Typing and bug squashes #1764
Merged
Typing and bug squashes #1764
+154
−115
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ieaves
requested review from
rhatdan,
ericcurtin,
bmahabirbu,
maxamillion,
swarajpande5,
jhjaggars,
cgruver and
engelmi
as code owners
Reviewer's GuideThis PR systematically enhances type safety and enforces interface contracts while squashing minor bugs and normalizing code style across the codebase. Class diagram for ModelBase and Model with typing and ABC changesclassDiagram
class ModelBase {
<<abstract>>
str model
str type
__not_implemented_error(param)
pull(args)
push(source_model, args)
<<abstract>> remove(args)
<<abstract>> bench(args)
<<abstract>> run(args)
<<abstract>> perplexity(args)
<<abstract>> serve(args)
<<abstract>> exists() bool
<<abstract>> inspect(args)
}
class Model {
str model
str type = "Model"
str directory
str filename
str _model_name
str _model_tag
...
Model(model: str, model_store_path: str)
}
Model --|> ModelBase
Class diagram for Rag class with improved typingclassDiagram
class Rag {
str model = ""
str target = ""
list~str~ urls = []
Rag(target: str)
build(source: str, target: str, args)
_handle_paths(path: str)
generate(args)
}
Class diagram for ModelFactory and New function with type aliasingclassDiagram
class ModelFactory {
str model
str store_path
str transport
bool ignore_stderr
ModelFactory(model: str, args: StoreArgType, transport: str, ignore_stderr: bool)
detect_model_model_type() Tuple[type[CLASS_MODEL_TYPES], Callable[[], CLASS_MODEL_TYPES]]
create()
create_huggingface()
create_ollama()
create_oci()
create_url()
create_modelscope()
}
class New {
<<function>>
New(name, args, transport: str | None = None) -> Union[Huggingface | ModelScope | Ollama | OCI | URL]
}
Class diagram for AccelType, GPUEnvVar, AccelEnvVar TypeAliasesclassDiagram
class AccelType {
<<typealias>>
Literal["asahi", "cuda", "cann", "hip", "intel", "musa"]
}
class GPUEnvVar {
<<typealias>>
Literal["ASAHI_VISIBLE_DEVICES", "ASCEND_VISIBLE_DEVICES", "CUDA_VISIBLE_DEVICES", "HIP_VISIBLE_DEVICES", "INTEL_VISIBLE_DEVICES", "MUSA_VISIBLE_DEVICES"]
}
class AccelEnvVar {
<<typealias>>
Literal["CUDA_LAUNCH_BLOCKING", "HSA_VISIBLE_DEVICES", "HSA_OVERRIDE_GFX_VERSION", "MUSA_VISIBLE_DEVICES"]
}
Class diagram for file loaders with updated load method signatureclassDiagram
class BaseFileManager {
<<abstract>>
_get_loader(file: str) base.BaseFileLoader
<<abstract>> load(*args, **kwargs)
supported_extensions()
}
class FileManager {
load(file_path: str) list~dict~
text_manager
image_manager
}
Class diagram for StoreFile, RefJSONFile, and StoreFileType with typing fixesclassDiagram
class StoreFileType {
GGUF_MODEL = "gguf"
MMPROJ = "mmproj"
CHAT_TEMPLATE = "chat_template"
}
class StoreFile {
hash: str
name: str
type: StoreFileType
}
class RefJSONFile {
hash: str
path: str
files: list~StoreFile~
}
Class diagram for updated config type aliasesclassDiagram
class PathStr {
<<typealias>>
str
}
class SUPPORTED_ENGINES {
<<typealias>>
Literal["podman", "docker"] | PathStr
}
class SUPPORTED_RUNTIMES {
<<typealias>>
Literal["llama.cpp", "vllm", "mlx"]
}
class COLOR_OPTIONS {
<<typealias>>
Literal["auto", "always", "never"]
}
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Summary of Changes
Hello @ieaves, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request primarily focuses on a comprehensive effort to improve the codebase's type safety and address several identified bugs. The changes aim to make the system more robust, easier to understand, and more maintainable for future development. It also includes minor refactorings to enhance code clarity and consistency.
Highlights
- Enhanced Type Safety: Introduced extensive type hints across the codebase, utilizing
TypeAlias,TypedDict,Sequence, and explicit annotations for variables, function parameters, and return types. This significantly improves code readability, maintainability, and enables more robust static analysis. - Critical Bug Fixes: Addressed several bugs, including incorrect attribute access in GGUF model inspection, potential
Nonedereferences in model store operations, and missing validation for engine configuration, enhancing the stability and reliability of the application. - Code Structure and Readability Improvements: Refactored
ModelBaseinto an Abstract Base Class (ABC) with abstract methods to enforce consistent interfaces. Minor code style adjustments, such as using the walrus operator and reformatting argument lists, were also applied to improve overall readability and conciseness. - New Unit Test: Added a new unit test for
GGUFModelInfoserialization to ensure the correctness of recent type-related fixes in model inspection.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments or fill out our survey to provide feedback.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
Hey @ieaves - I've reviewed your changes - here's some feedback:
- ModelBase defines several interface methods that raise NotImplementedError but aren’t decorated with @AbstractMethod (e.g. pull, push); consider marking them abstract so subclasses must implement them.
- In check_nvidia you cast the result of zip(*parsed_lines) directly to list[list[str]], but zip yields tuples—unpack into indices, uuids and convert each to list explicitly to satisfy the declared type.
- GGUF parsing now stores Tensor.type as a string but gguf_info.serialize still expects an enum with .name; unify the Tensor.type field across parser and serializer to avoid attribute errors.
Prompt for AI Agents
Please address the comments from this code review:
## Overall Comments
- ModelBase defines several interface methods that raise NotImplementedError but aren’t decorated with @abstractmethod (e.g. pull, push); consider marking them abstract so subclasses must implement them.
- In check_nvidia you cast the result of zip(*parsed_lines) directly to list[list[str]], but zip yields tuples—unpack into indices, uuids and convert each to list explicitly to satisfy the declared type.
- GGUF parsing now stores Tensor.type as a string but gguf_info.serialize still expects an enum with .name; unify the Tensor.type field across parser and serializer to avoid attribute errors.
## Individual Comments
### Comment 1
<location> `ramalama/common.py:343` </location>
<code_context>
return None
- indices, uuids = zip(*parsed_lines) if parsed_lines else (tuple(), tuple())
+ indices, uuids = cast(list[list[str]], zip(*parsed_lines))
# Get the list of devices specified by CUDA_VISIBLE_DEVICES, if any
cuda_visible_devices = os.environ.get("CUDA_VISIBLE_DEVICES", "")
</code_context>
<issue_to_address>
Use of cast with zip may not be type safe.
Instead of casting, use map(list, zip(*parsed_lines)) to ensure type safety, as zip returns tuples, not lists.
</issue_to_address>
### Comment 2
<location> `ramalama/model_store/go2jinja.py:375` </location>
<code_context>
children=[],
artificial=True,
)
+ for_node.next = cast(Node, for_node.next)
for_node.next.prev = initial_set_node
for_node.next = initial_set_node
</code_context>
<issue_to_address>
Use of cast may hide potential NoneType errors.
Add a check or assertion to ensure for_node.next is not None before casting.
</issue_to_address>
### Comment 3
<location> `test/unit/test_model_store.py:1` </location>
<code_context>
+import pytest
+from ramalama.model_inspect.gguf_info import GGUFModelInfo
+from ramalama.model_inspect.base_info import Tensor
</code_context>
<issue_to_address>
No new or updated tests for recent changes in model_store/store.py and related files.
Please add or update tests to cover:
- get_ref_file returning None in update_ref_file, get_cached_files, and _download_snapshot_files
- Handling of HTTPStatus.NOT_FOUND in _download_snapshot_files when ref_file is None
- The new Sequence type annotations for update_snapshot and validate_snapshot_files
This will help ensure robustness and proper edge case handling.
</issue_to_address>Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.
test/unit/test_model_store.py
Outdated
Comment on lines
1
to
6
| import pytest | ||
|
|
||
| from ramalama.model_store.snapshot_file import SnapshotFile, SnapshotFileType, validate_snapshot_files | ||
| from ramalama.model_store.global_store import ModelFile | ||
| from ramalama.model_store.reffile import StoreFile, StoreFileType | ||
|
|
||
| chat_template = SnapshotFile(name="chat-template", hash="", header={}, type=SnapshotFileType.ChatTemplate, url="") | ||
| model_file = SnapshotFile(name="model", hash="", header={}, type=SnapshotFileType.Model, url="") |
There was a problem hiding this comment.
suggestion (testing): No new or updated tests for recent changes in model_store/store.py and related files.
Please add or update tests to cover:
- get_ref_file returning None in update_ref_file, get_cached_files, and _download_snapshot_files
- Handling of HTTPStatus.NOT_FOUND in _download_snapshot_files when ref_file is None
- The new Sequence type annotations for update_snapshot and validate_snapshot_files
This will help ensure robustness and proper edge case handling.
There was a problem hiding this comment.
Code Review
This pull request does a great job of improving type safety and fixing bugs. The introduction of TypedDict, TypeAlias, and converting ModelBase to an ABC are all excellent changes that enhance maintainability.
I've pointed out a few areas for improvement, mainly around strengthening type hints and handling potential None values more robustly to prevent runtime errors. Overall, this is a solid contribution.
ieaves
commented
Jul 29, 2025
engelmi
reviewed
Jul 30, 2025
|
Please squash your commits. |
Signed-off-by: Ian Eaves <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary by Sourcery
Introduce comprehensive static typing to core modules, convert ModelBase to an abstract ABC, tighten ModelFactory initialization, and resolve GGUF parsing bugs with accompanying tests
Bug Fixes:
Enhancements:
Tests: