feat(argparse): 添加选择类型支持和错误处理优化

添加了 SCC_ARGPARSE_ERR_PNT_DEFAULT 错误类型用于默认操作处理,
实现了 scc_argparse_spec_setup_choices 函数支持枚举选择,
重构了错误处理流程使返回值更加一致。
修复了长选项名称匹配的逻辑错误。

feat(lexer): 添加换行符和注释符号的词法标记

新增 SCC_TOK_ENDLINE 和 SCC_TOK_SHARP 标记类型,
改进词法分析器对换行符和井号的识别处理。

feat(scc_core): 添加常用宏定义

添加 scc_min 和 scc_max 宏定义提供基础数值比较功能。

feat(main): 实现编译器主程序和命令行接口

创建主程序入口实现完整的编译流程,
集成预处理器、词法分析、语法分析和IR生成模块,
添加AST和IR输出功能支持调试查看中间表示。

chore(build): 配置项目构建依赖关系

创建 cbuild.toml 配置文件定义项目包信息和依赖库,
建立编译器各组件库之间的依赖关系管理。
This commit is contained in:
zzy
2026-02-13 17:26:50 +08:00
parent ffee07a03d
commit 088050c903
10 changed files with 370 additions and 22 deletions

13
cbuild.toml Normal file
View File

@@ -0,0 +1,13 @@
[package]
name = "scc"
version = "0.1.0"
dependencies = [
{ name = "argparse", path = "./libs/argparse" },
{ name = "pprocessor", path = "./libs/pprocessor" },
{ name = "lexer", path = "./libs/lexer" },
{ name = "parser", path = "./libs/parser" },
{ name = "ast", path = "./libs/ast" },
{ name = "ast2ir", path = "./libs/ast2ir" },
{ name = "ir", path = "./libs/ir" },
]

View File

@@ -0,0 +1,112 @@
#include <argparse.h>
#ifdef _WIN32
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#endif
typedef struct {
int verbose; // -v, --verbose
const char *output; // -o, --output
const char *config; // --config
const char *input; // input argument
const char *output_dir; // output_dir argument
cbool force; // -f, --force (for subcommand)
const char *path; // path argument (for subcommand)
} parsed_args_t;
int main(int argc, const char **argv, const char **envp) {
#ifdef _WIN32
SetConsoleOutputCP(CP_UTF8);
SetConsoleCP(CP_UTF8);
#endif
// 初始化解析结果结构体
parsed_args_t parsed = {0};
parsed.output = "a.out"; // 默认值
parsed.output_dir = "."; // 默认值
parsed.path = "."; // 默认值
// 初始化 argparse
scc_argparse_t argparse;
scc_argparse_init(&argparse, "demo", "一个命令行参数解析演示工具");
argparse.need_debug = true;
argparse.lang = SCC_ARGPARSE_LANG_ZH; // 中文帮助
scc_argparse_cmd_t *root = scc_argparse_get_root(&argparse);
// ---------- 添加选项 ----------
// 1. --verbose / -v (计数)
scc_argparse_opt_t verbose;
scc_argparse_opt_init(&verbose, 'v', "verbose",
"增加详细程度(可多次使用)");
scc_argparse_spec_setup_count(&verbose.spec, &parsed.verbose);
scc_argparse_cmd_add_opt(root, &verbose);
// 2. --output / -o (字符串)
scc_argparse_opt_t output;
scc_argparse_opt_init(&output, 'o', "output", "输出文件路径");
scc_argparse_spec_setup_string(&output.spec, &parsed.output);
scc_argparse_cmd_add_opt(root, &output);
// 3. --config (字符串)
scc_argparse_opt_t config;
scc_argparse_opt_init(&config, 0, "config", "配置文件路径");
scc_argparse_spec_setup_string(&config.spec, &parsed.config);
scc_argparse_cmd_add_opt(root, &config);
// ---------- 添加位置参数 ----------
// 必需参数: input
scc_argparse_arg_t input;
scc_argparse_arg_init(&input, "input", "输入文件");
scc_argparse_spec_setup_string(&input.spec, &parsed.input);
scc_argparse_spec_set_required(&input.spec, true);
scc_argparse_cmd_add_arg(root, &input);
// 可选参数: output_dir
scc_argparse_arg_t output_dir;
scc_argparse_arg_init(&output_dir, "output_dir", "输出目录");
scc_argparse_spec_setup_string(&output_dir.spec, &parsed.output_dir);
scc_argparse_spec_set_required(&output_dir.spec, false);
scc_argparse_cmd_add_arg(root, &output_dir);
// ---------- 子命令 init ----------
scc_argparse_cmd_t init_cmd;
scc_argparse_cmd_init(&init_cmd, "init", "初始化仓库");
// 子命令选项: --force / -f
scc_argparse_opt_t force;
scc_argparse_opt_init(&force, 'f', "force", "强制覆盖");
scc_argparse_spec_setup_bool(&force.spec, &parsed.force);
scc_argparse_cmd_add_opt(&init_cmd, &force);
// 子命令位置参数: path (可选)
scc_argparse_arg_t path;
scc_argparse_arg_init(&path, "path", "初始化路径");
scc_argparse_spec_setup_string(&path.spec, &parsed.path);
scc_argparse_spec_set_required(&path.spec, false);
scc_argparse_cmd_add_arg(&init_cmd, &path);
// 将子命令加入根命令
scc_argparse_cmd_add_subcmd(root, &init_cmd);
// 解析参数
int ret = scc_argparse_parse(&argparse, argc, argv);
// 打印解析结果
if (ret == 0) {
scc_printf("\n解析成功\n");
scc_printf("解析到的参数和选项:\n");
scc_printf(" verbose: %d\n", parsed.verbose);
scc_printf(" output: %s\n", parsed.output);
scc_printf(" config: %s\n",
parsed.config ? parsed.config : "(未指定)");
scc_printf(" input: %s\n", parsed.input);
scc_printf(" output_dir: %s\n", parsed.output_dir);
scc_printf(" force: %s\n", parsed.force ? "true" : "false");
scc_printf(" path: %s\n", parsed.path);
}
scc_argparse_drop(&argparse);
return ret;
}

View File

@@ -33,6 +33,7 @@ typedef enum scc_argparse_val_type {
typedef enum scc_argparse_err {
SCC_ARGPARSE_ERR_NONE,
SCC_ARGPARSE_ERR_PNT_DEFAULT,
SCC_ARGPARSE_ERR_UNKNOWN_ERR,
SCC_ARGPARSE_ERR_INVALID_ARG,
SCC_ARGPARSE_ERR_INVALID_VALUE,
@@ -206,6 +207,14 @@ static inline void scc_argparse_spec_setup_float(scc_argparse_spec_t *spec,
spec->store.float_store = store;
}
static inline void scc_argparse_spec_setup_choices(scc_argparse_spec_t *spec,
const char **values,
int count) {
spec->value_type = SCC_ARGPARSE_VAL_TYPE_ENUM;
spec->choices.values = values;
spec->choices.count = count;
}
#define SCC_ARGPARSE_MACRO_SETTER(attr) \
static inline void scc_argparse_spec_set_##attr(scc_argparse_spec_t *spec, \
cbool flag) { \

View File

@@ -132,29 +132,33 @@ static int handle_parse_error(scc_argparse_t *parser,
return error;
}
static void validate_and_cleanup(scc_argparse_context_t *ctx,
scc_argparse_t *parser) {
static int validate_and_cleanup(scc_argparse_context_t *ctx,
scc_argparse_t *parser, int errcode) {
if (errcode == SCC_ARGPARSE_ERR_NONE) {
// 检查必需参数是否都已提供
scc_vec_foreach(ctx->current_cmd->args, i) {
scc_argparse_arg_t *arg = &scc_vec_at(ctx->current_cmd->args, i);
if (arg->spec.flag_required && *arg->spec.store.str_store == NULL) {
scc_argparse_print_error(ctx, SCC_ARGPARSE_ERR_MISSING_ARG);
errcode = SCC_ARGPARSE_ERR_MISSING_ARG;
scc_argparse_print_error(ctx, errcode);
break;
}
}
}
// 清理资源
scc_vec_free(ctx->opts);
scc_optparse_drop(&ctx->optparse);
return errcode;
}
static void handle_option(scc_argparse_context_t *ctx, scc_argparse_t *parser) {
static int handle_option(scc_argparse_context_t *ctx, scc_argparse_t *parser) {
scc_argparse_opt_t *opt = (scc_argparse_opt_t *)ctx->result.opt->user_data;
if (parser->need_help && scc_strcmp(opt->long_name, "help") == 0) {
scc_argparse_print_help(parser, ctx->current_cmd);
ctx->parsing_done = true;
return;
return SCC_ARGPARSE_ERR_PNT_DEFAULT;
}
if (opt->spec.flag_store_as_count) {
@@ -190,16 +194,17 @@ static void handle_option(scc_argparse_context_t *ctx, scc_argparse_t *parser) {
// // TODO
// *org_opt->spec.store.str_store = opt_res.value;
// }
return SCC_ARGPARSE_ERR_NONE;
}
static void handle_positional_arg(scc_argparse_context_t *ctx,
static int handle_positional_arg(scc_argparse_context_t *ctx,
scc_argparse_t *parser) {
scc_argparse_cmd_t *subcmd =
is_subcommand(ctx->current_cmd, ctx->result.value);
if (subcmd != NULL) {
ctx->current_cmd = subcmd;
parse_cmd(&ctx->optparse, &ctx->opts, ctx->current_cmd);
return;
return SCC_ARGPARSE_ERR_NONE;
}
if (ctx->positional_index < scc_vec_size(ctx->current_cmd->args)) {
@@ -231,11 +236,13 @@ static void handle_positional_arg(scc_argparse_context_t *ctx,
// }
// }
// }
return SCC_ARGPARSE_ERR_NONE;
}
int scc_argparse_parse(scc_argparse_t *parser, int argc, const char **argv) {
scc_argparse_context_t ctx = {0};
init_context(&ctx, parser, argc, argv); // 初始化上下文
int errcode = SCC_ARGPARSE_ERR_NONE;
while (!ctx.parsing_done &&
scc_optparse_parse(&ctx.optparse, &ctx.result)) {
@@ -245,22 +252,22 @@ int scc_argparse_parse(scc_argparse_t *parser, int argc, const char **argv) {
ctx.result.opt ? ctx.result.opt->long_name : "--",
ctx.result.error, ctx.result.value);
}
if (ctx.result.error) {
handle_parse_error(parser, &ctx);
return 1;
errcode = handle_parse_error(parser, &ctx);
break;
}
if (ctx.result.opt != null) {
handle_option(&ctx, parser);
errcode = handle_option(&ctx, parser);
} else if (ctx.result.value != null) {
handle_positional_arg(&ctx, parser);
errcode = handle_positional_arg(&ctx, parser);
} else {
UNREACHABLE(); // 不应到达此处
}
}
validate_and_cleanup(&ctx, parser);
return 0;
return validate_and_cleanup(&ctx, parser, errcode);
}
void scc_argparse_cmd_init(scc_argparse_cmd_t *cmd, const char *name,

View File

@@ -174,6 +174,9 @@ void scc_argparse_print_help(scc_argparse_t *parser, scc_argparse_cmd_t *cmd) {
const char *scc_argparse_find_similar_arg(scc_argparse_cmd_t *cmd,
const char *arg) {
if (arg == null || cmd == null) {
return null;
}
if (arg[0] == '-' && arg[1] == '-' && arg[2] != '\0') {
// opt arg
scc_vec_foreach(cmd->opts, i) {

View File

@@ -43,7 +43,7 @@ scc_optparse_get_long_name(const scc_optparse_opt_t *opts, const char *name,
break;
}
}
if (name[i] == '\0' || name[i] == end)
if (opt->long_name[i] == '\0' && (name[i] == '\0' || name[i] == end))
return opt;
}
return 0;

View File

@@ -57,6 +57,8 @@ typedef enum scc_cstd {
X(unknown , SCC_TOK_SUBTYPE_INVALID, SCC_TOK_UNKNOWN ) \
X(EOF , SCC_TOK_SUBTYPE_EOF, SCC_TOK_EOF ) \
X(blank , SCC_TOK_SUBTYPE_EMPTYSPACE, SCC_TOK_BLANK ) \
X(endline , SCC_TOK_SUBTYPE_EMPTYSPACE, SCC_TOK_ENDLINE ) \
X("#" , SCC_TOK_SUBTYPE_EMPTYSPACE, SCC_TOK_SHARP ) \
X("==" , SCC_TOK_SUBTYPE_OPERATOR, SCC_TOK_EQ ) \
X("=" , SCC_TOK_SUBTYPE_OPERATOR, SCC_TOK_ASSIGN ) \
X("++" , SCC_TOK_SUBTYPE_OPERATOR, SCC_TOK_ADD_ADD ) \

View File

@@ -334,11 +334,11 @@ void scc_lexer_get_token(scc_lexer_t *lexer, scc_lexer_tok_t *token) {
scc_probe_stream_back(stream);
scc_lex_parse_skip_endline(stream, &lexer->pos);
scc_probe_stream_sync(stream);
token->type = SCC_TOK_BLANK;
token->type = SCC_TOK_ENDLINE;
goto END;
case '#':
parse_line(lexer, token);
token->type = SCC_TOK_BLANK;
token->type = SCC_TOK_SHARP;
goto END;
case '\0':
case scc_stream_eof:

View File

@@ -8,4 +8,7 @@
#define SCC_FUNC
#define scc_min(a, b) ((a) < (b) ? (a) : (b))
#define scc_max(a, b) ((a) > (b) ? (a) : (b))
#endif // __SCC_CORE_MACRO_H__

199
src/main.c Normal file
View File

@@ -0,0 +1,199 @@
#include <argparse.h>
#include <lexer.h>
#include <parser.h>
#include <pprocessor.h>
#include <ast_dump.h>
#include <ir_dump.h>
#include <scc_ast2ir.h>
#include <stdio.h>
static scc_probe_stream_t *from_file_stream(FILE *fp) {
if (fseek(fp, 0, SEEK_END) != 0) {
perror("fseek failed");
return NULL;
}
usize fsize = ftell(fp);
if (fseek(fp, 0, SEEK_SET)) {
perror("fseek failed");
return NULL;
}
char *buffer = (char *)scc_malloc(fsize);
scc_memset(buffer, 0, fsize);
usize read_ret = fread(buffer, 1, fsize, fp);
fclose(fp);
scc_probe_stream_t *stream =
scc_mem_probe_stream_alloc(buffer, read_ret, true);
return stream;
}
typedef struct {
const char *input_file;
const char *output_file;
int verbose;
cbool dump_ast;
cbool dump_ir;
} scc_config_t;
static void setup_argparse(scc_argparse_t *argparse, scc_config_t *config,
scc_argparse_lang_t lang) {
enum {
SCC_HINT_PROG_NAME,
SCC_HINT_DESCRIPTION,
SCC_HINT_OUTPUT_FILE,
SCC_HINT_INPUT_FILE,
SCC_HINT_VERBOSE,
SCC_HINT_EMIT_AST,
SCC_HINT_EMIT_IR,
};
static const char *scc_hints_en[] = {
[SCC_HINT_PROG_NAME] = "scc",
[SCC_HINT_DESCRIPTION] = "A simple C compiler",
[SCC_HINT_OUTPUT_FILE] = "Output file",
[SCC_HINT_INPUT_FILE] = "Input source file",
[SCC_HINT_VERBOSE] = "Increase verbosity (can be used multiple times)",
[SCC_HINT_EMIT_AST] = "Generate AST and exit",
[SCC_HINT_EMIT_IR] = "Generate IR and exit",
};
static const char *scc_hints_zh[] = {
[SCC_HINT_PROG_NAME] = "scc",
[SCC_HINT_DESCRIPTION] = "一个简单的C编译器",
[SCC_HINT_OUTPUT_FILE] = "输出文件",
[SCC_HINT_INPUT_FILE] = "输入源文件",
[SCC_HINT_VERBOSE] = "增加详细输出(可多次使用)",
[SCC_HINT_EMIT_AST] = "生成 AST 并退出",
[SCC_HINT_EMIT_IR] = "生成 IR 并退出",
};
const char **scc_hints;
switch (lang) {
case SCC_ARGPARSE_LANG_EN:
scc_hints = scc_hints_en;
break;
case SCC_ARGPARSE_LANG_ZH:
scc_hints = scc_hints_zh;
break;
default:
scc_hints = scc_hints_en;
break;
}
scc_argparse_init(argparse, scc_hints[SCC_HINT_PROG_NAME],
scc_hints[SCC_HINT_DESCRIPTION]);
argparse->lang = lang;
scc_argparse_cmd_t *root = scc_argparse_get_root(argparse);
// -o, --output
scc_argparse_opt_t opt_output;
scc_argparse_opt_init(&opt_output, 'o', "output",
scc_hints[SCC_HINT_OUTPUT_FILE]);
scc_argparse_spec_setup_string(&opt_output.spec, &(config->output_file));
scc_argparse_cmd_add_opt(root, &opt_output);
// input file (必需)
scc_argparse_arg_t arg_input;
scc_argparse_arg_init(&arg_input, "input", scc_hints[SCC_HINT_INPUT_FILE]);
scc_argparse_spec_setup_string(&arg_input.spec, &(config->input_file));
scc_argparse_spec_set_required(&arg_input.spec, true);
scc_argparse_cmd_add_arg(root, &arg_input);
// -v, --verbose (计数)
scc_argparse_opt_t opt_verbose;
scc_argparse_opt_init(&opt_verbose, 'v', "verbose",
scc_hints[SCC_HINT_VERBOSE]);
scc_argparse_spec_setup_count(&opt_verbose.spec, &(config->verbose));
scc_argparse_cmd_add_opt(root, &opt_verbose);
// -T, --ast
scc_argparse_opt_t opt_ast;
scc_argparse_opt_init(&opt_ast, 'T', "emit-ast",
scc_hints[SCC_HINT_EMIT_AST]);
scc_argparse_spec_setup_bool(&opt_ast.spec, &(config->dump_ast));
scc_argparse_cmd_add_opt(root, &opt_ast);
// -R, --ir
scc_argparse_opt_t opt_ir;
scc_argparse_opt_init(&opt_ir, 'R', "emit-ir", scc_hints[SCC_HINT_EMIT_IR]);
scc_argparse_spec_setup_bool(&opt_ir.spec, &(config->dump_ir));
scc_argparse_cmd_add_opt(root, &opt_ir);
}
#ifdef _WIN32
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#endif
int main(int argc, const char **argv, const char **envp) {
#ifdef _WIN32
SetConsoleOutputCP(CP_UTF8);
SetConsoleCP(CP_UTF8);
#endif
scc_config_t config = {
.input_file = NULL,
.output_file = "a.exe",
.verbose = 0,
.dump_ast = false,
.dump_ir = false,
};
scc_argparse_t argparse;
setup_argparse(&argparse, &config, SCC_ARGPARSE_LANG_ZH);
int ret = scc_argparse_parse(&argparse, argc, argv);
if (ret != 0) {
scc_argparse_drop(&argparse);
return ret;
}
scc_argparse_drop(&argparse);
setbuf(stdout, NULL);
FILE *fp = fopen(config.input_file, "r");
if (!fp) {
perror("fopen");
scc_argparse_drop(&argparse);
return 1;
}
scc_pproc_t pproc;
scc_probe_stream_t *source_code_stream = from_file_stream(fp);
// scc_probe_stream_t *pprocessed_code_stream =
// scc_pproc_init(&pproc, source_code_stream);
scc_lexer_t lexer;
scc_lexer_init(&lexer, source_code_stream);
scc_lexer_stream_t lexer_stream;
scc_lexer_to_stream(&lexer, &lexer_stream, false);
scc_parser_t parser;
scc_parser_init(&parser, &lexer_stream, null);
scc_ast_translation_unit_t *translation_unit =
scc_parse_translation_unit(&parser);
if (config.dump_ast) {
scc_tree_dump_ctx_t tree_dump;
scc_tree_dump_ctx_init(&tree_dump, true);
scc_ast_dump_node(&tree_dump, (scc_ast_node_t *)translation_unit);
scc_tree_dump_ctx_drop(&tree_dump);
return 0;
}
scc_ir_builder_t ir_builder;
scc_ast2ir(translation_unit, &ir_builder);
if (config.dump_ir) {
scc_ir_dump_ctx_t ir_dump_ctx;
scc_tree_dump_ctx_t tree_dump; // 仅为 ir dump 辅助
scc_tree_dump_ctx_init(&tree_dump, true);
scc_ir_dump_ctx_init(&ir_dump_ctx, &tree_dump, &ir_builder.cprog,
&ir_builder.ctx);
// scc_ir_dump_cprog(&ir_dump_ctx);
scc_ir_dump_cprog_linear(&ir_dump_ctx);
scc_tree_dump_ctx_drop(&tree_dump);
return 0;
}
scc_printf("output exe at %s", config.output_file);
return 0;
}