Skip to content

Fix IndexStats returning attribute names instead of values #1095

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
62 changes: 61 additions & 1 deletion meilisearch/models/index.py
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@ def __init__(self, doc: Dict[str, Any]) -> None:

def __getattr__(self, attr: str) -> Any:
if attr in self.__dict.keys():
return attr
return self.__dict[attr]
raise AttributeError(f"{self.__class__.__name__} object has no attribute {attr}")

def __iter__(self) -> Iterator:
Expand Down Expand Up @@ -62,3 +62,63 @@ class EmbedderDistribution(CamelBase):
class LocalizedAttributes(CamelBase):
attribute_patterns: List[str]
locales: List[str]


class OpenAiEmbedder(CamelBase):
source: str = "openAi"
url: Optional[str] = None
model: Optional[str] = None # Defaults to text-embedding-3-small
dimensions: Optional[int] = None # Uses the model default
api_key: Optional[str] = None # Can be provided through a CLI option or environment variable
document_template: Optional[str] = None
document_template_max_bytes: Optional[int] = None # Default to 400
distribution: Optional[EmbedderDistribution] = None
binary_quantized: Optional[bool] = None


class HuggingFaceEmbedder(CamelBase):
source: str = "huggingFace"
model: Optional[str] = None # Defaults to BAAI/bge-base-en-v1.5
revision: Optional[str] = None
document_template: Optional[str] = None
document_template_max_bytes: Optional[int] = None # Default to 400
distribution: Optional[EmbedderDistribution] = None
binary_quantized: Optional[bool] = None


class OllamaEmbedder(CamelBase):
source: str = "ollama"
url: Optional[str] = None
api_key: Optional[str] = None
model: str
document_template: Optional[str] = None
document_template_max_bytes: Optional[int] = None # Default to 400
distribution: Optional[EmbedderDistribution] = None
binary_quantized: Optional[bool] = None


class RestEmbedder(CamelBase):
source: str = "rest"
url: str
api_key: Optional[str] # required for protected APIs
document_template: Optional[str] = None
document_template_max_bytes: Optional[int] = None # Default to 400
request: Dict[str, Any]
response: Dict[str, Any]
distribution: Optional[EmbedderDistribution] = None
headers: Optional[Dict[str, Any]] = None
binary_quantized: Optional[bool] = None


class UserProvidedEmbedder(CamelBase):
source: str = "userProvided"
dimensions: int
distribution: Optional[EmbedderDistribution] = None
binary_quantized: Optional[bool] = None


class Embedders(CamelBase):
embedders: Dict[
str,
Union[OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder, UserProvidedEmbedder],
]
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Verification agent

🧩 Analysis chain

Potential code duplication with existing embedder classes.

The embedder classes defined here (OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder, UserProvidedEmbedder, Embedders) appear to duplicate functionality that already exists in meilisearch/models/embedders.py based on the relevant code snippets provided. This could lead to maintenance issues and confusion about which classes to use.

Please verify if these classes are intended to replace the existing ones or if there's a different purpose for having them in both locations:


🏁 Script executed:

#!/bin/bash
# Description: Check for duplicate embedder class definitions across the codebase
# Expected: Find if these classes exist in multiple files

echo "Searching for embedder class definitions..."
rg -A 2 "class.*Embedder.*CamelBase" 

echo -e "\nSearching for Embedders container class..."
rg -A 2 "class Embedders"

Length of output: 2856


Consolidate duplicate embedder definitions
The OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder, UserProvidedEmbedder and Embedders classes live in both:

  • meilisearch/models/index.py (lines 67–124)
  • meilisearch/models/embedders.py

Maintaining two copies will lead to drift and confusion. Please remove or de-duplicate one set of definitions—ideally centralizing all embedder models in meilisearch/models/embedders.py (or vice versa)—and update any imports accordingly.

🧰 Tools
🪛 Ruff (0.11.9)

123-123: Undefined name Union

(F821)

🤖 Prompt for AI Agents
In meilisearch/models/index.py around lines 67 to 124, the embedder classes
OpenAiEmbedder, HuggingFaceEmbedder, OllamaEmbedder, RestEmbedder,
UserProvidedEmbedder, and Embedders are duplicated from
meilisearch/models/embedders.py. To fix this, remove these class definitions
from index.py and update all imports in the codebase to reference the
centralized definitions in embedders.py, ensuring no duplicate embedder classes
remain.

2 changes: 1 addition & 1 deletion tests/models/test_index.py
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@

def test_getattr():
document = IndexStats({"field1": "test 1", "fiels2": "test 2"})
assert document.__getattr__("field1") == "field1"
assert document.__getattr__("field1") == "test 1"


def test_getattr_not_found():
Expand Down