runtime override not wired up
- Wired CLI search to pass provider_options to searcher.search()
- Enables runtime template override during search via --embedding-prompt-template

All 42 prompt template tests passing.

Fixes #155

* Fix: Prevent embedding server from applying templates during search

- Filter out all prompt templates (build_prompt_template, query_prompt_template, prompt_template) from provider_options when launching embedding server during search
- Templates are already applied in compute_query_embedding() before server call
- Prevents double-templating and ensures runtime override works correctly

This fixes the issue where --embedding-prompt-template during search was ignored because the server was applying build_prompt_template instead.

* Format code with ruff

Author: ww2283

packages/leann-core/src/leann/cli.py

@@ -1545,6 +1545,11 @@ async def search_documents(self, args):
                         print("Invalid input. Aborting search.")
                         return
 
+        # Build provider_options for runtime override
+        provider_options = {}
+        if args.embedding_prompt_template:
+            provider_options["prompt_template"] = args.embedding_prompt_template
+
         searcher = LeannSearcher(index_path=index_path)
         results = searcher.search(
             query,
@@ -1554,6 +1559,7 @@ async def search_documents(self, args):
             prune_ratio=args.prune_ratio,
             recompute_embeddings=args.recompute_embeddings,
             pruning_strategy=args.pruning_strategy,
+            provider_options=provider_options if provider_options else None,
         )
 
         print(f"Search results for '{query}' (top {len(results)}):")

packages/leann-core/src/leann/embedding_compute.py

@@ -740,7 +740,10 @@ def compute_embeddings_openai(
     print(f"len of texts: {len(texts)}")
 
     # Apply prompt template if provided
-    prompt_template = provider_options.get("prompt_template")
+    # Priority: build_prompt_template (new format) > prompt_template (old format)
+    prompt_template = provider_options.get("build_prompt_template") or provider_options.get(
+        "prompt_template"
+    )
 
     if prompt_template:
         logger.warning(f"Applying prompt template: '{prompt_template}'")
@@ -1031,7 +1034,10 @@ def compute_embeddings_ollama(
 
     # Apply prompt template if provided
     provider_options = provider_options or {}
-    prompt_template = provider_options.get("prompt_template")
+    # Priority: build_prompt_template (new format) > prompt_template (old format)
+    prompt_template = provider_options.get("build_prompt_template") or provider_options.get(
+        "prompt_template"
+    )
 
     if prompt_template:
         logger.warning(f"Applying prompt template: '{prompt_template}'")

packages/leann-core/src/leann/searcher_base.py

@@ -71,14 +71,23 @@ def _ensure_server_running(self, passages_source_file: str, port: int, **kwargs)
             or "mips"
         )
 
+        # Filter out ALL prompt templates from provider_options during search
+        # Templates are applied in compute_query_embedding (line 109-110) BEFORE server call
+        # The server should never apply templates during search to avoid double-templating
+        search_provider_options = {
+            k: v
+            for k, v in self.embedding_options.items()
+            if k not in ("build_prompt_template", "query_prompt_template", "prompt_template")
+        }
+
         server_started, actual_port = self.embedding_server_manager.start_server(
             port=port,
             model_name=self.embedding_model,
             embedding_mode=self.embedding_mode,
             passages_file=passages_source_file,
             distance_metric=distance_metric,
             enable_warmup=kwargs.get("enable_warmup", False),
-            provider_options=self.embedding_options,
+            provider_options=search_provider_options,
         )
         if not server_started:
             raise RuntimeError(f"Failed to start embedding server on port {actual_port}")