Meow

.clang-format

@@ -123,7 +123,7 @@ IncludeIsMainSourceRegex: ''
 IndentAccessModifiers: false
 IndentCaseBlocks: false
 IndentCaseLabels: false
-IndentExternBlock: AfterExternBlock
+IndentExternBlock: Indent
 IndentGotoLabels: true
 IndentPPDirectives: None
 IndentRequiresClause: true

src/dynarray.c

@@ -1,5 +1,7 @@
 
 #include "dynarray.h"
+#include <stdlib.h>
+#include <string.h>
 
 /*
 
@@ -19,34 +21,34 @@ To set the ith element of the array, use either bracket notation
 
 // Returns a pointer to the start of a new dynarray (after the header) which
 // has `init_cap` units of `stride` bytes.
-void *_dynarray_create(size_t init_cap, size_t stride) {
+void* _dynarray_create(size_t init_cap, size_t stride) {
     size_t header_size = DYNARRAY_FIELDS * sizeof(size_t);
     size_t arr_size = init_cap * stride;
-  size_t *arr = (size_t *)malloc(header_size + arr_size);
+    size_t* arr = (size_t*)malloc(header_size + arr_size);
     arr[CAPACITY] = init_cap;
     arr[LENGTH] = 0;
     arr[STRIDE] = stride;
-  return (void *)(arr + DYNARRAY_FIELDS);
+    return (void*)(arr + DYNARRAY_FIELDS);
 }
 
-void _dynarray_destroy(void *arr) {
+void _dynarray_destroy(void* arr) {
     free(arr - DYNARRAY_FIELDS * sizeof(size_t));
 }
 
 // Returns the dynarray's field which is specified by passing
 // one of CAPACITY, LENGTH, STRIDE.
-size_t _dynarray_field_get(void *arr, size_t field) {
-  return ((size_t *)(arr)-DYNARRAY_FIELDS)[field];
+size_t _dynarray_field_get(void* arr, size_t field) {
+    return ((size_t*)(arr)-DYNARRAY_FIELDS)[field];
 }
 
-void _dynarray_field_set(void *arr, size_t field, size_t value) {
-  ((size_t *)(arr)-DYNARRAY_FIELDS)[field] = value;
+void _dynarray_field_set(void* arr, size_t field, size_t value) {
+    ((size_t*)(arr)-DYNARRAY_FIELDS)[field] = value;
 }
 
 // Allocates a new dynarray with twice the size of the one passed in, and
 // retaining the values that the original stored.
-void *_dynarray_resize(void *arr) {
-  void *temp = _dynarray_create( // Allocate new dynarray w/ more space.
+void* _dynarray_resize(void* arr) {
+    void* temp = _dynarray_create( // Allocate new dynarray w/ more space.
         DYNARRAY_RESIZE_FACTOR * dynarray_capacity(arr), dynarray_stride(arr));
     memcpy(temp, arr,
            dynarray_length(arr) * dynarray_stride(arr)); // Copy erythin' over.
@@ -56,20 +58,37 @@ void *_dynarray_resize(void *arr) {
     return temp;
 }
 
-void *_dynarray_push(void *arr, void *xptr) {
-  if (dynarray_length(arr) >= dynarray_capacity(arr))
+void* _dynarray_push(void* arr, void* xptr) {
+    if (dynarray_length(arr) >= dynarray_capacity(arr)) {
         arr = _dynarray_resize(arr);
+    }
 
-  memcpy(arr + dynarray_length(arr) * dynarray_stride(arr), xptr,
-         dynarray_stride(arr));
+    memcpy(arr + dynarray_length(arr) * dynarray_stride(arr), xptr, dynarray_stride(arr));
     _dynarray_field_set(arr, LENGTH, dynarray_length(arr) + 1);
     return arr;
 }
 
 // Removes the last element in the array, but copies it to `*dest` first.
-void _dynarray_pop(void *arr, void *dest) {
-  memcpy(dest, arr + (dynarray_length(arr) - 1) * dynarray_stride(arr),
-         dynarray_stride(arr));
+int _dynarray_pop(void* arr, void* dest) {
+    if (dynarray_length(arr) < 1) {
+        return 1;
+    }
+    memcpy(dest, arr + (dynarray_length(arr) - 1) * dynarray_stride(arr), dynarray_stride(arr));
     _dynarray_field_set(arr, LENGTH,
                         dynarray_length(arr) - 1); // Decrement length.
+    return 0;
+}
+
+// copies everything in reverse into a temporary array, and then memcpy's everything into the old array;
+void _dynarray_reverse(void* arr) {
+    size_t stride = dynarray_stride(arr);
+    size_t length = dynarray_length(arr);
+    void* temp_arr = _dynarray_create(length, stride);
+
+    void* temp_item = malloc(stride);
+    for (int i = dynarray_length(arr); i > 0; i--) {
+        dynarray_pop(arr, temp_item);
+        memcpy(arr + i * stride, temp_item, stride);
+    }
+    memcpy(arr, temp_arr, stride * length);
 }

src/include/ast.h

@@ -0,0 +1,69 @@
+#ifndef _H_MORPH_AST
+#define _H_MORPH_AST
+
+#include "loc.h"
+#include "token.h"
+#include <stddef.h>
+
+typedef struct const_s {
+    const char* name;
+
+    union {
+        const char* str_v;
+        size_t int_v;
+    };
+} const_t;
+
+typedef struct memory_s {
+    const char* name;
+    size_t size;
+} memory_t;
+
+typedef struct ast_if_stat_s {
+    token_t* condition;
+    token_t* body;
+    bool is_elseif;
+
+    union {
+        struct ast_if_stat_s* elseif;
+        token_t else_body;
+    };
+} ast_if_stat_t;
+
+typedef struct ast_while_stat_s {
+    token_t* condition;
+    token_t* body;
+} ast_while_stat_t;
+
+typedef enum ast_op_type_e {
+    AOT_OP,
+    AOT_IF,
+    AOT_WHILE,
+} ast_op_type_t;
+
+typedef struct ast_op_s {
+    ast_op_type_t type;
+    loc_t loc;
+
+    union {
+        token_t op;
+        ast_if_stat_t if_stat;
+        ast_while_stat_t while_stat;
+    };
+} ast_op_t;
+
+typedef struct function_s {
+    loc_t loc;
+    const char* name;
+    const char** args;
+    const char** return_args;
+    ast_op_type_t* body;
+} function_t;
+
+typedef struct program_s {
+    const_t const_vars;
+    function_t funcs;
+    memory_t memories;
+} program_t;
+
+#endif // !_H_MORPH_AST

src/include/dynarray.h

@@ -24,7 +24,9 @@ void _dynarray_field_set(void* arr, size_t field, size_t value);
 void* _dynarray_resize(void* arr);
 
 void* _dynarray_push(void* arr, void* xptr);
-void _dynarray_pop(void* arr, void* dest);
+// OK = 0; ERR = 1
+int _dynarray_pop(void* arr, void* dest);
+void _dynarray_reverse(void* arr);
 
 #define DYNARRAY_DEFAULT_CAP 1
 #define DYNARRAY_RESIZE_FACTOR 2
@@ -41,6 +43,7 @@ void _dynarray_pop(void* arr, void* dest);
     } while (0)
 
 #define dynarray_pop(arr, xptr) _dynarray_pop(arr, xptr)
+#define dynarray_reverse(arr) _dynarray_reverse(arr)
 
 #define dynarray_capacity(arr) _dynarray_field_get(arr, CAPACITY)
 #define dynarray_length(arr) _dynarray_field_get(arr, LENGTH)

src/include/logger.h

@@ -0,0 +1,20 @@
+#ifndef _H_MORPH_LOG
+#define _H_MORPH_LOG
+
+#include <loc.h>
+
+typedef enum log_level_e {
+    ERROR,
+    WARN,
+    INFO,
+    DEBUG,
+} log_level_t;
+
+void _log(loc_t* loc, log_level_t level, const char* fmt, ...);
+
+#define log_error(loc, ...) _log(loc, ERROR, __VA_ARGS__)
+#define log_warn(loc, ...) _log(loc, WARN, __VA_ARGS__)
+#define log_info(loc, ...) _log(loc, INFO, __VA_ARGS__)
+#define log_debug(loc, ...) _log(loc, DEBUG, __VA_ARGS__)
+
+#endif // !_H_MORPH_LOG

src/include/parser.h

@@ -0,0 +1,8 @@
+#ifndef _H_MORPH_PARSER
+#define _H_MORPH_PARSER
+
+#include "ast.h"
+#include "token.h"
+program_t parse(const token_t* tokens);
+
+#endif // _H_MORPH_PARSER

src/include/token.h

@@ -33,6 +33,7 @@ typedef enum kw_type_e {
     KW_WHILE,
     KW_CONST,
     KW_MEMORY,
+    KW_INCLUDE,
     KW_COUNT__,
 } kw_type_t;
 
@@ -119,4 +120,6 @@ typedef struct token_s {
 
 extern const char* OP_LIST[];
 extern const char* KW_LIST[];
+const char* get_tok_str_dbg(token_t* tok);
+const char* get_tok_type_str_dbg(token_type_t tok);
 #endif // _H_MORPH_TOKEN

src/logger.c

@@ -0,0 +1,27 @@
+#include "loc.h"
+#include <logger.h>
+#include <stdarg.h>
+#include <stdio.h>
+
+#define ERROR_PFX "\x1b[1;31merror\x1b[0m"
+#define WARN_PFX "\x1b[1;33mwarn\x1b[0m"
+#define INFO_PFX "\x1b[1;32minfo\x1b[0m"
+#define DEBUG_PFX "\x1b[1;34mdebug\x1b[0m"
+
+void _log(loc_t* loc, log_level_t level, const char* fmt, ...) {
+    va_list args;
+    va_start(args, fmt);
+    if (loc) {
+        printf("%s:%d:%d ", loc->file, loc->line, loc->col);
+    }
+    switch (level) { // clang-format off
+    case (ERROR): printf(ERROR_PFX); break;
+    case (WARN): printf(WARN_PFX); break;
+    case (INFO): printf(INFO_PFX); break;
+    case (DEBUG): printf(DEBUG_PFX); break;
+    } // clang-format on
+    printf(": ");
+    vprintf(fmt, args);
+    printf("\n");
+    va_end(args);
+}

src/parser.c

@@ -0,0 +1,131 @@
+#include "logger.h"
+#include "tokeniser.h"
+#include "util.h"
+#include <assert.h>
+#include <stdlib.h>
+#include <token.h>
+#include <ast.h>
+#include <dynarray.h>
+#include <parser.h>
+
+typedef struct parser_state_s {
+    program_t prog;
+    token_t curr_tok;
+    token_t* tokens;
+} parser_state_t;
+
+token_t* expect_token_type(parser_state_t* state, token_type_t type) {
+    const char* exoected_type_str = get_tok_type_str_dbg(state->curr_tok.type);
+    if (dynarray_pop(state->tokens, &state->curr_tok) != 0) {
+        log_error(&state->curr_tok.loc, "Invalid word, expected %s, got nothing.", exoected_type_str);
+        return NULL;
+    }
+    if (state->curr_tok.type != type) {
+        const char* s1 = get_tok_str_dbg(&state->curr_tok);
+        log_error(&state->curr_tok.loc, "Invalid word, expected %s, got %s.", exoected_type_str, s1);
+        return NULL;
+    }
+    return &state->curr_tok;
+}
+
+token_t* expect_token_type_ex(parser_state_t* state, token_type_t type, int inner_type) {
+    const char* exoected_type_str = get_tok_type_str_dbg(state->curr_tok.type);
+    if (dynarray_pop(state->tokens, &state->curr_tok) != 0) {
+        log_error(&state->curr_tok.loc, "Invalid word, expected %s, got nothing.", exoected_type_str);
+        return NULL;
+    }
+    // NOTE: all inner types are in a union
+    if (state->curr_tok.type != type || state->curr_tok.kw_type != inner_type) {
+        const char* s1 = get_tok_str_dbg(&state->curr_tok);
+        log_error(&state->curr_tok.loc, "Invalid word, expected %s, got %s.", exoected_type_str, s1);
+        return NULL;
+    }
+    return &state->curr_tok;
+}
+
+// Tests token type, and inner type, so like, op type, kw type, etc.
+// Does not edit state.curr_tok
+token_t* test_token_type_ex(parser_state_t* state, token_type_t tok_type, int gen_type) {
+    const char* exoected_type_str = get_tok_type_str_dbg(state->curr_tok.type);
+    token_t* token = &state->tokens[dynarray_length(state->tokens - 1)];
+    // NOTE: tests inner type in a generic way
+    if (token->type != tok_type || ((int)token->kw_type) != gen_type) {
+        return NULL;
+    }
+    return token;
+}
+
+token_t* test_token_type(parser_state_t* state, token_type_t tok_type) {
+    const char* exoected_type_str = get_tok_type_str_dbg(state->curr_tok.type);
+    token_t* token = &state->tokens[dynarray_length(state->tokens - 1)];
+    if (token->type != tok_type) {
+        return NULL;
+    }
+    return token;
+}
+
+int _parse(parser_state_t* state);
+
+program_t parse(const token_t* tokens) {
+    // reverses tokens so its way easier to parse;
+    dynarray_reverse((void*)tokens);
+    parser_state_t state = {0};
+    state.tokens = (token_t*)tokens;
+    (void)tokens;
+    _parse(&state);
+    return state.prog;
+}
+
+int _parse(parser_state_t* state) {
+    while (dynarray_pop(state->tokens, &state->curr_tok) == 0) {
+        if (state->curr_tok.type != TT_KW) {
+            const char* dbg_s = get_tok_str_dbg(&state->curr_tok);
+            log_error(&state->curr_tok.loc, "Invalid word, expected Keyword, got %s.", dbg_s);
+            free((void*)dbg_s);
+        }
+        switch (state->curr_tok.kw_type) {
+        case (KW_INCLUDE): {
+            expect_token_type(state, TT_PUSH_STR);
+            const char* code = read_to_string(state->curr_tok.str_v);
+            if (!code) {
+                return 1;
+            }
+            token_t* tokens = tokenise_string((char*)state->curr_tok.str_v, (char*)code);
+            if (!tokens) {
+                return 1;
+            }
+            if (_parse(state) != 0) {
+                return 1;
+            }
+        }; break;
+        case (KW_CONST): {
+            // TODO: Implement compile time calculation of the value
+            token_t* tokens = dynarray_create(token_t);
+
+            while (test_token_type_ex(state, TT_KW, KW_END) == NULL) {
+                // clang-format off
+                if (
+                    test_token_type(state, TT_PUSH_STR) || test_token_type(state, TT_PUSH_CSTR) || 
+                    test_token_type(state, TT_PUSH_CHAR) || test_token_type(state, TT_PUSH_INT) || 
+                    test_token_type(state, TT_PUSH_FLOAT) || test_token_type(state, TT_PUSH_MEM)
+                ) { // clang-format on
+                    dynarray_pop(state->tokens, &state->curr_tok);
+                    dynarray_push(tokens, state->curr_tok);
+                }
+            }
+            if (!expect_token_type_ex(state, TT_KW, KW_END)) {
+                return 1;
+            }
+
+        }; break;
+        case (KW_MEMORY): {
+
+        }; break;
+        case (KW_FN): {
+
+        }; break;
+        default:
+            assert(true && "TODO: parse all kw's");
+        }
+    }
+}

src/token.c

@@ -1,3 +1,7 @@
+#include <assert.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
 #include <token.h>
 
 // clang-format off
@@ -59,7 +63,65 @@ const char* KW_LIST[] = {
     [KW_ELSE] = "else",
     [KW_WHILE] = "while",
     [KW_CONST] = "const",
-    [KW_MEMORY] = "memory"
+    [KW_MEMORY] = "memory",
+    [KW_INCLUDE] = "include",
 };
 
+const char* get_tok_type_str_dbg(token_type_t tok_t) {
+    switch (tok_t) {
+        case (TT_KW): return "Keyword";
+        case (TT_OP): return "Operator";
+        case (TT_IDENT): return "Identifier";
+        case (TT_PUSH_CHAR): return "Char Literal";
+        case (TT_PUSH_INT): return "Intager";
+        case (TT_PUSH_STR): return "String";
+        case (TT_PUSH_CSTR): return "CString";
+        case (TT_PUSH_MEM): return "Memory address";
+        case (TT_PUSH_FLOAT): return "Float";
+        case (TT_NONE): assert(true && "Invalid");
+    }
+    return "Unreachable";
+}
+const char* get_tok_str_dbg(token_t* tok) {
+    int buf_size = 1024*4;
+    char* buf = (char*)malloc(buf_size);
+    switch (tok->type) {
+        case (TT_KW): {
+            snprintf(buf, buf_size, "Keyword '%s'", KW_LIST[tok->kw_type]);
+        }; break;
+        case (TT_OP): {
+            snprintf(buf, buf_size, "Operator '%s'", OP_LIST[tok->op_type]);
+        }; break;
+        case (TT_IDENT): {
+            snprintf(buf, buf_size, "Identifier '%s'", tok->str_v);
+        }; break;
+        case (TT_PUSH_CHAR): {
+            snprintf(buf, buf_size, "'%c'", tok->char_v);
+        }; break;
+        case (TT_PUSH_INT): {
+            snprintf(buf, buf_size, "%zu", tok->int_v);
+        }; break;
+        case (TT_PUSH_STR): {
+            while (strlen(tok->str_v) > buf_size-2) {
+                buf = realloc(buf, buf_size*=2);
+            }
+            snprintf(buf, buf_size, "\"%s\"", tok->str_v);
+        }; break;
+        case (TT_PUSH_CSTR): {
+            while (strlen(tok->str_v) > buf_size-3) {
+                buf = realloc(buf, buf_size*=2);
+            }
+            snprintf(buf, buf_size, "c\"%s\"", tok->str_v);
+        }; break;
+        case (TT_PUSH_MEM): {
+            snprintf(buf, buf_size, "Memory address label '%s'", tok->str_v);
+        }; break;
+        case (TT_PUSH_FLOAT): {
+            snprintf(buf, buf_size, "%f", tok->float_v);
+        }; break;
+        case (TT_NONE): assert(true && "Invalid");
+    }
+    return buf;
+}
+
 // clang-format on

src/tokeniser.c

@@ -1,4 +1,5 @@
 #include <limits.h>
+#include <math.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
@@ -78,9 +79,65 @@ token_t* tokenise_string(char* file_path, char* str) {
             loc.line += 1;
             continue;
         }
+        case ('\''): {
+            char* buf = malloc(32);
+            bool escaped = false;
+            for (int y = 0;;) {
+                if (str[i] == '\'' && !escaped) {
+                    break;
+                }
+                if (str[i] == '\\' && !escaped) {
+                    escaped = true;
+                    continue;
+                }
+                if (escaped) {
+                    switch (str[i]) {
+                    case ('r'):
+                        buf[0] = '\r';
+                        break;
+                    case ('\''):
+                        buf[0] = '\'';
+                        break;
+                    case ('n'):
+                        buf[0] = '\n';
+                        break;
+                    case ('t'):
+                        buf[0] = '\t';
+                        break;
+                    case ('\\'):
+                        buf[0] = '\\';
+                        break;
+                    }
+                    y = 1;
+                }
+                if (str[i] == '\n') {
+                    printf("%s:%d:%d: Error: Multi line chars not supported\n", loc.file, loc.line, loc.col);
+                    return NULL;
+                }
+                if (y > 0) {
+                    printf("%s:%d:%d: Error: Character literals must only have 1 char inside of them\n", loc.file, loc.line, loc.col);
+                    return NULL;
+                }
+                loc.col += 1;
+                buf[y++] = str[i++];
+            }
+
+            token_t tok = {
+                .type = TT_PUSH_CHAR,
+                .str_v = buf,
+            };
+            dynarray_push(tokens, tok);
+        } break;
         case ('c'):
         case ('"'): {
+            bool is_cstr = false;
             if (str[i] == 'c') {
+                if (str[i + 1] != '"') {
+                    goto tokenise_ident;
+                }
+                i += 1;
+                loc.col += 1;
+                is_cstr = true;
             }
             char* buf = malloc(1024 * 4);
             bool escaped = false;
@@ -95,16 +152,22 @@ token_t* tokenise_string(char* file_path, char* str) {
                 if (str[i] == '\\' && escaped) {
                     escaped = false;
                 }
+                if (str[i] == '\n') {
+                    printf("%s:%d:%d: Error: Multi line strings not supported\n", loc.file, loc.line, loc.col);
+                    return NULL;
+                }
+                loc.col += 1;
                 buf[y++] = str[i++];
             }
 
             token_t tok = {
-                .type = TT_PUSH_STR,
+                .type = is_cstr ? TT_PUSH_CSTR : TT_PUSH_STR,
                 .str_v = buf,
             };
             dynarray_push(tokens, tok);
         } break;
         default: {
+        tokenise_ident:
             // TODO: manage memory better
             // INFO: If you make an ident longer than 4kb i will murder you
             char* buf = malloc(1024 * 4);
@@ -115,6 +178,7 @@ token_t* tokenise_string(char* file_path, char* str) {
                     break;
                 }
                 buf[buf_counter++] = str[i];
+                loc.col += 1;
                 i += 1;
             }
             buf[buf_counter + 1] = '\0';