#include #include #include #include #include "include/lexer.h" #include "include/dstr.h" #include "include/token.h" #include "include/util.h" Lexer* thelexer = NULL; void lexer_init(char* src) { thelexer = malloc(sizeof(Lexer)); thelexer->src = src; thelexer->srcln = strlen(src); thelexer->cchar = thelexer->src; thelexer->tokens = calloc(TOKENS_MAX, sizeof(Token*)); thelexer->ntokens = 0; thelexer->state = LEXER_STATE_CONFUSED; log_dbgf("created thelexer @ %p", thelexer); } void lexer_destroy(Lexer* lexer) { // Does not free lexer->src. for (int i = 0; i < lexer->ntokens; i++) token_destroy(lexer->tokens[i]); } void lexer_lex(Lexer* lexer) { while (*lexer->cchar) { switch (lexer->state) { case LEXER_STATE_CONFUSED: lexer_do_confused(lexer); break; case LEXER_STATE_NUM: lexer_do_number(lexer); break; case LEXER_STATE_CALL: lexer_do_call(lexer); break; default: break; } } } void lexer_do_confused(Lexer* lexer) { log_dbgf("lexer @ %p entered confused mode @ char '%c' (%d)", lexer, *lexer->cchar, (int)*lexer->cchar); if (isspace(*lexer->cchar)) lexer_inc(lexer); if (isdigit(*lexer->cchar)) { lexer->state = LEXER_STATE_NUM; lexer_do_number(lexer); } else { lexer->state = LEXER_STATE_CALL; lexer_do_call(lexer); } } void lexer_do_number(Lexer* lexer) { log_dbgf("lexer @ %p entered number mode @ char '%c' (%d)", lexer, *lexer->cchar, (int)*lexer->cchar); // Length of the number string. size_t numln; // Where the number string starts. char* start = lexer->cchar; for (numln = 0; *lexer->cchar && isdigit(*lexer->cchar); numln++) lexer_inc(lexer); char* num = malloc(numln + 1); memcpy(num, start, numln); num[numln] = '\0'; lexer_add_token(lexer, token_init(TOKEN_TYPE_NUMBER, num, numln)); lexer->state = LEXER_STATE_CONFUSED; } void lexer_do_call(Lexer* lexer) { log_dbgf("lexer @ %p entered call mode @ char '%c' (%d)", lexer, *lexer->cchar, (int)*lexer->cchar); // Size of the call string. size_t callln; // Where the call string starts. char* start = lexer->cchar; for (callln = 0; *lexer->cchar && (!isdigit(*lexer->cchar) && !isspace(*lexer->cchar)); callln++) lexer_inc(lexer); char* call = malloc(callln + 1); memcpy(call, start, callln); call[callln] = '\0'; lexer_add_token(lexer, token_init(TOKEN_TYPE_CALL, call, callln)); lexer->state = LEXER_STATE_CONFUSED; } void lexer_inc(Lexer* lexer) { lexer->cchar += sizeof(char); } void lexer_add_token(Lexer* lexer, Token* token) { assert(lexer->ntokens < TOKENS_MAX); if (lexer->ntokens < TOKENS_MAX - 1) { lexer->tokens[lexer->ntokens] = token; lexer->ntokens++; log_dbgf("added token (total: %ld)", lexer->ntokens); } } void lexer_print(Lexer* lexer) { lexer_print_i(lexer, 0); } void lexer_print_i(Lexer* lexer, int ilvl) { INDENT_BEGIN(ilvl); INDENT_TITLE("Lexer", lexer); INDENT_FIELD_NONL_START("state") lexerstate_print_raw(lexer->state); INDENT_FIELD_NONL_END INDENT_FIELD("srcln", "%ld", lexer->srcln); INDENT_FIELD_NL("src", "\"%s\"", lexer->src); INDENT_FIELD("cchar", "'%c'", *lexer->cchar); INDENT_FIELD("ntokens", "%ld", lexer->ntokens); INDENT_FIELD_LIST("tokens", lexer->tokens, lexer->ntokens, token_print_i); #if 0 printf("%s tokens: [\n", INDENT_spacing->buf); for (int i = 0; i < lexer->ntokens; i++) { token_print_i(lexer->tokens[i], ilvl + 2); printf(",\n\n"); } #endif } void lexerstate_print_raw(LexerState s) { if (s > LEXER_STATE_MAX) { printf("Unknown (%d)", s); log_dbgf("%d is not a valid LexerState (max: %d)", s, TOKEN_TYPE_MAX); } else printf("%s", lexerstate_names[s]); } #include "../build/grammars/grammar.tab.h" int yylex(void) { if (*thelexer->cchar == '\0') return YYEOF; switch (*thelexer->cchar) { case ' ': case '\t': thelexer->cchar++; } // Assign & consume current character. int c = *thelexer->cchar++; switch (c) { case '+': return PLUS; } if (isdigit(c)) { int value = c - '0'; // Start with the first digit while (isdigit(*thelexer->cchar)) { value = value * 10 + (*thelexer->cchar - '0'); // Accumulate value thelexer++; } yylval.intval = value; // Set the token value return NUM; // Return the INTEGER token type } fprintf(stderr, "Unexpected character: %c\n", c); return 0; }