diff options
Diffstat (limited to 'lexer.c')
| -rw-r--r-- | lexer.c | 127 |
1 files changed, 121 insertions, 6 deletions
@@ -2,6 +2,97 @@ #include <stdlib.h> #include <ctype.h> #include <stdio.h> +#include <string.h> + +struct keywords_cell { + token_type type; + struct keywords_cell *next; + char *keyword; + size_t keyword_length; +}; + +static struct keywords_cell *keywords[TOKEN_VOLATILE - TOKEN_IF] = {0}; + +static size_t hash_keyword(char *keyword, size_t length) +{ + size_t sum = 0; + for (size_t i=0; i < length; i++) { + sum += (uint8_t) keyword[i]; + } + + return sum % (TOKEN_VOLATILE - TOKEN_IF); +} + +static void insert_keyword(char *keyword, size_t length, token_type type) +{ + struct keywords_cell *next = keywords[hash_keyword(keyword, length)]; + if (!next) { + keywords[hash_keyword(keyword, length)] = calloc(1, sizeof(struct keywords_cell)); + keywords[hash_keyword(keyword, length)]->keyword = keyword; + keywords[hash_keyword(keyword, length)]->keyword_length = length; + keywords[hash_keyword(keyword, length)]->type = type; + keywords[hash_keyword(keyword, length)]->next = NULL; + return; + } + + while (next->next) { + next = next->next; + } + + next->next = calloc(1, sizeof(struct keywords_cell)); + next->next->keyword = keyword; + next->next->keyword_length = length; + next->next->type = type; + next->next->next = NULL; +} + +static token_type get_keyword(char *keyword, size_t length) +{ + struct keywords_cell *cell = keywords[hash_keyword(keyword, length)]; + if (!cell) return 0; + if (cell->next == NULL) return cell->type; + while (cell->next != NULL) { + if (cell->keyword_length != length) { + cell = cell->next; + continue; + } + if (strncmp(cell->keyword, keyword, length) == 0) { + return cell->type; + } + cell = cell->next; + } + + if (strncmp(cell->keyword, keyword, length) == 0) { + return cell->type; + } + + return 0; +} + +static void populate_keywords(void) +{ + insert_keyword("if", 2, TOKEN_IF); + insert_keyword("else", 4, TOKEN_ELSE); + insert_keyword("while", 5, TOKEN_WHILE); + insert_keyword("for", 3, TOKEN_FOR); + insert_keyword("break", 5, TOKEN_BREAK); + insert_keyword("case", 4, TOKEN_CASE); + insert_keyword("const", 5, TOKEN_CONST); + insert_keyword("do", 2, TOKEN_DO); + insert_keyword("enum", 4, TOKEN_ENUM); + insert_keyword("extern", 6, TOKEN_EXTERN); + insert_keyword("goto", 4, TOKEN_GOTO); + insert_keyword("inline", 6, TOKEN_INLINE); + insert_keyword("return", 6, TOKEN_RETURN); + insert_keyword("signed", 6, TOKEN_SIGNED); + insert_keyword("unsigned", 8, TOKEN_UNSIGNED); + insert_keyword("static", 6, TOKEN_STATIC); + insert_keyword("struct", 6, TOKEN_STRUCT); + insert_keyword("switch", 6, TOKEN_SWITCH); + insert_keyword("typedef", 7, TOKEN_TYPEDEF); + insert_keyword("union", 5, TOKEN_UNION); + insert_keyword("volatile", 8, TOKEN_VOLATILE); +} static token *token_create(char *lexeme, size_t length, token_type type) { @@ -151,24 +242,34 @@ size_t check_symbol(char *src, size_t index, token **last) } case '{': token_add(last, token_create(src+index, 1, TOKEN_LCURLY)); + return 1; case '}': token_add(last, token_create(src+index, 1, TOKEN_RCURLY)); + return 1; case '(': token_add(last, token_create(src+index, 1, TOKEN_LPAREN)); + return 1; case ')': token_add(last, token_create(src+index, 1, TOKEN_RPAREN)); + return 1; case '[': token_add(last, token_create(src+index, 1, TOKEN_LSQUARE)); + return 1; case ']': token_add(last, token_create(src+index, 1, TOKEN_RSQUARE)); + return 1; case ':': token_add(last, token_create(src+index, 1, TOKEN_COLON)); + return 1; case ';': token_add(last, token_create(src+index, 1, TOKEN_SEMICOLON)); + return 1; case ',': token_add(last, token_create(src+index, 1, TOKEN_COMMA)); + return 1; case '.': token_add(last, token_create(src+index, 1, TOKEN_DOT)); + return 1; default: return 0; @@ -177,10 +278,18 @@ size_t check_symbol(char *src, size_t index, token **last) token *lexer_parse(char *src) { + populate_keywords(); size_t index = 0; token *head = token_create(src, 0, TOKEN_HEAD); token *last = head; + size_t row = 1, col = 1; while (src[index] != '\0') { + if (src[index] == '\n') { + index++; + col++; + row = 1; + continue; + } if (isspace(src[index])) { index++; continue; @@ -194,7 +303,7 @@ token *lexer_parse(char *src) if (src[index] == '\'') { if (src[index+1] != '\\' && src[index+2] != '\'') { - // TODO: error + fprintf(stderr, "%zu:%zu: invalid character\n", row, col); return NULL; } token_add(&last, token_create(src+index+1, 1, TOKEN_CHAR)); @@ -208,7 +317,7 @@ token *lexer_parse(char *src) size_t string_length = index; while (src[index++] != '\"') { if (src[index] == '\0') { - // TODO: error + fprintf(stderr, "%zu:%zu: unterminated string\n", row, col); return NULL; } } @@ -224,7 +333,13 @@ token *lexer_parse(char *src) index++; } identifier_length = index - identifier_length; - token_add(&last, token_create(identifier_start, identifier_length, TOKEN_IDENTIFIER)); + + token_type keyword = get_keyword(identifier_start, identifier_length); + if (keyword == 0) { + token_add(&last, token_create(identifier_start, identifier_length, TOKEN_IDENTIFIER)); + } else { + token_add(&last, token_create(identifier_start, identifier_length, keyword)); + } continue; } @@ -233,7 +348,7 @@ token *lexer_parse(char *src) size_t number_length = index; uint8_t is_float = 0; - if (src[index] == '0' && !isdigit(src[index+1]) && src[index+1] != '.') { + if (src[index] == '0' && !isdigit(src[index+1]) && src[index+1] != '.' && (isdigit(src[index+2]) || ((src[index+2] >= 65 && src[index+2] <= 70) && (src[index+2] >= 97 && src[index+2] <= 102)))) { if (src[index+1] == 'x') { index += 2; } else if (src[index+1] == 'o') { @@ -241,7 +356,7 @@ token *lexer_parse(char *src) } else if (src[index+1] == 'b') { index += 2; } else { - // TODO: error + fprintf(stderr, "%zu:%zu: invalid integer\n", row, col); return NULL; } } @@ -249,7 +364,7 @@ token *lexer_parse(char *src) while (isdigit(src[index]) || src[index] == '.') { if (src[index] == '.') { if (is_float) { - // TODO: error + fprintf(stderr, "%zu:%zu: invalid float\n", row, col); return NULL; } is_float = 1; |