Function extract_python_code

applications/camel/eval/code_extractor.py:7–171

Extract Python code from text. Priority extraction: 1. Code block containing entry_point function 2. Python code in markdown code blocks 3. Directly included Python code Args: text: Text containing code. entry_point: Function entry point name (optio

(text: str, entry_point: Optional[str] = None)

def extract_python_code(text: str, entry_point: Optional[str] = None) -> str:
    """
    Extract Python code from text.
    
    Priority extraction:
    1. Code block containing entry_point function
    2. Python code in markdown code blocks
    3. Directly included Python code
    
    Args:
        text: Text containing code.
        entry_point: Function entry point name (optional, for locating target function).
    
    Returns:
        Extracted Python code string.
    """
    if not text:
        return ""
    
    text = text.strip()
    
    # If entry_point exists, try to extract code containing that function
    if entry_point:
        # Try to match code block containing entry_point
        pattern = rf'(def\s+{re.escape(entry_point)}\s*\([^)]*\):.*?)(?=\n\ndef\s+|\nclass\s+|$)'
        match = re.search(pattern, text, re.DOTALL)
        if match:
            code = match.group(1).strip()
            # Ensure complete function body is included
            if code.count('def') == 1:
                # Try to find function end position (next def or class or end of file)
                lines = text.split('\n')
                start_idx = text.find(code)
                if start_idx != -1:
                    start_line = text[:start_idx].count('\n')
                    # Find function definition line
                    for i, line in enumerate(lines[start_line:], start_line):
                        if line.strip().startswith(f'def {entry_point}'):
                            # Found function definition, extract to next def/class or end of file
                            code_lines = []
                            indent_level = None
                            for j in range(i, len(lines)):
                                current_line = lines[j]
                                if j == i:
                                    code_lines.append(current_line)
                                    # Determine function body indentation level
                                    if ':' in current_line:
                                        indent_level = len(current_line) - len(current_line.lstrip()) + 4
                                elif indent_level is not None:
                                    # Check if it's part of function body
                                    if current_line.strip() == '':
                                        code_lines.append(current_line)
                                    elif current_line.strip().startswith('#'):
                                        code_lines.append(current_line)
                                    elif len(current_line) - len(current_line.lstrip()) >= indent_level:
                                        code_lines.append(current_line)
                                    elif current_line.strip().startswith(('def ', 'class ')):
                                        break