Merge pull request #3 from Awesome-Embedded-Learning-Studio/feat/phase1

feat: 16 new tools

Author: Charliechen114514

CMakeLists.txt

@@ -20,9 +20,17 @@ include(cmake/Config.cmake)
 
 # ── Applet sources (conditional on config) ───────────────────
 set(CFBOX_APPLET_SOURCES)
+
+# Multi-file applets
+if(CFBOX_ENABLE_SH)
+    file(GLOB_RECURSE SH_SOURCES ${CMAKE_CURRENT_SOURCE_DIR}/src/applets/sh/*.cpp)
+    list(APPEND CFBOX_APPLET_SOURCES ${SH_SOURCES})
+endif()
+
+# Single-file applets
 foreach(applet IN LISTS CFBOX_APPLETS)
     string(TOUPPER "${applet}" APPLET_UPPER)
-    if(CFBOX_ENABLE_${APPLET_UPPER})
+    if(NOT applet STREQUAL "sh" AND CFBOX_ENABLE_${APPLET_UPPER})
         list(APPEND CFBOX_APPLET_SOURCES ${CMAKE_CURRENT_SOURCE_DIR}/src/applets/${applet}.cpp)
     endif()
 endforeach()

Roadmap.md

@@ -2,7 +2,7 @@
 
 ## Context
 
-CFBox 是一个 C++23 BusyBox 替代品，当前版本有 17 个 applet（echo, printf, cat, head, tail, wc, sort, uniq, grep, sed, mkdir, rm, cp, mv, ls, find, init）。项目使用注册表分发模式（`APPLET_REGISTRY`）、`std::expected` 错误处理、自定义参数解析器，CI 覆盖原生构建、交叉编译和 QEMU 测试。
+CFBox 是一个 C++23 BusyBox 替代品，当前版本有 33 个 applet。项目使用注册表分发模式（`APPLET_REGISTRY`）、`std::expected` 错误处理、自定义参数解析器，CI 覆盖原生构建、交叉编译和 QEMU 测试。
 
 **目标**：全面对齐 BusyBox，覆盖嵌入式、容器、救援和通用场景。Shell 是最关键的组件，必须最先实现。
 
@@ -15,7 +15,7 @@ CFBox 是一个 C++23 BusyBox 替代品，当前版本有 17 个 applet（echo,
 | 阶段 | 主题 | 新增 Applet | 核心基础设施 | 累计 |
 |------|------|------------|-------------|------|
 | 0 | 构建系统现代化 ✅ | 0 | CMake 配置、help 系统、UTF-8、彩色输出 | 17 |
-| 1 | POSIX Shell + Coreutils I | ~17 | Shell 引擎、进程管理、信号处理 | ~34 |
+| 1 | POSIX Shell + Coreutils I ✅ | ~17 | Shell 引擎、进程管理、信号处理 | ~34 |
 | 2 | Coreutils II + findutils | ~41 | 流处理管线、校验和框架 | ~75 |
 | 3 | 编辑器 + 归档 + 压缩 | ~15 | 终端抽象、压缩框架 | ~90 |
 | 4 | 进程/Init + util-linux | ~38 | /proc 解析器、TUI 框架 | ~128 |
@@ -48,32 +48,33 @@ CFBox 是一个 C++23 BusyBox 替代品，当前版本有 17 个 applet（echo,
 
 ---
 
-## Phase 1：POSIX Shell + Coreutils 第一批（最高优先级）
+## Phase 1：POSIX Shell + Coreutils 第一批 ✅
 
 **目标**：实现交互式 POSIX Shell——这是在真实 Linux 上使用的基础。同时实现简单的 coreutils 建立势头。
 
-### Shell 架构（`src/applets/sh/`）
+### Coreutils 第一批 ✅
 
-Shell 是最复杂的单一组件，按模块拆分：
+16 个简单 coreutils 已完成，覆盖脚本基础所需：
 
-| 模块 | 文件 | 功能 |
+`basename` ✅, `dirname` ✅, `true` ✅, `false` ✅, `yes` ✅, `sleep` ✅, `pwd` ✅, `tty` ✅, `uname` ✅, `whoami` ✅, `hostname` ✅, `id` ✅, `logname` ✅, `nproc` ✅, `test` ✅, `link` ✅
+
+### POSIX Shell ✅
+
+Shell 已实现为第一个多文件 applet（`src/applets/sh/`，8 个模块，~2400 行）：
+
+| 模块 | 文件 | 状态 |
 |------|------|------|
-| 词法分析 | `sh_lexer.cpp` | 分词、引号处理、here-document |
-| 语法解析 | `sh_parser.cpp` | AST 构建：管道、列表、复合命令、函数定义 |
-| 执行器 | `sh_executor.cpp` | AST 遍历执行，fork/exec，管道/重定向设置 |
-| 内建命令 | `sh_builtins.cpp` | cd, export, exit, test, read, trap, umask, eval, source 等 |
-| 行编辑 | `sh_lineedit.cpp` | 原生实现（不依赖 readline）：行编辑、历史记录、Tab 补全 |
-| 作业控制 | `sh_jobs.cpp` | 进程组、会话管理、jobs/fg/bg、Ctrl+Z |
-| 变量系统 | `sh_vars.cpp` | 环境变量、Shell 变量、特殊参数、位置参数 |
-| 词展开 | `sh_expand.cpp` | 波浪号、参数展开、命令替换、算术展开、文件名展开 |
-
-### 需要的基础设施
-- **进程管理** `include/cfbox/process.hpp`：fork/exec/pipe/dup2/waitpid RAII 封装
-- **信号处理** `include/cfbox/signal.hpp`：RAII 信号处理器
-
-### Coreutils 第一批（简单 applet，50-200 行/个）
-
-`basename`, `dirname`, `true`, `false`, `yes`, `sleep`, `pwd`, `tty`, `uname`, `whoami`, `hostname`, `id`, `logname`, `nproc`, `test`, `link`
+| 共享类型 | `sh.hpp` | ✅ AST 节点、Token、ShellState |
+| 词法分析 | `sh_lexer.cpp` | ✅ 分词、引号、$展开、运算符 |
+| 语法解析 | `sh_parser.cpp` | ✅ 递归下降：管道、列表、if/while/for、子shell、花括号组 |
+| 执行器 | `sh_executor.cpp` | ✅ fork/exec/pipe、重定向 < > >>、内置命令 |
+| 内建命令 | `sh_builtins.cpp` | ✅ echo/cd/pwd/export/exit/set/unset/shift/read/eval/source |
+| 词展开 | `sh_expand.cpp` | ✅ $VAR、${VAR}、$?、$$、$#/$@/$*、命令替换、波浪号、glob |
+| 变量系统 | `sh_vars.cpp` | ✅ 变量存储、环境同步、位置参数、特殊参数 |
+
+**已实现功能**：管道 `|`、重定向 `< > >>`、`&&` `||`、`if/elif/else/fi`、`while/until/do/done`、`for/in/do/done`、子shell `()`、花括号组 `{}`、变量赋值、变量展开、命令替换 `$(...)`、tilde 展开、glob 展开、15 个内置命令。
+
+**待后续迭代**：here-document `<<`、函数定义、算术展开 `$(( ))`、行编辑（readline 风格）、作业控制（jobs/fg/bg）、`case/esac`。
 
 ### 验证
 - `./cfbox sh -c "echo hello | wc -l"` 管道工作

cmake/Config.cmake

@@ -6,6 +6,11 @@ include_guard()
 set(CFBOX_APPLETS
     echo printf cat head tail wc sort uniq
     mkdir rm cp mv ls grep find sed init
+    true false yes pwd
+    basename dirname uname nproc link
+    hostname logname whoami tty
+    sleep id test
+    sh
 )
 
 foreach(applet IN LISTS CFBOX_APPLETS)

include/cfbox/applet_config.hpp.in

@@ -21,3 +21,20 @@
 #cmakedefine01 CFBOX_ENABLE_FIND
 #cmakedefine01 CFBOX_ENABLE_SED
 #cmakedefine01 CFBOX_ENABLE_INIT
+#cmakedefine01 CFBOX_ENABLE_TRUE
+#cmakedefine01 CFBOX_ENABLE_FALSE
+#cmakedefine01 CFBOX_ENABLE_YES
+#cmakedefine01 CFBOX_ENABLE_PWD
+#cmakedefine01 CFBOX_ENABLE_BASENAME
+#cmakedefine01 CFBOX_ENABLE_DIRNAME
+#cmakedefine01 CFBOX_ENABLE_UNAME
+#cmakedefine01 CFBOX_ENABLE_NPROC
+#cmakedefine01 CFBOX_ENABLE_LINK
+#cmakedefine01 CFBOX_ENABLE_HOSTNAME
+#cmakedefine01 CFBOX_ENABLE_LOGNAME
+#cmakedefine01 CFBOX_ENABLE_WHOAMI
+#cmakedefine01 CFBOX_ENABLE_TTY
+#cmakedefine01 CFBOX_ENABLE_SLEEP
+#cmakedefine01 CFBOX_ENABLE_ID
+#cmakedefine01 CFBOX_ENABLE_TEST
+#cmakedefine01 CFBOX_ENABLE_SH

include/cfbox/applets.hpp

@@ -55,6 +55,57 @@ extern auto sed_main(int argc, char* argv[]) -> int;
 #if CFBOX_ENABLE_INIT
 extern auto init_main(int argc, char* argv[]) -> int;
 #endif
+#if CFBOX_ENABLE_TRUE
+extern auto true_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_FALSE
+extern auto false_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_YES
+extern auto yes_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_PWD
+extern auto pwd_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_BASENAME
+extern auto basename_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_DIRNAME
+extern auto dirname_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_UNAME
+extern auto uname_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_NPROC
+extern auto nproc_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_LINK
+extern auto link_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_HOSTNAME
+extern auto hostname_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_LOGNAME
+extern auto logname_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_WHOAMI
+extern auto whoami_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_TTY
+extern auto tty_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_SLEEP
+extern auto sleep_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_ID
+extern auto id_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_TEST
+extern auto test_main(int argc, char* argv[]) -> int;
+#endif
+#if CFBOX_ENABLE_SH
+extern auto sh_main(int argc, char* argv[]) -> int;
+#endif
 
 // registry — one line per applet, conditionally compiled
 constexpr auto APPLET_REGISTRY = std::to_array<cfbox::applet::AppEntry>({
@@ -109,4 +160,56 @@ constexpr auto APPLET_REGISTRY = std::to_array<cfbox::applet::AppEntry>({
 #if CFBOX_ENABLE_INIT
     {"init",   init_main,   "system init for boot testing (PID 1)"},
 #endif
+#if CFBOX_ENABLE_TRUE
+    {"true",   true_main,   "do nothing, exit with status 0"},
+#endif
+#if CFBOX_ENABLE_FALSE
+    {"false",  false_main,  "do nothing, exit with status 1"},
+#endif
+#if CFBOX_ENABLE_YES
+    {"yes",    yes_main,    "output a string repeatedly until killed"},
+#endif
+#if CFBOX_ENABLE_PWD
+    {"pwd",    pwd_main,    "print working directory"},
+#endif
+#if CFBOX_ENABLE_BASENAME
+    {"basename", basename_main, "strip directory and suffix from file names"},
+#endif
+#if CFBOX_ENABLE_DIRNAME
+    {"dirname",  dirname_main,  "strip last component from file name"},
+#endif
+#if CFBOX_ENABLE_UNAME
+    {"uname",  uname_main,  "print system information"},
+#endif
+#if CFBOX_ENABLE_NPROC
+    {"nproc",  nproc_main,  "print number of available processors"},
+#endif
+#if CFBOX_ENABLE_LINK
+    {"link",   link_main,   "create a hard link"},
+#endif
+#if CFBOX_ENABLE_HOSTNAME
+    {"hostname", hostname_main, "show or set the system host name"},
+#endif
+#if CFBOX_ENABLE_LOGNAME
+    {"logname", logname_main,  "print the user's login name"},
+#endif
+#if CFBOX_ENABLE_WHOAMI
+    {"whoami", whoami_main, "print effective user ID"},
+#endif
+#if CFBOX_ENABLE_TTY
+    {"tty",    tty_main,    "print the file name of the terminal connected to stdin"},
+#endif
+#if CFBOX_ENABLE_SLEEP
+    {"sleep",  sleep_main,  "delay for a specified amount of time"},
+#endif
+#if CFBOX_ENABLE_ID
+    {"id",     id_main,     "print real and effective user and group IDs"},
+#endif
+#if CFBOX_ENABLE_TEST
+    {"test",   test_main,   "evaluate conditional expression"},
+    {"[",      test_main,   "evaluate conditional expression"},
+#endif
+#if CFBOX_ENABLE_SH
+    {"sh",     sh_main,     "POSIX shell command interpreter"},
+#endif
 });

include/cfbox/fs_util.hpp

@@ -112,6 +112,18 @@ inline auto rename(std::string_view old_path, std::string_view new_path) -> base
     return {};
 }
 
+inline auto create_hard_link(std::string_view target, std::string_view link_path) -> base::Result<void> {
+    std::error_code ec;
+    std::filesystem::create_hard_link(
+        std::filesystem::path{target},
+        std::filesystem::path{link_path},
+        ec);
+    if (ec) {
+        return std::unexpected(base::Error{static_cast<int>(ec.value()), ec.message()});
+    }
+    return {};
+}
+
 inline auto current_path() -> base::Result<std::string> {
     std::error_code ec;
     auto p = std::filesystem::current_path(ec);

src/applets/basename.cpp

@@ -0,0 +1,70 @@
+#include <cstdio>
+#include <string>
+#include <string_view>
+
+#include <cfbox/args.hpp>
+#include <cfbox/help.hpp>
+
+namespace {
+constexpr cfbox::help::HelpEntry HELP = {
+    .name    = "basename",
+    .version = CFBOX_VERSION_STRING,
+    .one_line = "strip directory and suffix from file names",
+    .usage   = "basename PATH [SUFFIX]",
+    .options = "",
+    .extra   = "",
+};
+
+auto do_basename(std::string_view path, std::string_view suffix) -> std::string {
+    // Remove trailing slashes (but keep root "/")
+    while (path.size() > 1 && path.back() == '/') {
+        path.remove_suffix(1);
+    }
+
+    std::string result;
+    if (path.empty()) {
+        result = ".";
+    } else if (path == "/") {
+        result = "/";
+    } else {
+        // Find last component after last '/'
+        auto pos = path.rfind('/');
+        if (pos == std::string_view::npos) {
+            result = std::string{path};
+        } else {
+            result = std::string{path.substr(pos + 1)};
+        }
+    }
+
+    // Strip suffix if given and result ends with it (but not if suffix == result)
+    if (!suffix.empty() && result.size() > suffix.size()) {
+        if (result.compare(result.size() - suffix.size(), suffix.size(), suffix) == 0) {
+            result.erase(result.size() - suffix.size());
+        }
+    }
+
+    return result;
+}
+
+} // namespace
+
+auto basename_main(int argc, char* argv[]) -> int {
+    auto parsed = cfbox::args::parse(argc, argv, {});
+
+    if (parsed.has_long("help"))    { cfbox::help::print_help(HELP); return 0; }
+    if (parsed.has_long("version")) { cfbox::help::print_version(HELP); return 0; }
+
+    const auto& pos = parsed.positional();
+    if (pos.empty()) {
+        std::fprintf(stderr, "cfbox basename: missing operand\n");
+        return 1;
+    }
+
+    std::string_view suffix;
+    if (pos.size() >= 2) {
+        suffix = pos[1];
+    }
+
+    std::puts(do_basename(pos[0], suffix).c_str());
+    return 0;
+}

src/applets/dirname.cpp

@@ -0,0 +1,57 @@
+#include <cstdio>
+#include <string>
+#include <string_view>
+
+#include <cfbox/args.hpp>
+#include <cfbox/help.hpp>
+
+namespace {
+constexpr cfbox::help::HelpEntry HELP = {
+    .name    = "dirname",
+    .version = CFBOX_VERSION_STRING,
+    .one_line = "strip last component from file name",
+    .usage   = "dirname PATH",
+    .options = "",
+    .extra   = "",
+};
+
+auto do_dirname(std::string_view path) -> std::string {
+    if (path.empty()) return ".";
+
+    // Remove trailing slashes
+    while (path.size() > 1 && path.back() == '/') {
+        path.remove_suffix(1);
+    }
+
+    if (path.empty()) return "/";
+
+    auto pos = path.rfind('/');
+    if (pos == std::string_view::npos) return ".";
+
+    // Remove trailing slashes from the parent part
+    auto parent = path.substr(0, pos);
+    while (parent.size() > 1 && parent.back() == '/') {
+        parent.remove_suffix(1);
+    }
+
+    if (parent.empty()) return "/";
+    return std::string{parent};
+}
+
+} // namespace
+
+auto dirname_main(int argc, char* argv[]) -> int {
+    auto parsed = cfbox::args::parse(argc, argv, {});
+
+    if (parsed.has_long("help"))    { cfbox::help::print_help(HELP); return 0; }
+    if (parsed.has_long("version")) { cfbox::help::print_version(HELP); return 0; }
+
+    const auto& pos = parsed.positional();
+    if (pos.empty()) {
+        std::fprintf(stderr, "cfbox dirname: missing operand\n");
+        return 1;
+    }
+
+    std::puts(do_dirname(pos[0]).c_str());
+    return 0;
+}