auto-coder 0.1.315__py3-none-any.whl → 0.1.317__py3-none-any.whl

autocoder/common/v2/code_editblock_manager.py

@@ -0,0 +1,275 @@
+from typing import List, Dict, Tuple, Optional, Any
+import os
+import json
+import time
+from concurrent.futures import ThreadPoolExecutor
+
+import byzerllm
+from byzerllm.utils.client import code_utils
+
+from autocoder.common.types import Mode, CodeGenerateResult, MergeCodeWithoutEffect
+from autocoder.common import AutoCoderArgs, git_utils, SourceCodeList
+from autocoder.common import sys_prompt
+from autocoder.privacy.model_filter import ModelPathFilter
+from autocoder.common.utils_code_auto_generate import chat_with_continue, stream_chat_with_continue, ChatWithContinueResult
+from autocoder.utils.auto_coder_utils.chat_stream_out import stream_out
+from autocoder.common.stream_out_type import CodeGenerateStreamOutType
+from autocoder.common.auto_coder_lang import get_message_with_format
+from autocoder.common.printer import Printer
+from autocoder.rag.token_counter import count_tokens
+from autocoder.utils import llms as llm_utils
+from autocoder.memory.active_context_manager import ActiveContextManager
+from autocoder.common.v2.code_auto_generate_editblock import CodeAutoGenerateEditBlock
+from autocoder.common.v2.code_auto_merge_editblock import CodeAutoMergeEditBlock
+from autocoder.shadows.shadow_manager import ShadowManager
+from autocoder.linters.shadow_linter import ShadowLinter
+from autocoder.linters.models import IssueSeverity
+from loguru import logger
+from autocoder.common.global_cancel import global_cancel
+from autocoder.linters.models import ProjectLintResult
+from autocoder.common.token_cost_caculate import TokenCostCalculator
+
+
+class CodeEditBlockManager:
+    """
+    A class that combines code generation, linting, and merging with automatic error correction.
+    It generates code, lints it, and if there are errors, regenerates the code up to 5 times
+    before merging the final result.
+    """
+    
+    def __init__(
+        self,
+        llm: byzerllm.ByzerLLM,
+        args: AutoCoderArgs,
+        action=None,
+        fence_0: str = "```",
+        fence_1: str = "```",
+    ) -> None:
+        self.llm = llm
+        self.args = args
+        self.action = action
+        self.fence_0 = fence_0
+        self.fence_1 = fence_1
+        self.generate_times_same_model = args.generate_times_same_model
+        self.max_correction_attempts = args.auto_fix_lint_max_attempts
+        self.printer = Printer()
+        
+        # Initialize sub-components
+        self.code_generator = CodeAutoGenerateEditBlock(llm, args, action, fence_0, fence_1)
+        self.code_merger = CodeAutoMergeEditBlock(llm, args, fence_0, fence_1)
+        
+        # Create shadow manager for linting
+        self.shadow_manager = ShadowManager(args.source_dir)
+        self.shadow_linter = ShadowLinter(self.shadow_manager, verbose=False)
+
+    @byzerllm.prompt()
+    def fix_linter_errors(self, query: str, lint_issues: str) -> str:
+        """        
+        Linter 检测到的问题:
+        <lint_issues>
+        {{ lint_issues }}
+        </lint_issues>
+        
+        用户原始需求:
+        <user_query_wrapper>
+        {{ query }}
+        </user_query_wrapper>
+
+        修复上述问题，请确保代码质量问题被解决，同时保持代码的原有功能。
+        请严格遵守*SEARCH/REPLACE block*的格式。
+        """
+        
+    def _create_shadow_files_from_edits(self, generation_result: CodeGenerateResult) -> Dict[str, str]:
+        """
+        从编辑块内容中提取代码并创建临时影子文件用于检查。
+        
+        参数:
+            generation_result (CodeGenerateResult): 包含SEARCH/REPLACE块的内容
+            
+        返回:
+            Dict[str, str]: 映射 {影子文件路径: 内容}
+        """
+        result = self.code_merger.choose_best_choice(generation_result)
+        merge = self.code_merger._merge_code_without_effect(result.contents[0])        
+        shadow_files = {}
+        for file_path, new_content in merge.success_blocks:
+            self.shadow_manager.update_file(file_path, new_content)
+            shadow_files[self.shadow_manager.to_shadow_path(file_path)] = new_content
+
+        return shadow_files
+
+    def _format_lint_issues(self, lint_results:ProjectLintResult,level:IssueSeverity) -> str:
+        """
+        将linter结果格式化为字符串供模型使用
+        
+        参数:
+            lint_results: Linter结果对象
+            level: 过滤问题的级别
+            
+        返回:
+            str: 格式化的问题描述
+        """
+        formatted_issues = []
+        
+        for file_path, result in lint_results.file_results.items():                        
+            file_has_issues = False
+            file_issues = []
+            
+            for issue in result.issues:                
+                if issue.severity.value != level.value:
+                    continue
+                    
+                if not file_has_issues:
+                    file_has_issues = True
+                    file_issues.append(f"文件: {file_path}")
+                
+                severity = "错误" if issue.severity == IssueSeverity.ERROR else "警告" if issue.severity == IssueSeverity.WARNING else "信息"
+                line_info = f"第{issue.position.line}行"
+                if issue.position.column:
+                    line_info += f", 第{issue.position.column}列"
+                
+                file_issues.append(
+                    f"  - [{severity}] {line_info}: {issue.message} (规则: {issue.code})"
+                )
+            
+            if file_has_issues:
+                formatted_issues.extend(file_issues)
+                formatted_issues.append("")  # 空行分隔不同文件
+            
+        return "\n".join(formatted_issues)
+
+    def _count_errors(self, lint_results:ProjectLintResult) -> int:
+        """
+        计算lint结果中的错误数量
+        
+        参数:
+            lint_results: Linter结果对象
+            
+        返回:
+            int: 错误数量
+        """
+        error_count = 0
+        
+        for _, result in lint_results.file_results.items():                
+            error_count += result.error_count
+                
+        return error_count
+
+    def generate_and_fix(self, query: str, source_code_list: SourceCodeList) -> CodeGenerateResult:
+        """
+        生成代码，运行linter，修复错误，最多尝试指定次数
+        
+        参数:
+            query (str): 用户查询
+            source_code_list (SourceCodeList): 源代码列表
+            
+        返回:
+            CodeGenerateResult: 生成的代码结果
+        """
+        # 初始代码生成
+        self.printer.print_in_terminal("generating_initial_code")
+        start_time = time.time()
+        generation_result = self.code_generator.single_round_run(query, source_code_list)
+        
+        token_cost_calculator = TokenCostCalculator(args=self.args)
+        token_cost_calculator.track_token_usage_by_generate(
+            llm=self.llm,
+            generate=generation_result,
+            operation_name="code_generation_complete",
+            start_time=start_time,
+            end_time=time.time()
+        )
+        
+        # 确保结果非空
+        if not generation_result.contents:
+            self.printer.print_in_terminal("generation_failed", style="red")
+            return generation_result
+                            
+        
+        # 最多尝试修复5次
+        for attempt in range(self.max_correction_attempts):
+            global_cancel.check_and_raise()
+            # 代码生成结果更新到影子文件里去
+            shadow_files = self._create_shadow_files_from_edits(generation_result)
+            
+            if not shadow_files:
+                self.printer.print_in_terminal("no_files_to_lint", style="yellow")
+                break
+                
+            # 运行linter
+            lint_results = self.shadow_linter.lint_all_shadow_files()
+            error_count = self._count_errors(lint_results)
+            # print(f"error_count: {error_count}")
+            # print(f"lint_results: {json.dumps(lint_results.model_dump(), indent=4,ensure_ascii=False)}")            
+            
+            # 如果没有错误则完成
+            if error_count == 0:
+                self.printer.print_in_terminal("no_lint_errors_found", style="green")
+                break
+
+            # 格式化lint问题
+            formatted_issues = self._format_lint_issues(lint_results, IssueSeverity.ERROR)            
+
+            # 打印当前错误
+            self.printer.print_in_terminal(
+                "lint_attempt_status", 
+                style="yellow", 
+                attempt=(attempt + 1), 
+                max_correction_attempts=self.max_correction_attempts,
+                error_count=error_count,
+                formatted_issues=formatted_issues
+            )
+            
+            if attempt == self.max_correction_attempts - 1:
+                self.printer.print_in_terminal("max_attempts_reached", style="yellow")
+                break
+                
+            
+            
+            # 准备修复提示
+            fix_prompt = self.fix_linter_errors.prompt(
+                query=query,
+                lint_issues=formatted_issues
+            )
+            
+            # for source in source_code_list.sources:
+            #     print(f"file_path: {source.module_name}")                
+            # print(f"fix_prompt: {fix_prompt}")
+            
+            # 将 shadow_files 转化为 source_code_list
+            start_time = time.time()
+            source_code_list = self.code_merger.get_source_code_list_from_shadow_files(shadow_files)
+            generation_result = self.code_generator.single_round_run(fix_prompt, source_code_list) 
+            token_cost_calculator.track_token_usage_by_generate(
+                llm=self.llm,
+                generate=generation_result,
+                operation_name="code_generation_complete",
+                start_time=start_time,
+                end_time=time.time()
+            )           
+
+        
+        # 清理临时影子文件
+        self.shadow_manager.clean_shadows()
+        
+        # 返回最终结果
+        return generation_result
+
+    def run(self, query: str, source_code_list: SourceCodeList) -> CodeGenerateResult:
+        """
+        执行完整的代码生成、修复、合并流程
+        
+        参数:
+            query (str): 用户查询
+            source_code_list (SourceCodeList): 源代码列表
+            
+        返回:
+            CodeGenerateResult: 生成和修复的代码结果
+        """
+        # 生成代码并自动修复lint错误
+        generation_result = self.generate_and_fix(query, source_code_list)
+        global_cancel.check_and_raise()
+        # 合并代码        
+        self.code_merger.merge_code(generation_result)
+        
+        return generation_result 

autocoder/common/v2/code_manager.py

@@ -0,0 +1,238 @@
+from typing import List, Dict, Tuple, Optional, Any
+import os
+import json
+import time
+from concurrent.futures import ThreadPoolExecutor
+
+import byzerllm
+from byzerllm.utils.client import code_utils
+
+from autocoder.common.types import Mode, CodeGenerateResult, MergeCodeWithoutEffect
+from autocoder.common import AutoCoderArgs, git_utils, SourceCodeList
+from autocoder.common import sys_prompt
+from autocoder.privacy.model_filter import ModelPathFilter
+from autocoder.common.utils_code_auto_generate import chat_with_continue, stream_chat_with_continue, ChatWithContinueResult
+from autocoder.utils.auto_coder_utils.chat_stream_out import stream_out
+from autocoder.common.stream_out_type import CodeGenerateStreamOutType
+from autocoder.common.auto_coder_lang import get_message_with_format
+from autocoder.common.printer import Printer
+from autocoder.rag.token_counter import count_tokens
+from autocoder.utils import llms as llm_utils
+from autocoder.memory.active_context_manager import ActiveContextManager
+from autocoder.shadows.shadow_manager import ShadowManager
+from autocoder.linters.shadow_linter import ShadowLinter
+from autocoder.linters.models import IssueSeverity
+from loguru import logger
+from autocoder.common.global_cancel import global_cancel
+from autocoder.linters.models import ProjectLintResult
+from autocoder.common.token_cost_caculate import TokenCostCalculator
+
+
+class CodeManager:
+    """
+    Base class for code generation managers that combines code generation, linting, and merging with automatic error correction.
+    """
+    
+    def __init__(
+        self,
+        llm: byzerllm.ByzerLLM,
+        args: AutoCoderArgs,
+        action=None,
+    ) -> None:
+        self.llm = llm
+        self.args = args
+        self.action = action
+        self.generate_times_same_model = args.generate_times_same_model
+        self.max_correction_attempts = args.auto_fix_lint_max_attempts
+        self.printer = Printer()
+        
+        # Create shadow manager for linting
+        self.shadow_manager = ShadowManager(args.source_dir)
+        self.shadow_linter = ShadowLinter(self.shadow_manager, verbose=False)
+
+    @byzerllm.prompt()
+    def fix_linter_errors(self, query: str, lint_issues: str) -> str:
+        """        
+        Linter 检测到的问题:
+        <lint_issues>
+        {{ lint_issues }}
+        </lint_issues>
+        
+        用户原始需求:
+        <user_query_wrapper>
+        {{ query }}
+        </user_query_wrapper>
+
+        修复上述问题，请确保代码质量问题被解决，同时保持代码的原有功能。
+        """
+        
+    def _format_lint_issues(self, lint_results:ProjectLintResult,level:IssueSeverity) -> str:
+        """
+        将linter结果格式化为字符串供模型使用
+        
+        参数:
+            lint_results: Linter结果对象
+            level: 过滤问题的级别
+            
+        返回:
+            str: 格式化的问题描述
+        """
+        formatted_issues = []
+        
+        for file_path, result in lint_results.file_results.items():                        
+            file_has_issues = False
+            file_issues = []
+            
+            for issue in result.issues:                
+                if issue.severity.value != level.value:
+                    continue
+                    
+                if not file_has_issues:
+                    file_has_issues = True
+                    file_issues.append(f"文件: {file_path}")
+                
+                severity = "错误" if issue.severity == IssueSeverity.ERROR else "警告" if issue.severity == IssueSeverity.WARNING else "信息"
+                line_info = f"第{issue.position.line}行"
+                if issue.position.column:
+                    line_info += f", 第{issue.position.column}列"
+                
+                file_issues.append(
+                    f"  - [{severity}] {line_info}: {issue.message} (规则: {issue.code})"
+                )
+            
+            if file_has_issues:
+                formatted_issues.extend(file_issues)
+                formatted_issues.append("")  # 空行分隔不同文件
+            
+        return "\n".join(formatted_issues)
+
+    def _count_errors(self, lint_results:ProjectLintResult) -> int:
+        """
+        计算lint结果中的错误数量
+        
+        参数:
+            lint_results: Linter结果对象
+            
+        返回:
+            int: 错误数量
+        """
+        error_count = 0
+        
+        for _, result in lint_results.file_results.items():                
+            error_count += result.error_count
+                
+        return error_count
+
+    def generate_and_fix(self, query: str, source_code_list: SourceCodeList) -> CodeGenerateResult:
+        """
+        生成代码，运行linter，修复错误，最多尝试指定次数
+        
+        参数:
+            query (str): 用户查询
+            source_code_list (SourceCodeList): 源代码列表
+            
+        返回:
+            CodeGenerateResult: 生成的代码结果
+        """
+        # 初始代码生成
+        self.printer.print_in_terminal("generating_initial_code")
+        start_time = time.time()
+        generation_result = self.code_generator.single_round_run(query, source_code_list)
+        
+        token_cost_calculator = TokenCostCalculator(args=self.args)
+        token_cost_calculator.track_token_usage_by_generate(
+            llm=self.llm,
+            generate=generation_result,
+            operation_name="code_generation_complete",
+            start_time=start_time,
+            end_time=time.time()
+        )
+        
+        # 确保结果非空
+        if not generation_result.contents:
+            self.printer.print_in_terminal("generation_failed", style="red")
+            return generation_result
+                            
+        
+        # 最多尝试修复5次
+        for attempt in range(self.max_correction_attempts):
+            global_cancel.check_and_raise()
+            # 代码生成结果更新到影子文件里去
+            shadow_files = self._create_shadow_files_from_edits(generation_result)
+            
+            if not shadow_files:
+                self.printer.print_in_terminal("no_files_to_lint", style="yellow")
+                break
+                
+            # 运行linter
+            lint_results = self.shadow_linter.lint_all_shadow_files()
+            error_count = self._count_errors(lint_results)
+            
+            # 如果没有错误则完成
+            if error_count == 0:
+                self.printer.print_in_terminal("no_lint_errors_found", style="green")
+                break
+
+            # 格式化lint问题
+            formatted_issues = self._format_lint_issues(lint_results, IssueSeverity.ERROR)            
+
+            # 打印当前错误
+            self.printer.print_in_terminal(
+                "lint_attempt_status", 
+                style="yellow", 
+                attempt=(attempt + 1), 
+                max_correction_attempts=self.max_correction_attempts,
+                error_count=error_count,
+                formatted_issues=formatted_issues
+            )
+            
+            if attempt == self.max_correction_attempts - 1:
+                self.printer.print_in_terminal("max_attempts_reached", style="yellow")
+                break
+                
+            # 准备修复提示
+            fix_prompt = self.fix_linter_errors.prompt(
+                query=query,
+                lint_issues=formatted_issues
+            )
+            
+            for source in source_code_list.sources:
+                print(f"file_path: {source.module_name}")                
+            print(f"fix_prompt: {fix_prompt}")
+            
+            # 将 shadow_files 转化为 source_code_list
+            source_code_list = self.code_merger.get_source_code_list_from_shadow_files(shadow_files)
+            start_time = time.time()
+            generation_result = self.code_generator.single_round_run(fix_prompt, source_code_list)            
+            token_cost_calculator.track_token_usage_by_generate(
+                llm=self.llm,
+                generate=generation_result,
+                operation_name="code_generation_complete",
+                start_time=start_time,
+                end_time=time.time()
+            )
+        
+        # 清理临时影子文件
+        self.shadow_manager.clean_shadows()
+        
+        # 返回最终结果
+        return generation_result
+
+    def run(self, query: str, source_code_list: SourceCodeList) -> CodeGenerateResult:
+        """
+        执行完整的代码生成、修复、合并流程
+        
+        参数:
+            query (str): 用户查询
+            source_code_list (SourceCodeList): 源代码列表
+            
+        返回:
+            CodeGenerateResult: 生成和修复的代码结果
+        """
+        # 生成代码并自动修复lint错误
+        generation_result = self.generate_and_fix(query, source_code_list)
+        global_cancel.check_and_raise()
+        # 合并代码        
+        self.code_merger.merge_code(generation_result)
+        
+        return generation_result