SourceCodeExtractionTool/txt解析版/txt解析.py

import os
import shutil
import re
from datetime import datetime
import time


def find_latest_txt_file():
    """
    在当前运行目录下查找最新修改的 .txt 文件，排除 config.txt
    """
    cwd = os.getcwd()
    txt_files = [
        os.path.join(cwd, f)
        for f in os.listdir(cwd)
        if f.lower().endswith(".txt") and not f.lower().startswith("config")
    ]

    if not txt_files:
        raise FileNotFoundError("当前目录下未找到任何非config的 .txt 文件")

    latest_txt = max(txt_files, key=os.path.getmtime)
    return latest_txt


def load_extensions(config_path):
    """
    从 config.txt 读取需要的文件后缀集合
    """
    exts = set()
    with open(config_path, "r", encoding="utf-8") as f:
        for line in f:
            line = line.strip().lower()
            if not line or line.startswith("#"):
                continue
            if not line.startswith("."):
                line = "." + line
            exts.add(line)
    return exts


def extract_and_copy_files(compile_order_path):
    """
    从 report_compile_order 输出文件中：
    - 提取 Source Compile Order（Design Sources）第一个段落
    - 只保留 config.txt 中指定的需要的文件类型
    - 按完整路径复制文件
    """

    try:
        # 读取 config.txt 中需要的文件后缀
        config_path = os.path.join(os.getcwd(), "config.txt")
        if not os.path.exists(config_path):
            raise FileNotFoundError("当前目录下未找到 config.txt")
        
        include_exts = load_extensions(config_path)
        print(f"从 config.txt 加载需要的文件后缀: {include_exts}")

        # ---------- 读取文件 ----------
        with open(compile_order_path, "r", encoding="utf-8") as f:
            lines = f.readlines()

        # ---------- 定位 Source Compile Order 段落 ----------
        start_idx = None
        end_idx = None

        for i, line in enumerate(lines):
            if line.startswith("Source compile order for 'synthesis'"):
                start_idx = i
                continue

            if start_idx is not None and line.strip() == "":
                end_idx = i
                break

        if start_idx is None:
            raise RuntimeError("未找到 Source compile order for 'synthesis' 段落")

        if end_idx is None:
            end_idx = len(lines)

        section = lines[start_idx:end_idx]

        # ---------- 跳过标题 + 表头 ----------
        data_lines = section[2:]

        extracted_paths = []

        for line in data_lines:
            line = line.rstrip()
            if not line:
                continue

            # 行首必须是索引号
            if not re.match(r"\s*\d+\s+", line):
                continue

            # 用 2 个及以上空格分割，最后一列一定是完整路径
            parts = re.split(r"\s{2,}", line)
            if not parts:
                continue

            full_path = parts[-1]

            # ---------- 只保留 config.txt 中指定的需要的文件类型 ----------
            file_ext = os.path.splitext(full_path)[1].lower()
            if include_exts and file_ext not in include_exts:
                continue

            extracted_paths.append(full_path)

        print(f"提取到 {len(extracted_paths)} 个需要的文件")

        # ---------- 创建输出目录 ----------
        now = datetime.now()
        out_dir = now.strftime("Copy_Files_%Y%m%d_%H%M%S")
        os.makedirs(out_dir)
        print(f"创建输出目录: {out_dir}")

        # ---------- 复制文件 ----------
        copied = 0
        skipped = 0

        for src in extracted_paths:
            if not os.path.exists(src):
                print(f"文件不存在，跳过: {src}")
                skipped += 1
                continue

            base = os.path.basename(src)
            dst = os.path.join(out_dir, base)

            # 避免重名覆盖
            if os.path.exists(dst):
                name, ext = os.path.splitext(base)
                idx = 1
                while True:
                    new_dst = os.path.join(out_dir, f"{name}_{idx}{ext}")
                    if not os.path.exists(new_dst):
                        dst = new_dst
                        break
                    idx += 1

            shutil.copy2(src, dst)
            copied += 1

        print("\n复制完成")
        print(f"  成功复制: {copied}")
        print(f"  跳过文件: {skipped}")
        print(f"  输出目录: {os.path.abspath(out_dir)}")

        return extracted_paths

    except Exception as e:
        print("处理过程中发生错误:")
        print(e)
        return None


if __name__ == "__main__":
    try:
        input_file = find_latest_txt_file()
        print(f"使用最新的 txt 文件: {input_file}")

        result = extract_and_copy_files(input_file)

        if not result:
            print("未提取到任何文件")
        

    except Exception as e:
        print("程序异常退出:")
        print(e)

    finally:
        input("\n按回车键关闭窗口...")