ragit 0.7.1__tar.gz → 0.7.3__tar.gz

{ragit-0.7.1 → ragit-0.7.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ragit
-Version: 0.7.1
+Version: 0.7.3
 Summary: Automatic RAG Pattern Optimization Engine
 Author: RODMENA LIMITED
 Maintainer-email: RODMENA LIMITED <info@rodmena.co.uk>
@@ -37,6 +37,10 @@ Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Requires-Dist: pytest-cov; extra == "test"
 Requires-Dist: pytest-mock; extra == "test"
+Provides-Extra: docs
+Requires-Dist: sphinx>=7.0; extra == "docs"
+Requires-Dist: sphinx-rtd-theme>=2.0; extra == "docs"
+Requires-Dist: sphinx-copybutton>=0.5; extra == "docs"
 Dynamic: license-file
 
 # ragit

{ragit-0.7.1 → ragit-0.7.3}/pyproject.toml

@@ -57,6 +57,12 @@ dev = [
 
 test = ["pytest", "pytest-cov", "pytest-mock"]
 
+docs = [
+    "sphinx>=7.0",
+    "sphinx-rtd-theme>=2.0",
+    "sphinx-copybutton>=0.5",
+]
+
 [tool.setuptools.dynamic]
 version = { attr = "ragit.version.__version__" }
 

{ragit-0.7.1 → ragit-0.7.3}/ragit/providers/ollama.py

@@ -56,6 +56,9 @@ class OllamaProvider(BaseLLMProvider, BaseEmbeddingProvider):
         "qwen3-embedding:8b": 4096,
     }
 
+    # Max characters per embedding request (safe limit for 512 token models)
+    MAX_EMBED_CHARS = 1500
+
     def __init__(
         self,
         base_url: str | None = None,
@@ -162,17 +165,21 @@ class OllamaProvider(BaseLLMProvider, BaseEmbeddingProvider):
         self._current_embed_model = model
         self._current_dimensions = self.EMBEDDING_DIMENSIONS.get(model, 768)
 
+        # Truncate oversized inputs to prevent context length errors
+        if len(text) > self.MAX_EMBED_CHARS:
+            text = text[: self.MAX_EMBED_CHARS]
+
         try:
             response = requests.post(
-                f"{self.embedding_url}/api/embed",
+                f"{self.embedding_url}/api/embeddings",
                 headers=self._get_headers(include_auth=False),
-                json={"model": model, "input": text},
+                json={"model": model, "prompt": text},
                 timeout=self.timeout,
             )
             response.raise_for_status()
             data = response.json()
 
-            embedding = data.get("embeddings", [[]])[0]
+            embedding = data.get("embedding", [])
             if not embedding:
                 raise ValueError("Empty embedding returned from Ollama")
 
@@ -189,33 +196,42 @@ class OllamaProvider(BaseLLMProvider, BaseEmbeddingProvider):
             raise ConnectionError(f"Ollama embed failed: {e}") from e
 
     def embed_batch(self, texts: list[str], model: str) -> list[EmbeddingResponse]:
-        """Generate embeddings for multiple texts (uses embedding_url, no auth for local)."""
+        """Generate embeddings for multiple texts (uses embedding_url, no auth for local).
+
+        Note: Ollama /api/embeddings only supports single prompts, so we loop.
+        """
         self._current_embed_model = model
         self._current_dimensions = self.EMBEDDING_DIMENSIONS.get(model, 768)
 
+        results = []
         try:
-            response = requests.post(
-                f"{self.embedding_url}/api/embed",
-                headers=self._get_headers(include_auth=False),
-                json={"model": model, "input": texts},
-                timeout=self.timeout,
-            )
-            response.raise_for_status()
-            data = response.json()
-
-            embeddings = data.get("embeddings", [])
-            if embeddings:
-                self._current_dimensions = len(embeddings[0])
-
-            return [
-                EmbeddingResponse(
-                    embedding=tuple(emb),
-                    model=model,
-                    provider=self.provider_name,
-                    dimensions=len(emb),
+            for text in texts:
+                # Truncate oversized inputs to prevent context length errors
+                if len(text) > self.MAX_EMBED_CHARS:
+                    text = text[: self.MAX_EMBED_CHARS]
+
+                response = requests.post(
+                    f"{self.embedding_url}/api/embeddings",
+                    headers=self._get_headers(include_auth=False),
+                    json={"model": model, "prompt": text},
+                    timeout=self.timeout,
+                )
+                response.raise_for_status()
+                data = response.json()
+
+                embedding = data.get("embedding", [])
+                if embedding:
+                    self._current_dimensions = len(embedding)
+
+                results.append(
+                    EmbeddingResponse(
+                        embedding=tuple(embedding),
+                        model=model,
+                        provider=self.provider_name,
+                        dimensions=len(embedding),
+                    )
                 )
-                for emb in embeddings
-            ]
+            return results
         except requests.RequestException as e:
             raise ConnectionError(f"Ollama batch embed failed: {e}") from e
 

{ragit-0.7.1 → ragit-0.7.3}/ragit/version.py

@@ -2,4 +2,4 @@
 # Copyright RODMENA LIMITED 2025
 # SPDX-License-Identifier: Apache-2.0
 #
-__version__ = "0.7.1"
+__version__ = "0.7.3"

{ragit-0.7.1 → ragit-0.7.3}/ragit.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ragit
-Version: 0.7.1
+Version: 0.7.3
 Summary: Automatic RAG Pattern Optimization Engine
 Author: RODMENA LIMITED
 Maintainer-email: RODMENA LIMITED <info@rodmena.co.uk>
@@ -37,6 +37,10 @@ Provides-Extra: test
 Requires-Dist: pytest; extra == "test"
 Requires-Dist: pytest-cov; extra == "test"
 Requires-Dist: pytest-mock; extra == "test"
+Provides-Extra: docs
+Requires-Dist: sphinx>=7.0; extra == "docs"
+Requires-Dist: sphinx-rtd-theme>=2.0; extra == "docs"
+Requires-Dist: sphinx-copybutton>=0.5; extra == "docs"
 Dynamic: license-file
 
 # ragit

{ragit-0.7.1 → ragit-0.7.3}/ragit.egg-info/requires.txt

@@ -14,6 +14,11 @@ issuedb[web]
 ruff
 mypy
 
+[docs]
+sphinx>=7.0
+sphinx-rtd-theme>=2.0
+sphinx-copybutton>=0.5
+
 [test]
 pytest
 pytest-cov