Resolve issue patchflow

[CTY-git]

No description provided.

[codelion]

fixes #35

[patched-admin]

The code changes in this update include the addition of a new step to propose relevant files to issues in the Patchwork project workflow using Poetry, an update to the .gitignore file to exclude certain file types and folders, import statement changes for run_chat function, method additions in scm.py for issue extraction and commenting, new module imports and function definitions in utils.py, the introduction of new classes for resolving and creating issue comments, restructuring import statements, and the creation of new modules for embedding generation, embedding query, and reading issues. Additionally, changes were made to the pyproject.toml file updating the patchwork-cli project version and dependencies.

---

• File changed: .github/workflows/test.yml
  A new step Propose relevant file to issues was added to the workflow which runs a command using Poetry to propose a relevant file to issues, specifically for the Patchwork project by resolving issue number 20 on GitHub.

• File changed: .gitignore
  The diff in the .gitignore file adds the exclusion of files with the .iml extension in the .idea folder and also excludes the benchmark folder.

• File changed: patchwork/app.py
  The code is updated to import the function run_chat from patchwork.patchwork_interpreter instead of patchwork.interpreter, which suggests a restructuring or renaming of the interpreter file.

• File changed: patchwork/common/client/scm.py
  The diff in the file scm.py includes modifications to introduce methods for extracting slug and issue IDs from URLs, finding issues by URL or ID, and creating issue comments in both GitHub and GitLab clients in a Python application.

• File changed: patchwork/common/utils.py
  Added import statements for new modules tiktoken, chromadb, and related functions in the utils.py file. Defined new functions for working with openai and huggingface models, as well as a generic function to get an embedding function based on provided inputs in the patchwork project.

• File changed: patchwork/patchflows/ResolveIssue/ResolveIssue.py
  The diff introduces a new Python file named ResolveIssue.py that defines a class ResolveIssue inheriting from Step. This class reads inputs from a defaults.yml file, updates the inputs, runs several steps, modifies the inputs during execution, and finally returns the updated inputs after running those steps.

• File changed: patchwork/patchflows/ResolveIssue/init.py
  There are no changes to the __init__.py file in the ResolveIssue module under patchflows in the patchwork repository.

• File changed: patchwork/patchflows/ResolveIssue/defaults.yml
  The diff adds inputs for ReadIssues, including options for github_api_key and gitlab_api_key, as well as inputs for GenerateEmbeddings for choosing between OpenAI and HuggingFace models and providing the appropriate API keys.

• File changed: patchwork/patchflows/init.py
  A new import statement for ResolveIssue module was added and the all list was updated to include ResolveIssue, ensuring that the ResolveIssue module is exposed for use outside the init.py file.

• File changed: patchwork/patchwork_interpreter.py
  There are no changes in the patchwork_interpreter.py file.

• File changed: patchwork/steps/CreateIssueComment/CreateIssueComment.py
  This diff introduces a new class named CreateIssueComment that extends a Step class and is responsible for creating comments on issues. The class requires specific input keys such as "issue_url" and "issue_text" and can interact with SCM platforms like GitHub and GitLab based on provided API keys. The run method of the class fetches the issue slug and ID from the URL, creates an issue comment using the SCM client, and returns the URL of the created issue comment.

• File changed: patchwork/steps/CreateIssueComment/README.md
  This diff adds documentation for the CreateIssueComment step, outlining the inputs required (such as issue URL, issue text, API keys) and the outputs (like the run() method that creates a comment on a specified issue and returns the comment URL.

• File changed: patchwork/steps/CreateIssueComment/init.py
  There are no changes in the __init__.py file of the CreateIssueComment module in the patchwork/steps directory.

• File changed: patchwork/steps/ExtractCode/ExtractCode.py
  The diff involves restructuring the import statements in the ExtractCode.py file by removing unnecessary imports and rearranging the required ones. Additionally, it changes the function count_tokens to count_openai_tokens and updates its usage within the code.

• File changed: patchwork/steps/GenerateCodeRepositoryEmbeddings/GenerateCodeRepositoryEmbeddings.py
  The given diff adds a new Python script called GenerateCodeRepositoryEmbeddings.py, which includes functions to filter files, batch operations, hash text, and a class GenerateCodeRepositoryEmbeddings that gathers and processes code repository data for embedding generation. The script also uses external libraries like git and chromadb for functionalities like version control and database interactions.

• File changed: patchwork/steps/GenerateCodeRepositoryEmbeddings/README.md
  The diff shows the addition of a README.md file explaining the functionality of a code documentation related to generating embeddings for code files in a repository. It describes the inputs, process of generating embeddings, functions related to file filtering, hashing text content, managing the code repository, and producing results for further processing with the GenerateEmbeddings class.

• File changed: patchwork/steps/GenerateCodeRepositoryEmbeddings/init.py
  No changes were made to the file 'patchwork/steps/GenerateCodeRepositoryEmbeddings/init.py'.

• File changed: patchwork/steps/GenerateEmbeddings/GenerateEmbeddings.py
  This diff introduces a new Python file "GenerateEmbeddings.py" to the path "patchwork/steps/GenerateEmbeddings/", which contains code for generating embeddings, applying split_text function, and running the GenerateEmbeddings class that manages embedding creation procedures using embedded documents.

• File changed: patchwork/steps/GenerateEmbeddings/README.md
  The added content provides a summary of the GenerateEmbeddings.py script, detailing its inputs, outputs, and usage instructions. It explains the functions and class structure, highlighting the steps for using GenerateEmbeddings to generate document embeddings and store them in a database collection.

• File changed: patchwork/steps/GenerateEmbeddings/init.py
  No changes were made in the file __init__.py within the GenerateEmbeddings module in the patchwork repository under the steps directory.

• File changed: patchwork/steps/PreparePrompt/PreparePrompt.py
  The diff in the PreparePrompt.py file involves changing the way a dictionary is defined for the required_keys attribute in the PreparePrompt class, from using string formatting to directly referencing the keys.

• File changed: patchwork/steps/QueryEmbeddings/QueryEmbeddings.py
  The diff shows the addition of a new class QueryEmbeddings within the QueryEmbeddings.py file, which contains methods to fetch embeddings for query texts. It includes methods to get embedding functions, run queries against a collection, and return results sorted by distance.

• File changed: patchwork/steps/QueryEmbeddings/README.md
  This diff adds a documentation section for a Python module QueryEmbeddings.py, outlining its inputs, outputs, and usage. It describes the module's attributes, methods, classes, and functionality of querying embeddings for a list of texts, along with details on the expected inputs and the output structure.

• File changed: patchwork/steps/QueryEmbeddings/init.py
  No changes were made to the file __init__.py in the QueryEmbeddings module of the patchwork project.

• File changed: patchwork/steps/ReadIssues/README.md
  This diff introduces documentation for a Patchwork module called ReadIssues, outlining the inputs, outputs, and usage of the module. The ReadIssues class expects inputs in a dictionary format, interacts with source code management platforms like GitHub or GitLab, and returns the issue text associated with a provided URL. The documentation also explains how users can utilize the ReadIssues class to retrieve issue text data.

• File changed: patchwork/steps/ReadIssues/ReadIssues.py
  A new class ReadIssues is added that is a subclass of Step. The class requires issue_url as a key in the input dictionary and initializes a GithubClient or GitlabClient based on the presence of github_api_key or gitlab_api_key in the inputs. It then fetches and returns the issue text by calling find_issue_by_url from the selected ScmPlatformClientProtocol implementation.

• File changed: patchwork/steps/ReadIssues/init.py
  There are no changes in the file __init__.py within the ReadIssues module under the patchwork/steps directory.

• File changed: patchwork/steps/init.py
  A new module called CreateIssueComment has been added to the patchwork.steps package. Additional modules called GenerateCodeRepositoryEmbeddings, QueryEmbeddings, and ReadIssues have also been included in the package.

• File changed: pyproject.toml
  The version for the "patchwork-cli" project in the pyproject.toml file has been updated from 0.0.6 to 0.0.7. Additionally, there were changes made to the dependencies section including updates to various packages such as owasp-depscan, pygithub, gitpython, pydantic, tiktoken, libcst, open-interpreter, chromadb, sentence-transformers, and the addition of datasets.