aria-code 0.167.0__py3-none-win_amd64.whl

aria_code/tools/project_tools.py

@@ -0,0 +1,525 @@
+"""
+project_tools.py — 项目文件夹分析引擎
+========================================
+提供 Claude Code / Codex 同等能力：
+  - 递归扫描项目目录，构建文件树
+  - 自动检测项目类型（Python / Node / Go / Rust / Java / …）
+  - 智能读取关键文件注入 LLM 上下文
+  - 跨文件 grep / 符号索引
+  - Git 状态集成
+  - 按 token 预算动态裁剪上下文
+
+ProjectSession 生命周期：
+  1. scan(path)         — 扫描目录，建立索引
+  2. build_llm_context  — 生成注入 system prompt 的上下文块
+  3. get_file / grep    — LLM 工具调用时按需读取
+"""
+
+from __future__ import annotations
+
+import fnmatch
+import json
+import os
+import re
+import subprocess
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set, Tuple
+
+# ── 忽略规则 ──────────────────────────────────────────────────────────────────
+
+_IGNORE_DIRS: Set[str] = {
+    ".git", ".hg", ".svn",
+    "node_modules", ".pnp",
+    "__pycache__", ".mypy_cache", ".ruff_cache", ".pytest_cache",
+    ".venv", "venv", "env", ".env",
+    "dist", "build", "out", "target", ".next", ".nuxt",
+    "coverage", ".coverage", "htmlcov",
+    ".idea", ".vscode", ".DS_Store",
+    "eggs", "*.egg-info",
+}
+
+_IGNORE_FILE_PATTERNS: List[str] = [
+    "*.pyc", "*.pyo", "*.pyd",
+    "*.so", "*.dll", "*.dylib",
+    "*.class", "*.jar",
+    "*.min.js", "*.min.css",
+    "*.map",
+    "*.lock",           # package-lock.json 例外：内容有用，但体积太大
+    "*.log",
+    ".DS_Store", "Thumbs.db",
+    "*.bin", "*.exe", "*.whl", "*.tar.gz", "*.zip",
+    "*.jpg", "*.jpeg", "*.png", "*.gif", "*.webp", "*.ico",
+    "*.mp4", "*.mp3", "*.pdf",
+]
+
+_BINARY_EXTENSIONS: Set[str] = {
+    ".pyc",".pyo",".pyd",".so",".dll",".dylib",".class",".jar",
+    ".bin",".exe",".whl",".tar",".gz",".zip",".rar",
+    ".jpg",".jpeg",".png",".gif",".bmp",".webp",".ico",
+    ".mp4",".mp3",".wav",".avi",".mkv",
+    ".pdf",".doc",".xls",".ppt",
+}
+
+# ── 项目类型指纹 ──────────────────────────────────────────────────────────────
+
+_PROJECT_SIGNATURES: List[Tuple[str, List[str], str]] = [
+    # (类型名, 指纹文件列表, 主入口候选)
+    ("Python",      ["pyproject.toml", "setup.py", "requirements.txt"],  "main.py"),
+    ("Node.js",     ["package.json"],                                     "index.js"),
+    ("TypeScript",  ["tsconfig.json"],                                    "src/index.ts"),
+    ("Go",          ["go.mod"],                                           "main.go"),
+    ("Rust",        ["Cargo.toml"],                                       "src/main.rs"),
+    ("Java",        ["pom.xml", "build.gradle"],                         "src/main/java"),
+    ("Kotlin",      ["build.gradle.kts"],                                 "src/main/kotlin"),
+    ("Ruby",        ["Gemfile"],                                          "main.rb"),
+    ("PHP",         ["composer.json"],                                    "index.php"),
+    ("C#",          ["*.csproj", "*.sln"],                                "Program.cs"),
+    ("C/C++",       ["CMakeLists.txt", "Makefile"],                       "main.c"),
+    ("Swift",       ["Package.swift"],                                    "Sources/main.swift"),
+    ("Dart/Flutter",["pubspec.yaml"],                                     "lib/main.dart"),
+    ("Elixir",      ["mix.exs"],                                          "lib/main.ex"),
+]
+
+# 关键配置文件（权重高，一定要读）
+_KEY_CONFIG_FILES: List[str] = [
+    "README.md", "README.rst", "README.txt",
+    "pyproject.toml", "setup.py", "requirements.txt",
+    "package.json",
+    "go.mod",
+    "Cargo.toml",
+    "pom.xml",
+    "Makefile",
+    ".env.example", ".env.sample",
+    "docker-compose.yml", "Dockerfile",
+    ".github/workflows",  # CI
+]
+
+# ── 数据结构 ──────────────────────────────────────────────────────────────────
+
+@dataclass
+class FileNode:
+    rel_path: str
+    abs_path: str
+    size:     int
+    ext:      str
+    is_dir:   bool = False
+    language: str  = ""
+
+
+@dataclass
+class ProjectSession:
+    """Loaded project state — persisted in ArtheraTerminal._project_session."""
+    root:         str            = ""
+    name:         str            = ""
+    project_type: str            = ""
+    languages:    List[str]      = field(default_factory=list)
+    files:        List[FileNode] = field(default_factory=list)
+    key_contents: Dict[str, str] = field(default_factory=dict)  # rel_path → text
+    git_info:     Dict[str, Any] = field(default_factory=dict)
+    stats:        Dict[str, Any] = field(default_factory=dict)
+    _tree_cache:  str            = field(default="", repr=False)
+
+    # ── Scan ──────────────────────────────────────────────────────────────────
+
+    def scan(self, path: str, max_files: int = 2000) -> "ProjectSession":
+        root = Path(path).expanduser().resolve()
+        if not root.exists():
+            raise FileNotFoundError(f"路径不存在: {root}")
+        if not root.is_dir():
+            raise ValueError(f"不是目录: {root}")
+
+        self.root = str(root)
+        self.name = root.name
+        self.files = []
+        self._tree_cache = ""
+
+        # Walk directory
+        for dirpath, dirnames, filenames in os.walk(root):
+            # Prune ignored dirs in-place
+            dirnames[:] = [
+                d for d in dirnames
+                if d not in _IGNORE_DIRS
+                and not d.startswith(".")
+                or d in {".github", ".gitlab"}
+            ]
+            # Also skip dirs matching patterns
+            dirnames[:] = [
+                d for d in dirnames
+                if not any(fnmatch.fnmatch(d, p) for p in ["*.egg-info", "*.dist-info"])
+            ]
+
+            for fname in filenames:
+                if len(self.files) >= max_files:
+                    break
+                if any(fnmatch.fnmatch(fname, p) for p in _IGNORE_FILE_PATTERNS):
+                    continue
+                fpath = Path(dirpath) / fname
+                try:
+                    size = fpath.stat().st_size
+                except OSError:
+                    continue
+                ext  = fpath.suffix.lower()
+                if ext in _BINARY_EXTENSIONS:
+                    continue
+                rel = str(fpath.relative_to(root))
+                self.files.append(FileNode(
+                    rel_path=rel,
+                    abs_path=str(fpath),
+                    size=size,
+                    ext=ext,
+                    language=_ext_to_lang(ext),
+                ))
+
+        # Detect project type
+        root_names = {f.rel_path for f in self.files}
+        self.project_type = "Unknown"
+        detected_langs: List[str] = []
+        for ptype, fingerprints, _ in _PROJECT_SIGNATURES:
+            for fp in fingerprints:
+                if "*" in fp:
+                    if any(fnmatch.fnmatch(r, fp) for r in root_names):
+                        self.project_type = ptype
+                        detected_langs.append(ptype)
+                        break
+                elif fp in root_names:
+                    self.project_type = ptype
+                    detected_langs.append(ptype)
+                    break
+        # Language frequency from extensions
+        lang_counts: Dict[str, int] = {}
+        for f in self.files:
+            if f.language:
+                lang_counts[f.language] = lang_counts.get(f.language, 0) + 1
+        top_langs = sorted(lang_counts, key=lang_counts.get, reverse=True)[:4]
+        self.languages = list(dict.fromkeys(detected_langs + top_langs))[:5]
+
+        # Stats
+        total_lines = 0
+        for f in self.files:
+            if f.size < 500_000 and f.ext not in _BINARY_EXTENSIONS:
+                try:
+                    total_lines += Path(f.abs_path).read_text(errors="replace").count("\n")
+                except Exception:
+                    pass
+        self.stats = {
+            "total_files": len(self.files),
+            "total_lines": total_lines,
+            "total_size_kb": sum(f.size for f in self.files) // 1024,
+            "languages": lang_counts,
+        }
+
+        # Read key files
+        self.key_contents = {}
+        self._load_key_files(root)
+
+        # Git info
+        self.git_info = _get_git_info(str(root))
+
+        return self
+
+    def _load_key_files(self, root: Path, max_chars_each: int = 4000):
+        """Read README + config files into key_contents."""
+        for rel in _KEY_CONFIG_FILES:
+            candidate = root / rel
+            if candidate.is_file():
+                try:
+                    text = candidate.read_text(errors="replace")[:max_chars_each]
+                    self.key_contents[rel] = text
+                except Exception:
+                    pass
+
+        # Auto-detect main entry point
+        for _, fingerprints, entry in _PROJECT_SIGNATURES:
+            for fp in fingerprints:
+                if "*" in fp:
+                    continue
+                if (root / fp).exists():
+                    ep = root / entry
+                    if ep.is_file() and entry not in self.key_contents:
+                        try:
+                            self.key_contents[entry] = ep.read_text(errors="replace")[:4000]
+                        except Exception:
+                            pass
+                    break
+
+    # ── Tree rendering ─────────────────────────────────────────────────────────
+
+    def get_tree(self, max_lines: int = 80) -> str:
+        """Return ASCII file tree string."""
+        if self._tree_cache:
+            return self._tree_cache
+
+        root = Path(self.root)
+        # Build nested dict
+        tree: Dict[str, Any] = {}
+        for f in self.files:
+            parts = Path(f.rel_path).parts
+            node = tree
+            for part in parts[:-1]:
+                node = node.setdefault(part, {})
+            node[parts[-1]] = f.size
+
+        lines: List[str] = [f"{self.name}/"]
+        _render_tree(tree, lines, "", max_lines)
+        if len(lines) > max_lines:
+            lines = lines[:max_lines]
+            lines.append(f"  … (仅显示前 {max_lines} 条，共 {len(self.files)} 个文件)")
+        result = "\n".join(lines)
+        self._tree_cache = result
+        return result
+
+    # ── LLM context ───────────────────────────────────────────────────────────
+
+    def build_llm_context(self, max_chars: int = 16_000) -> str:
+        """Build the system-prompt context block injected for every message."""
+        parts: List[str] = []
+        budget = max_chars
+
+        # Header
+        git_branch = self.git_info.get("branch", "")
+        header = (
+            f"=== 已加载项目: {self.name} ===\n"
+            f"路径: {self.root}\n"
+            f"类型: {self.project_type}  |  语言: {', '.join(self.languages[:3])}\n"
+            f"文件: {self.stats.get('total_files',0)} 个  "
+            f"代码行: {self.stats.get('total_lines',0):,}  "
+            f"大小: {self.stats.get('total_size_kb',0)} KB"
+        )
+        if git_branch:
+            header += f"\nGit 分支: {git_branch}"
+        if self.git_info.get("recent_commits"):
+            commits = self.git_info["recent_commits"][:3]
+            header += "\n最近提交:\n" + "\n".join(f"  {c}" for c in commits)
+        parts.append(header)
+        budget -= len(header)
+
+        # File tree (cap at 3000 chars)
+        tree_str = self.get_tree(max_lines=60)
+        tree_block = f"\n--- 文件结构 ---\n{tree_str}"
+        if len(tree_block) <= min(budget, 3000):
+            parts.append(tree_block)
+            budget -= len(tree_block)
+
+        # Key file contents
+        if budget > 1000:
+            for rel, content in self.key_contents.items():
+                if budget <= 500:
+                    break
+                snippet = content[:min(len(content), budget // max(1, len(self.key_contents)))]
+                block = f"\n--- {rel} ---\n{snippet}"
+                parts.append(block)
+                budget -= len(block)
+
+        # Tool capability hint
+        tool_hint = (
+            "\n--- 可用工具 ---\n"
+            "你可以调用以下工具来完成任务：\n"
+            "  read_file(path)            — 读取项目中任意文件\n"
+            "  write_file(path, content)  — 创建或覆写文件\n"
+            "  edit_file(path, old, new)  — 精确查找替换\n"
+            "  search_code(pattern, path) — 跨文件正则搜索\n"
+            "  list_files(path)           — 列出目录内容\n"
+            "  run_command(command, cwd)  — 在项目目录执行 shell 命令\n"
+            "所有路径相对于项目根目录，或使用绝对路径。\n"
+            "执行任务时先规划步骤，再逐步调用工具完成。"
+        )
+        parts.append(tool_hint)
+
+        return "\n".join(parts)
+
+    # ── File reading (safe, within project root) ───────────────────────────────
+
+    def read_file(self, rel_or_abs: str, max_chars: int = 40_000) -> Tuple[bool, str]:
+        """Read a file within the project. Returns (ok, content_or_error)."""
+        p = Path(rel_or_abs)
+        if not p.is_absolute():
+            p = Path(self.root) / rel_or_abs
+        p = p.resolve()
+        root = Path(self.root).resolve()
+        if not str(p).startswith(str(root)):
+            return False, f"安全限制：路径 {p} 在项目目录之外"
+        if not p.exists():
+            return False, f"文件不存在: {p}"
+        if not p.is_file():
+            return False, f"不是文件: {p}"
+        try:
+            content = p.read_text(errors="replace")[:max_chars]
+            return True, content
+        except Exception as e:
+            return False, str(e)
+
+    # ── Grep ──────────────────────────────────────────────────────────────────
+
+    def grep(self, pattern: str, glob: str = "**/*",
+             max_results: int = 60) -> List[Dict[str, Any]]:
+        """Regex search across project files."""
+        results: List[Dict[str, Any]] = []
+        try:
+            rx = re.compile(pattern, re.IGNORECASE)
+        except re.error as e:
+            return [{"error": f"正则表达式错误: {e}"}]
+
+        root = Path(self.root)
+        for f in self.files:
+            if len(results) >= max_results:
+                break
+            if f.ext in _BINARY_EXTENSIONS or f.size > 2_000_000:
+                continue
+            if glob != "**/*":
+                if not fnmatch.fnmatch(f.rel_path, glob):
+                    continue
+            try:
+                lines = Path(f.abs_path).read_text(errors="replace").splitlines()
+                for i, line in enumerate(lines, 1):
+                    if rx.search(line):
+                        results.append({
+                            "file":    f.rel_path,
+                            "line":    i,
+                            "content": line.strip()[:200],
+                        })
+                        if len(results) >= max_results:
+                            break
+            except Exception:
+                continue
+        return results
+
+    # ── Summary ───────────────────────────────────────────────────────────────
+
+    def summary(self) -> Dict[str, Any]:
+        lang_counts = self.stats.get("languages", {})
+        top_langs = sorted(lang_counts, key=lang_counts.get, reverse=True)[:5]
+        return {
+            "name":         self.name,
+            "root":         self.root,
+            "type":         self.project_type,
+            "languages":    top_langs,
+            "total_files":  self.stats.get("total_files", 0),
+            "total_lines":  self.stats.get("total_lines", 0),
+            "total_size_kb":self.stats.get("total_size_kb", 0),
+            "key_files":    list(self.key_contents.keys()),
+            "git":          self.git_info,
+        }
+
+
+# ── Tree renderer ─────────────────────────────────────────────────────────────
+
+def _render_tree(node: dict, lines: list, prefix: str, max_lines: int):
+    items = sorted(node.items(), key=lambda kv: (isinstance(kv[1], dict), kv[0]))
+    for i, (name, val) in enumerate(items):
+        if len(lines) >= max_lines:
+            return
+        is_last = (i == len(items) - 1)
+        connector = "└── " if is_last else "├── "
+        if isinstance(val, dict):
+            lines.append(f"{prefix}{connector}{name}/")
+            ext_prefix = prefix + ("    " if is_last else "│   ")
+            _render_tree(val, lines, ext_prefix, max_lines)
+        else:
+            size_str = _fmt_size(val)
+            lines.append(f"{prefix}{connector}{name}  [{size_str}]")
+
+
+def _fmt_size(size: int) -> str:
+    if size < 1024:     return f"{size}B"
+    if size < 1048576:  return f"{size//1024}KB"
+    return f"{size//1048576}MB"
+
+
+# ── Extension → language ──────────────────────────────────────────────────────
+
+_EXT_LANG: Dict[str, str] = {
+    ".py":"Python", ".pyx":"Python", ".pyi":"Python",
+    ".js":"JavaScript", ".jsx":"JavaScript", ".mjs":"JavaScript",
+    ".ts":"TypeScript", ".tsx":"TypeScript",
+    ".go":"Go",
+    ".rs":"Rust",
+    ".java":"Java", ".kt":"Kotlin", ".kts":"Kotlin",
+    ".rb":"Ruby",
+    ".php":"PHP",
+    ".cs":"C#",
+    ".c":"C", ".h":"C", ".cpp":"C++", ".cc":"C++", ".hpp":"C++",
+    ".swift":"Swift",
+    ".dart":"Dart",
+    ".ex":"Elixir", ".exs":"Elixir",
+    ".html":"HTML", ".htm":"HTML",
+    ".css":"CSS", ".scss":"CSS", ".sass":"CSS", ".less":"CSS",
+    ".json":"JSON", ".jsonl":"JSON",
+    ".yaml":"YAML", ".yml":"YAML",
+    ".toml":"TOML",
+    ".md":"Markdown", ".mdx":"Markdown",
+    ".sh":"Shell", ".bash":"Shell", ".zsh":"Shell",
+    ".sql":"SQL",
+    ".tf":"Terraform", ".hcl":"Terraform",
+    ".proto":"Protobuf",
+    ".r":"R", ".R":"R",
+    ".lua":"Lua",
+    ".vim":"VimScript",
+}
+
+def _ext_to_lang(ext: str) -> str:
+    return _EXT_LANG.get(ext, "")
+
+
+# ── Git integration ───────────────────────────────────────────────────────────
+
+def _get_git_info(root: str) -> Dict[str, Any]:
+    info: Dict[str, Any] = {}
+    if not (Path(root) / ".git").exists():
+        return info
+
+    def _git(*args, capture=True) -> str:
+        try:
+            r = subprocess.run(
+                ["git", *args], cwd=root,
+                capture_output=capture, text=True, timeout=5
+            )
+            return r.stdout.strip() if r.returncode == 0 else ""
+        except Exception:
+            return ""
+
+    info["branch"]   = _git("rev-parse", "--abbrev-ref", "HEAD")
+    info["remote"]   = _git("remote", "get-url", "origin")
+    status_out = _git("status", "--short")
+    if status_out:
+        changed = [l for l in status_out.splitlines() if l.strip()]
+        info["changed_files"] = changed[:20]
+        info["changed_count"] = len(changed)
+    else:
+        info["changed_files"] = []
+        info["changed_count"] = 0
+    log_out = _git("log", "--oneline", "-8")
+    if log_out:
+        info["recent_commits"] = log_out.splitlines()
+    diff_stat = _git("diff", "--stat", "HEAD")
+    if diff_stat:
+        info["diff_stat"] = diff_stat[:800]
+    return info
+
+
+# ── Standalone helpers (called by /project commands) ─────────────────────────
+
+def scan_project(path: str, max_files: int = 2000) -> ProjectSession:
+    """Convenience: create + scan a new ProjectSession."""
+    s = ProjectSession()
+    s.scan(path, max_files=max_files)
+    return s
+
+
+def format_grep_results(results: List[Dict[str, Any]],
+                         pattern: str, max_show: int = 40) -> str:
+    """Format grep results for Rich console display."""
+    if not results:
+        return f'没有找到匹配 "{pattern}" 的内容'
+    if results and "error" in results[0]:
+        return results[0]["error"]
+    lines = [f'搜索 "{pattern}" — {len(results)} 个匹配\n']
+    cur_file = None
+    for r in results[:max_show]:
+        if r["file"] != cur_file:
+            cur_file = r["file"]
+            lines.append(f"\n  📄 {cur_file}")
+        lines.append(f"    {r['line']:>4} │ {r['content']}")
+    if len(results) > max_show:
+        lines.append(f"\n  … 还有 {len(results) - max_show} 个结果")
+    return "\n".join(lines)