failing to impl stringbuilder

Makefile

@@ -1,104 +1,243 @@
+# Basic configuration
 TARGET := langc
-
-
 AFLAGS = -felf64 -F dwarf -g
 
 ifdef DEBUG
 	AFLAGS += -DDEBUG_BUILD
 endif
 
+# Directory structure
 SRCDIR := src
+OBJDIR := obj
+LIBDIR := lib
+
+# Core directories
 COREDIR := $(SRCDIR)/core
 MATHDIR := $(COREDIR)/math
 VECDIR := $(COREDIR)/vector
 STRDIR := $(COREDIR)/string
+SBDIR := $(COREDIR)/string_builder
 PRINTDIR := $(COREDIR)/print
 MEMDIR := $(COREDIR)/mem
 SYSCALLDIR := $(COREDIR)/syscall
 FILEDIR := $(COREDIR)/file
+
+# Parser directories
 PARSEDIR := $(SRCDIR)/parse
 EXPRDIR := $(PARSEDIR)/expression
 TOKDIR := $(PARSEDIR)/token
+
+# Lexer directories
 LEXDIR := $(SRCDIR)/lexer
 VARDIR := $(LEXDIR)/vars
+
+# Other directories
 GLOBALDIR := $(SRCDIR)/global
 
-# Define source files
+# Object file directories (mirroring source structure)
+OBJCOREDIR := $(OBJDIR)/core
+OBJMATHDIR := $(OBJCOREDIR)/math
+OBJVECDIR := $(OBJCOREDIR)/vector
+OBJSTRDIR := $(OBJCOREDIR)/string
+OBJSBDIR := $(OBJCOREDIR)/string_builder
+OBJPRINTDIR := $(OBJCOREDIR)/print
+OBJMEMDIR := $(OBJCOREDIR)/mem
+OBJSYSCALLDIR := $(OBJCOREDIR)/syscall
+OBJFILEDIR := $(OBJCOREDIR)/file
+OBJPARSEDIR := $(OBJDIR)/parse
+OBJEXPRDIR := $(OBJPARSEDIR)/expression
+OBJTOKDIR := $(OBJPARSEDIR)/token
+OBJLEXDIR := $(OBJDIR)/lexer
+OBJVARDIR := $(OBJLEXDIR)/vars
+OBJGLOBALDIR := $(OBJDIR)/global
+
+# All object directories in dependency order
+OBJDIRS := $(OBJDIR) \
+           $(OBJCOREDIR) \
+           $(OBJMATHDIR) \
+           $(OBJVECDIR) \
+           $(OBJSTRDIR) \
+           $(OBJSBDIR) \
+           $(OBJPRINTDIR) \
+           $(OBJMEMDIR) \
+           $(OBJSYSCALLDIR) \
+           $(OBJFILEDIR) \
+           $(OBJPARSEDIR) \
+           $(OBJEXPRDIR) \
+           $(OBJTOKDIR) \
+           $(OBJLEXDIR) \
+           $(OBJVARDIR) \
+           $(OBJGLOBALDIR)
+
+# Source file definitions by module
 MATHSRC := $(addprefix $(MATHDIR)/, $(addsuffix .s, \
             operators \
             ))
+
 STRSRC := $(addprefix $(STRDIR)/, $(addsuffix .s, \
             strlen split strcpy substr is_num strcmp is_alpha \
             ))
+
 MEMSRC := $(addprefix $(MEMDIR)/, $(addsuffix .s, \
             malloc memchr memcpy memset \
             ))
+
 VECSRC := $(addprefix $(VECDIR)/, $(addsuffix .s, \
 			vec_create vec_push vec_get vec_pop \
             ))
+
 PRINTSRC := $(addprefix $(PRINTDIR)/, $(addsuffix .s, \
             print putnumber \
             ))
+
 FILESRC := $(addprefix $(FILEDIR)/, $(addsuffix .s, \
             read_file get_file_content \
             ))
+
 SYSCALLSRC := $(addprefix $(SYSCALLDIR)/, $(addsuffix .s, \
-            exit file_ops syscall_err\
+            exit file_ops syscall_err fork\
             ))
+
 TOKSRC := $(addprefix $(TOKDIR)/, $(addsuffix .s, \
             parse_tokens debug_token \
             ))
+
 EXPRSRC := $(addprefix $(EXPRDIR)/, $(addsuffix .s, \
             create_expressions debug_expression \
             ))
+
 LEXSRC := $(addprefix $(LEXDIR)/, $(addsuffix .s, \
             lexer lex_err lex_load lex_func program_prologue \
 			func_boiler_plate \
             ))
+
 VARSRC := $(addprefix $(VARDIR)/, $(addsuffix .s, \
             get_vars insert_var \
             ))
+
 GLOBALSRC := $(addprefix $(GLOBALDIR)/, $(addsuffix .s, \
             function_table regs \
             ))
 
-# Collect all source files
+SBSRC := $(addprefix $(SBDIR)/, $(addsuffix .s, \
-SRC := $(SRCDIR)/start.s $(MATHSRC) $(STRSRC) $(PRINTSRC) $(FILESRC) $(VARSRC) $(SYSCALLSRC) $(MEMSRC) $(TOKSRC) $(EXPRSRC) $(LEXSRC) $(GLOBALSRC) $(VECSRC)
+			string_builder\
+            ))
 
-# Fix: Preserve directory structure in object files
+# Collect all sources and objects
-OBJDIR := obj
+MAIN_SRC := $(SRCDIR)/start.s
-OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(SRC))
+ALL_SRC := $(MAIN_SRC) $(MATHSRC) $(STRSRC) $(SBSRC) $(PRINTSRC) $(FILESRC) $(VARSRC) $(SYSCALLSRC) $(MEMSRC) $(TOKSRC) $(EXPRSRC) $(LEXSRC) $(GLOBALSRC) $(VECSRC)
 
-LIBOBJ := $(filter-out start.o, $(OBJ))
+# Generate object file paths
-LIBNAME := core.lang
+ALL_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(ALL_SRC))
 
-all: $(TARGET)
+# Library settings
+LIBNAME := $(LIBDIR)/core.a
+LIB_OBJ := $(filter-out $(OBJDIR)/start.o, $(ALL_OBJ))
 
-lib: $(OBJ)
+# Module-specific object files for staged compilation
-	ar rcs $(LIBNAME) $(LIBOBJ)
+MATH_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(MATHSRC))
+STR_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(STRSRC))
+MEM_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(MEMSRC))
+VEC_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(VECSRC))
+PRINT_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(PRINTSRC))
+FILE_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(FILESRC))
+SYSCALL_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(SYSCALLSRC))
+TOK_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(TOKSRC))
+EXPR_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(EXPRSRC))
+LEX_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(LEXSRC))
+VAR_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(VARSRC))
+GLOBAL_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(GLOBALSRC))
+SB_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(SBSRC))
+MAIN_OBJ := $(patsubst $(SRCDIR)/%.s,$(OBJDIR)/%.o,$(MAIN_SRC))
 
-# Create output directories
+# Main targets
-$(OBJDIR)/core/math $(OBJDIR)/core/string $(OBJDIR)/core/print $(OBJDIR)/core/mem $(OBJDIR)/core/syscall $(OBJDIR)/core/file $(OBJDIR)/parse/expression $(OBJDIR)/parse/token $(OBJDIR)/lexer $(OBJDIR)/lexer/vars $(OBJDIR)/global:
+all: prepare-build build-core build-parser build-lexer build-global build-main link-executable create-library
+
+# Stage 1: Prepare build environment
+prepare-build: create-directories
+
+create-directories: | $(OBJDIRS) $(LIBDIR)
+
+$(OBJDIRS):
 	mkdir -p $@
 
-# Main target
+$(LIBDIR):
-$(TARGET): $(OBJDIR) $(OBJDIR)/core/math $(OBJDIR)/core/string $(OBJDIR)/core/print $(OBJDIR)/core/mem $(OBJDIR)/core/syscall $(OBJDIR)/core/file $(OBJDIR)/parse/expression $(OBJDIR)/parse/token $(OBJDIR)/lexer $(OBJDIR)/lexer/vars $(OBJDIR)/global $(OBJ)
+	mkdir -p $@
-	ld -g -o $@ $(OBJ) -nostdlib -static
 
-# Fix: Use a more specific pattern rule that preserves paths
+# Stage 2: Build core modules
-$(OBJDIR)/%.o: $(SRCDIR)/%.s
+build-core: build-math build-string build-memory build-vector build-print build-file build-syscall build-string-builder
-	mkdir -p $(dir $@)
+
+build-math: $(MATH_OBJ)
+
+build-string: $(STR_OBJ)
+
+build-memory: $(MEM_OBJ)
+
+build-vector: $(VEC_OBJ)
+
+build-print: $(PRINT_OBJ)
+
+build-file: $(FILE_OBJ)
+
+build-syscall: $(SYSCALL_OBJ)
+
+build-string-builder: $(SB_OBJ)
+
+# Stage 3: Build parser modules
+build-parser: build-tokens build-expressions
+
+build-tokens: $(TOK_OBJ)
+
+build-expressions: $(EXPR_OBJ)
+
+# Stage 4: Build lexer modules
+build-lexer: build-lex-core build-lex-vars
+
+build-lex-core: $(LEX_OBJ)
+
+build-lex-vars: $(VAR_OBJ)
+
+# Stage 5: Build global modules
+build-global: $(GLOBAL_OBJ)
+
+# Stage 6: Build main entry point
+build-main: $(MAIN_OBJ)
+
+# Stage 7: Link executable
+link-executable: $(TARGET)
+
+$(TARGET): $(ALL_OBJ)
+	ld -g -o $@ $(ALL_OBJ) -nostdlib -static
+
+# Stage 8: Create library
+create-library: $(LIBNAME)
+
+$(LIBNAME): $(LIB_OBJ) | $(LIBDIR)
+	ar rcs $@ $(LIB_OBJ)
+
+# Individual file compilation rule
+$(OBJDIR)/%.o: $(SRCDIR)/%.s | $(OBJDIRS)
 	nasm $(AFLAGS) $< -o $@
 
-$(OBJDIR):
+# Utility targets
-	mkdir -p $@
-
 clean:
 	rm -rf $(OBJDIR)
 
-fclean: clean
+clean-library:
-	rm -rf $(TARGET)
+	rm -f $(LIBNAME)
 
-re: clean all
+clean-executable:
+	rm -f $(TARGET)
+
+fclean: clean clean-library clean-executable
+
+re: fclean all
+
+# Debug target to show what would be built
+show-config:
+	@echo "Target: $(TARGET)"
+	@echo "Flags: $(AFLAGS)"
+	@echo "Source files: $(words $(ALL_SRC)) files"
+	@echo "Object files: $(words $(ALL_OBJ)) files"
+	@echo "Object directories: $(OBJDIRS)"
 
-.PHONY: all clean re

src/core/string_builder/sappend.s

@@ -0,0 +1,77 @@
+%include "./src/core/string_builder/sb.s"
+
+section .text
+    extern strlen
+    extern strcpy
+    extern malloc
+    extern err_malloc
+
+%define SB [rbp - 16]
+%define SB_LEN [rbp - 4]
+%define SB_CAP [rbp - 8]
+%define APPENDIX [rbp - 24]
+
+sb_append:      ; (rdi: *sb, rsi: char*)
+    push rbp
+    mov rbp, rsp
+    sub rsp, 32
+
+    test rsi, rsi
+    jz .done
+    ; store sb on stack
+    mov SB, rdi
+
+    ; store new str on stack
+    mov APPENDIX, rsi
+
+    ; get sb len
+    mov eax, dword [rdi + STR_LEN]
+    mov dword SB_LEN, eax
+
+    mov eax, dword [rdi + STR_CAP]
+    mov dword SB_CAP, eax
+
+    push rdi
+    mov rdi, rsi
+    call strlen
+
+    add rax, SB_LEN
+    cmp rax, SB_CAP
+    jl .copy_string
+
+    mov r9, SB_CAP
+    ; new string will be to large for current cap, need to realloc
+.get_new_len:
+    imul r9, 2
+    cmp r9, rax
+    jl .get_new_len
+    mov SB_CAP, r9
+    push r9
+    push rax
+    mov rdi, r9
+    call malloc
+    test rax, rax
+    jz err_malloc
+    mov rdi, rax
+    mov rsi, SB
+    mov rsi, [rsi + STR_DATA]
+    call strcpy
+    pop rax
+    mov rsi, SB
+    pop r9
+    mov [SB + STR_CAP], r9
+
+.copy_string:
+    pop rdi
+    mov r9, SB
+    mov dword [r9 + STR_LEN], eax
+    mov rdi, [r9 + STR_DATA]
+    mov eax, dword [rbp - 4]
+    lea rdi, [rdi + eax]
+    mov rsi, [rbp - 24]
+    call strcpy
+
+.done:
+    mov rsp, rbp
+    pop rbp
+    ret

src/core/string_builder/sb.s

@@ -0,0 +1,12 @@
+%define STRING_INIT_CAP 1024
+
+
+%define STR_SIZE    16
+%define STR_LEN     0
+%define STR_CAP     4
+%define STR_DATA    8
+
+; struct string:
+;   .len    0   uint
+;   .cap    4   uint
+;   .data   8   char*

src/core/string_builder/string_builder.s

@@ -0,0 +1,56 @@
+%include "./src/core/string_builder/sb.s"
+
+section .text
+    extern malloc
+    extern err_malloc
+
+    extern memcpy
+    extern strlen
+
+global sb_new
+sb_new:     ; rax: str*(rdi: char* || NULL, rsi: *hidden_copy_ptr)
+    push rbx
+    push rdi
+    push rsi
+
+    mov rbx, STRING_INIT_CAP
+    xor r9, r9
+
+    test rdi, rdi
+    jz .alloc_string
+
+    push rdi
+    call strlen
+    pop rdi
+    mov r9, rax
+    cmp r9, STRING_INIT_CAP
+    jl .alloc_string
+
+.calc_init_len:
+    add rbx, STRING_INIT_CAP
+    cmp r9, rbx
+    jg .calc_init_len
+
+
+.alloc_string:
+    push rdi
+    mov rdi, rbx
+    call malloc
+    test rax, rax
+    jz err_malloc
+    pop rdi
+
+    pop rsi
+    mov dword [rsi + STR_CAP], ebx
+    mov dword [rsi + STR_LEN], r9d
+    mov [rsi + STR_DATA], rax
+    test r9, r9
+    jz .done
+    mov rdi, rax
+    mov rdx, r9
+    pop rsi
+    call memcpy
+
+.done:
+    pop rbx
+    ret

src/core/syscall/fork.s

@@ -0,0 +1,9 @@
+%define SYS_FORK 57
+
+section .text
+
+global fork
+fork:   ; rax: pid()
+    mov rdi, SYS_FORK
+    syscall
+    ret

src/core/syscall/syscall_err.s

@@ -1,7 +1,7 @@
 section .data
-    EARGCNT: db "[ERROR] Invalid arg count: expected 1", 0xa, 0
+    EARGCNT: db 0xa, "[ERROR] Invalid arg count: expected 1", 0xa, 0
-    EMALLOC: db "[ERROR] Malloc fialed!", 0xa, 0
+    EMALLOC: db 0xa, "[ERROR] Malloc failed!", 0xa, 0
-    ELSEEK: db "[ERROR] lseek failed!", 0xa, 0
+    ELSEEK: db 0xa, "[ERROR] lseek failed!", 0xa, 0
 
 section .text
     global err_args

src/lexer/lexer.s

@@ -19,6 +19,7 @@ section .text
 
     extern func_prologue
     extern func_epilogue
+    extern memset
 
     extern program_prologue
 
@@ -37,6 +38,14 @@ lex:            ; rax: lex* (rdi: char *file_content)
     cmp rax, 0
     je err_malloc
 
+    ; zero out lexer
+    push rax
+    mov rdi, rax
+    mov rsi, 0
+    mov rdx, LEX_SIZE
+    call memset
+
+    pop rax
     mov [rbp - 24], rax     ; store lex on stack
 
     lea rsi, [rbp - 16]     ; int* expr_cnt

src/start.s

@@ -13,31 +13,40 @@ section .text
     extern vec_get
     extern putchar
     extern vec_pop
+    extern sb_new
 
 print_usage:
     mov rdi, usage
     call putstr
 
+
 _start:
-    pop rdi
-    cmp rdi, 2
-    jne err_args
-    mov rdi, [rsp + 8]          ; argv[1]
     push rbp
     mov rbp, rsp
-
     sub rsp, 16
+    mov rdi, usage
+    lea rsi, [rsp - 16]
+    call sb_new
 
-    call get_file_content
+;    pop rdi
-    mov rdi, rax
+;    cmp rdi, 2
-    mov [rbp - 8], rax
+;    jne err_args
-
+;    mov rdi, [rsp + 8]          ; argv[1]
-    mov rdi, rax
+;    push rbp
-    call lex
+;    mov rbp, rsp
-
+;
-    mov rsp, rbp
+;    sub rsp, 16
-    pop rbp
+;
-
+;    call get_file_content
+;    mov rdi, rax
+;    mov [rbp - 8], rax
+;
+;    mov rdi, rax
+;    call lex
+;
+;    mov rsp, rbp
+;    pop rbp
+;
 done:
-    xor rdi, rdi
+;    xor rdi, rdi
-    call exit
+;    call exit