Implement Customize Build Commands and Test Commands

[dcloud347]

This pull request makes several significant changes across the codebase, primarily focused on removing the build command functionality from the issue handling workflow, improving error handling and validation, and refining context extraction and test case retrieval processes. The most notable changes are grouped below by theme.

Removal of Build Command Support:

• Removed all references to build_commands and passed_build from the issue processing pipeline, including the API layer (issue.py), service layer (issue_service.py), models (issue.py), state definitions (issue_state.py), and graph/subgraph nodes (issue_graph.py, issue_bug_subgraph_node.py). This includes removing build command validation, result reporting, and related error messaging. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

Validation and Error Handling Improvements:

• Added validation to ensure that build and test commands are provided when required by the issue request, returning clear error messages if missing.
• Improved exception handling in the context retrieval subgraph node to log and propagate recursion errors more transparently. [1] [2]

Container and Command Execution Enhancements:

• Refactored the container command execution logic to use a more robust command invocation format and replaced the restart_container method with a reset_repository method for cleaning the git state inside containers. [1] [2]
• Updated the file update method in the container to clarify parameter names and documentation.

Context Extraction and Test Case Retrieval Refinements:

• Improved the format of the human message for context extraction by clearly delimiting the original query and context sections, and refactored the code for clarity and robustness. [1] [2] [3]
• Increased the number of relevant test cases to retrieve from 3 to 5 in bug reproduction context queries, updating prompts and documentation accordingly. [1] [2] [3]

Type Annotations and Code Quality:

• Improved type annotations in several nodes and service methods, replacing generic Dict types with more specific types such as IssueBugState for better type safety and readability. [1] [2] [3]

These changes collectively simplify the issue processing workflow, enhance validation and error reporting, and improve the clarity and robustness of context and test case handling.