#include <config.h>
-#include "lexer_t.h"
+#include "lexer.h"
#include "token_t.h"
#include "symbol_table_t.h"
#include "adt/error.h"
+#include "adt/strset.h"
#include <assert.h>
#include <errno.h>
#include <ctype.h>
//#define DEBUG_CHARS
-#define MAX_PUTBACK 1
+#define MAX_PUTBACK 3
+
+static int c;
+token_t lexer_token;
+static FILE *input;
+static char buf[1024 + MAX_PUTBACK];
+static const char *bufend;
+static const char *bufpos;
+static strset_t stringset;
+//static FILE **input_stack;
+//static char **buf_stack;
static
-void error_prefix_at(lexer_t *this, const char *input_name, unsigned linenr)
+void error_prefix_at(const char *input_name, unsigned linenr)
{
- (void) this;
fprintf(stderr, "%s:%d: Error: ", input_name, linenr);
}
static
-void error_prefix(lexer_t *this)
+void error_prefix(void)
{
- error_prefix_at(this, this->source_position.input_name,
- this->source_position.linenr);
+ error_prefix_at(lexer_token.source_position.input_name,
+ lexer_token.source_position.linenr);
}
static
-void parse_error(lexer_t *this, const char *msg)
+void parse_error(const char *msg)
{
- error_prefix(this);
+ error_prefix();
fprintf(stderr, "%s\n", msg);
}
static inline
-void next_char(lexer_t *this)
+void next_char(void)
{
- this->bufpos++;
- if(this->bufpos >= this->bufend) {
- size_t s = fread(this->buf + MAX_PUTBACK, 1,
- sizeof(this->buf) - MAX_PUTBACK, this->input);
+ bufpos++;
+ if(bufpos >= bufend) {
+ size_t s = fread(buf + MAX_PUTBACK, 1, sizeof(buf) - MAX_PUTBACK,
+ input);
if(s == 0) {
- this->c = EOF;
+ c = EOF;
return;
}
- this->bufpos = this->buf + MAX_PUTBACK;
- this->bufend = this->buf + MAX_PUTBACK + s;
+ bufpos = buf + MAX_PUTBACK;
+ bufend = buf + MAX_PUTBACK + s;
}
- this->c = *(this->bufpos);
+ c = *(bufpos);
#ifdef DEBUG_CHARS
- printf("nchar '%c'\n", this->c);
+ printf("nchar '%c'\n", c);
#endif
}
static inline
-void put_back(lexer_t *this, int c)
+void put_back(int pc)
{
- char *p = (char*) this->bufpos - 1;
- this->bufpos--;
- assert(p >= this->buf);
- *p = c;
+ char *p = (char*) bufpos - 1;
+ bufpos--;
+ assert(p >= buf);
+ *p = pc;
#ifdef DEBUG_CHARS
- printf("putback '%c'\n", c);
+ printf("putback '%c'\n", pc);
#endif
}
+
static
-int replace_trigraph(lexer_t *this)
+int replace_trigraph(void)
{
#define MATCH_TRIGRAPH(ch,replacement) \
case ch: \
- this->c = replacement; \
+ c = replacement; \
return 1;
- switch(this->c) {
+ switch(c) {
MATCH_TRIGRAPH('=', '#')
MATCH_TRIGRAPH('(', '[')
MATCH_TRIGRAPH('/', '\\')
return 0;
}
+#define SKIP_TRIGRAPHS(custom_putback, no_trigraph_code) \
+ case '?': \
+ next_char(); \
+ if(c != '?') { \
+ custom_putback; \
+ put_back(c); \
+ c = '?'; \
+ no_trigraph_code; \
+ } \
+ next_char(); \
+ if(replace_trigraph()) { \
+ break; \
+ } \
+ custom_putback; \
+ put_back('?'); \
+ put_back(c); \
+ c = '?'; \
+ no_trigraph_code;
+
+#define EAT_NEWLINE(newline_code) \
+ if(c == '\r') { \
+ next_char(); \
+ if(c == '\n') \
+ next_char(); \
+ lexer_token.source_position.linenr++; \
+ newline_code; \
+ } else if(c == '\n') { \
+ next_char(); \
+ lexer_token.source_position.linenr++; \
+ newline_code; \
+ }
+
+#define SYMBOL_CHARS \
+ case 'a': \
+ case 'b': \
+ case 'c': \
+ case 'd': \
+ case 'e': \
+ case 'f': \
+ case 'g': \
+ case 'h': \
+ case 'i': \
+ case 'j': \
+ case 'k': \
+ case 'l': \
+ case 'm': \
+ case 'n': \
+ case 'o': \
+ case 'p': \
+ case 'q': \
+ case 'r': \
+ case 's': \
+ case 't': \
+ case 'u': \
+ case 'v': \
+ case 'w': \
+ case 'x': \
+ case 'y': \
+ case 'z': \
+ case 'A': \
+ case 'B': \
+ case 'C': \
+ case 'D': \
+ case 'E': \
+ case 'F': \
+ case 'G': \
+ case 'H': \
+ case 'I': \
+ case 'J': \
+ case 'K': \
+ case 'L': \
+ case 'M': \
+ case 'N': \
+ case 'O': \
+ case 'P': \
+ case 'Q': \
+ case 'R': \
+ case 'S': \
+ case 'T': \
+ case 'U': \
+ case 'V': \
+ case 'W': \
+ case 'X': \
+ case 'Y': \
+ case 'Z': \
+ case '_':
+
+#define DIGITS \
+ case '0': \
+ case '1': \
+ case '2': \
+ case '3': \
+ case '4': \
+ case '5': \
+ case '6': \
+ case '7': \
+ case '8': \
+ case '9':
+
static
-void parse_symbol(lexer_t *this, token_t *token)
+void parse_symbol(void)
{
symbol_t *symbol;
char *string;
- obstack_1grow(&symbol_obstack, this->c);
- next_char(this);
+ obstack_1grow(&symbol_obstack, c);
+ next_char();
while(1) {
- switch(this->c) {
+ switch(c) {
case '\\':
- next_char(this);
- if(this->c == '\n') {
- next_char(this);
- this->source_position.linenr++;
- break;
- }
+ next_char();
+ EAT_NEWLINE(break;)
+ goto end_symbol;
- case 'A' ... 'Z':
- case 'a' ... 'z':
- case '_':
- obstack_1grow(&symbol_obstack, this->c);
- next_char(this);
+ DIGITS
+ SYMBOL_CHARS
+ obstack_1grow(&symbol_obstack, c);
+ next_char();
break;
case '?':
- next_char(this);
- if(this->c != '?') {
- put_back(this, this->c);
- this->c = '?';
+ next_char();
+ if(c != '?') {
+ put_back(c);
+ c = '?';
goto end_symbol;
}
- next_char(this);
- if(replace_trigraph(this))
+ next_char();
+ if(replace_trigraph())
break;
- put_back(this, '?');
- put_back(this, this->c);
- this->c = '?';
+ put_back('?');
+ put_back(c);
+ c = '?';
goto end_symbol;
default:
string = obstack_finish(&symbol_obstack);
symbol = symbol_table_insert(string);
- if(symbol->ID > 0) {
- token->type = symbol->ID;
- } else {
- token->type = T_IDENTIFIER;
- }
- token->v.symbol = symbol;
+ lexer_token.type = symbol->ID;
+ lexer_token.v.symbol = symbol;
if(symbol->string != string) {
obstack_free(&symbol_obstack, string);
}
}
-#if 0
static
-preprocessor_token_type_t parse_pp_symbol(lexer_t *this)
+void parse_number_hex(void)
{
- do {
- obstack_1grow(&symbol_obstack, this->c);
- next_char(this);
- } while(is_ident_char(this->c));
- obstack_1grow(&symbol_obstack, '\0');
-
- char *string = obstack_finish(&symbol_obstack);
- symbol_t *symbol = preprocessor_symbol_table_find(string);
- obstack_free(&symbol_obstack, string);
-
- if(symbol == 0)
- return TP_ERROR;
-
- return symbol->ID;
-}
-#endif
-
-static
-void parse_number_hex(lexer_t *this, token_t *token)
-{
- assert(this->c == 'x' || this->c == 'X');
- next_char(this);
-
- if (!isdigit(this->c) &&
- !('A' <= this->c && this->c <= 'F') &&
- !('a' <= this->c && this->c <= 'f')) {
- parse_error(this, "premature end of hex number literal");
- token->type = T_ERROR;
+ assert(c == 'x' || c == 'X');
+ next_char();
+
+ if (!isdigit(c) &&
+ !('A' <= c && c <= 'F') &&
+ !('a' <= c && c <= 'f')) {
+ parse_error("premature end of hex number literal");
+ lexer_token.type = T_ERROR;
return;
}
int value = 0;
for(;;) {
- if (isdigit(this->c)) {
- value = 16 * value + this->c - '0';
- } else if ('A' <= this->c && this->c <= 'F') {
- value = 16 * value + this->c - 'A' + 10;
- } else if ('a' <= this->c && this->c <= 'f') {
- value = 16 * value + this->c - 'a' + 10;
+ if (isdigit(c)) {
+ value = 16 * value + c - '0';
+ } else if ('A' <= c && c <= 'F') {
+ value = 16 * value + c - 'A' + 10;
+ } else if ('a' <= c && c <= 'f') {
+ value = 16 * value + c - 'a' + 10;
} else {
- token->type = T_INTEGER;
- token->v.intvalue = value;
+ lexer_token.type = T_INTEGER;
+ lexer_token.v.intvalue = value;
return;
}
- next_char(this);
+ next_char();
}
}
static
-void parse_number_oct(lexer_t *this, token_t *token)
+void parse_number_oct(void)
{
- assert(this->c == 'o' || this->c == 'O');
- next_char(this);
+ assert(c == 'o' || c == 'O');
+ next_char();
int value = 0;
for(;;) {
- if ('0' <= this->c && this->c <= '7') {
- value = 8 * value + this->c - '0';
+ if ('0' <= c && c <= '7') {
+ value = 8 * value + c - '0';
} else {
- token->type = T_INTEGER;
- token->v.intvalue = value;
+ lexer_token.type = T_INTEGER;
+ lexer_token.v.intvalue = value;
return;
}
- next_char(this);
+ next_char();
}
}
static
-void parse_number_dec(lexer_t *this, token_t *token, int first_char)
+void parse_number_dec(int first_char)
{
int value = 0;
if(first_char > 0) {
}
for(;;) {
- if (isdigit(this->c)) {
- value = 10 * value + this->c - '0';
+ if (isdigit(c)) {
+ value = 10 * value + c - '0';
} else {
- token->type = T_INTEGER;
- token->v.intvalue = value;
+ lexer_token.type = T_INTEGER;
+ lexer_token.v.intvalue = value;
return;
}
- next_char(this);
+ next_char();
}
}
static
-void parse_number(lexer_t *this, token_t *token)
+void parse_number(void)
{
// TODO check for overflow
// TODO check for various invalid inputs sequences
- if (this->c == '0') {
- next_char(this);
- switch (this->c) {
+ if (c == '0') {
+ next_char();
+ switch (c) {
case 'X':
- case 'x': parse_number_hex(this, token); break;
+ case 'x': parse_number_hex(); break;
case 'o':
- case 'O': parse_number_oct(this, token); break;
- default: parse_number_dec(this, token, '0');
+ case 'O': parse_number_oct(); break;
+ default: parse_number_dec('0');
}
} else {
- parse_number_dec(this, token, 0);
+ parse_number_dec(0);
}
}
static
-int parse_escape_sequence(lexer_t *this)
+int parse_escape_sequence(void)
{
- int c = this->c;
- next_char(this);
+ while(1) {
+ int ec = c;
+ next_char();
- switch(c) {
- case '"': return '"';
- case '\'': return'\'';
- case '\\':
- if(this->c == '\n') {
- this->source_position.linenr++;
- next_char(this);
- return parse_escape_sequence(this);
- }
- return '\\';
- case 'a': return '\a';
- case 'b': return '\b';
- case 'f': return '\f';
- case 'n': return '\n';
- case 'r': return '\r';
- case 't': return '\t';
- case 'v': return '\v';
- case 'x': /* TODO parse hex number ... */
- parse_error(this, "hex escape sequences not implemented yet");
- return EOF;
- case 0 ... 8: /* TODO parse octal number ... */
- parse_error(this, "octal escape sequences not implemented yet");
- return EOF;
- case '?':
- if(this->c != '?') {
+ switch(ec) {
+ case '"': return '"';
+ case '\'': return'\'';
+ case '\\':
+ EAT_NEWLINE(break;)
+ return '\\';
+ case 'a': return '\a';
+ case 'b': return '\b';
+ case 'f': return '\f';
+ case 'n': return '\n';
+ case 'r': return '\r';
+ case 't': return '\t';
+ case 'v': return '\v';
+ case 'x': /* TODO parse hex number ... */
+ parse_error("hex escape sequences not implemented yet");
+ return EOF;
+ case '0':
+ case '1':
+ case '2':
+ case '3':
+ case '4':
+ case '5':
+ case '6':
+ case '7':
+ /* TODO parse octal number ... */
+ parse_error("octal escape sequences not implemented yet");
+ return EOF;
+ case '?':
+ if(c != '?') {
+ return '?';
+ }
+ /* might be a trigraph */
+ next_char();
+ if(replace_trigraph()) {
+ break;
+ }
+ put_back(c);
+ c = '?';
return '?';
+
+ case EOF:
+ parse_error("reached end of file while parsing escape sequence");
+ return EOF;
+ default:
+ parse_error("unknown escape sequence");
+ return EOF;
}
- /* might be a trigraph */
- next_char(this);
- if(replace_trigraph(this)) {
- return parse_escape_sequence(this);
- }
- put_back(this, '?');
- return '?';
+ }
+}
- case EOF:
- parse_error(this, "reached end of file while parsing escape sequence");
- return EOF;
- default:
- parse_error(this, "unknown escape sequence\n");
- return EOF;
+const char *concat_strings(const char *s1, const char *s2)
+{
+ size_t len1 = strlen(s1);
+ size_t len2 = strlen(s2);
+
+ char *concat = obstack_alloc(&symbol_obstack, len1 + len2 + 1);
+ memcpy(concat, s1, len1);
+ memcpy(concat + len1, s2, len2 + 1);
+
+ const char *result = strset_insert(&stringset, concat);
+ if(result != concat) {
+ obstack_free(&symbol_obstack, concat);
}
+
+ return result;
}
static
-void parse_string_literal(lexer_t *this, token_t *token)
+void parse_string_literal(void)
{
- unsigned start_linenr = this->source_position.linenr;
+ unsigned start_linenr = lexer_token.source_position.linenr;
char *string;
const char *result;
- assert(this->c == '"');
- next_char(this);
+ assert(c == '"');
+ next_char();
while(1) {
- switch(this->c) {
- case '?':
- next_char(this);
- if(this->c != '?') {
- obstack_1grow(&symbol_obstack, '?');
- break;
- }
- next_char(this);
- if(replace_trigraph(this))
- break;
+ switch(c) {
+ SKIP_TRIGRAPHS(,
obstack_1grow(&symbol_obstack, '?');
- put_back(this, this->c);
- this->c = '?';
+ next_char();
break;
+ )
case '\\':
- next_char(this);
- if(this->c == '\n') {
- next_char(this);
- this->source_position.linenr++;
- break;
- }
- int c = parse_escape_sequence(this);
- obstack_1grow(&symbol_obstack, c);
+ next_char();
+ EAT_NEWLINE(break;)
+ int ec = parse_escape_sequence();
+ obstack_1grow(&symbol_obstack, ec);
break;
case EOF:
- error_prefix_at(this, this->source_position.input_name,
+ error_prefix_at(lexer_token.source_position.input_name,
start_linenr);
fprintf(stderr, "string has no end\n");
- token->type = T_ERROR;
+ lexer_token.type = T_ERROR;
return;
case '"':
- next_char(this);
+ next_char();
goto end_of_string;
default:
- obstack_1grow(&symbol_obstack, this->c);
- next_char(this);
+ obstack_1grow(&symbol_obstack, c);
+ next_char();
break;
}
}
string = obstack_finish(&symbol_obstack);
/* check if there is already a copy of the string */
- result = strset_insert(&this->stringset, string);
+ result = strset_insert(&stringset, string);
if(result != string) {
obstack_free(&symbol_obstack, string);
}
- token->type = T_STRING_LITERAL;
- token->v.string = result;
+ lexer_token.type = T_STRING_LITERAL;
+ lexer_token.v.string = result;
+}
+
+#define MATCH_NEWLINE(code) \
+ case '\r': \
+ next_char(); \
+ if(c == '\n') { \
+ next_char(); \
+ } \
+ lexer_token.source_position.linenr++; \
+ code; \
+ case '\n': \
+ next_char(); \
+ lexer_token.source_position.linenr++; \
+ code;
+
+static
+void parse_character_constant(void)
+{
+ assert(c == '\'');
+ next_char();
+
+ int found_char = 0;
+ while(1) {
+ switch(c) {
+ SKIP_TRIGRAPHS(,
+ found_char = '?';
+ break;
+ )
+
+ case '\\':
+ next_char();
+ EAT_NEWLINE(break;)
+ found_char = '\\';
+ break;
+
+ MATCH_NEWLINE(
+ parse_error("newline while parsing character constant");
+ break;
+ )
+
+ case '\'':
+ next_char();
+ goto end_of_char_constant;
+
+ case EOF:
+ parse_error("EOF while parsing character constant");
+ lexer_token.type = T_ERROR;
+ return;
+
+ default:
+ if(found_char != 0) {
+ parse_error("more than 1 characters in character "
+ "constant");
+ goto end_of_char_constant;
+ } else {
+ found_char = c;
+ next_char();
+ }
+ break;
+ }
+ }
+
+end_of_char_constant:
+ lexer_token.type = T_INTEGER;
+ lexer_token.v.intvalue = found_char;
}
static
-void skip_multiline_comment(lexer_t *this)
+void skip_multiline_comment(void)
{
- unsigned start_linenr = this->source_position.linenr;
+ unsigned start_linenr = lexer_token.source_position.linenr;
+ int had_star = 0;
while(1) {
- switch(this->c) {
+ switch(c) {
case '*':
- next_char(this);
- if(this->c == '/') {
- next_char(this);
+ next_char();
+ had_star = 1;
+ break;
+
+ case '/':
+ next_char();
+ if(had_star) {
return;
}
+ had_star = 0;
break;
+
+ case '\\':
+ next_char();
+ EAT_NEWLINE(break;)
+ had_star = 0;
+ break;
+
case '?':
- next_char(this);
- if(this->c != '?')
+ next_char();
+ if(c != '?') {
+ had_star = 0;
break;
- next_char(this);
- if(replace_trigraph(this))
+ }
+ next_char();
+ if(replace_trigraph())
break;
- put_back(this, '?');
+ put_back(c);
+ c = '?';
+ had_star = 0;
/* we don't put back the 2nd ? as the comment text is discarded
* anyway */
break;
- case '\n':
- this->source_position.linenr++;
- next_char(this);
- break;
+ MATCH_NEWLINE(had_star = 0; break;)
+
case EOF:
- error_prefix_at(this, this->source_position.input_name,
+ error_prefix_at(lexer_token.source_position.input_name,
start_linenr);
fprintf(stderr, "at end of file while looking for comment end\n");
return;
default:
- next_char(this);
+ had_star = 0;
+ next_char();
break;
}
}
}
static
-void skip_line_comment(lexer_t *this)
+void skip_line_comment(void)
{
while(1) {
- switch(this->c) {
+ switch(c) {
case '?':
- next_char(this);
- if(this->c != '?')
+ next_char();
+ if(c != '?')
break;
- next_char(this);
- if(replace_trigraph(this))
+ next_char();
+ if(replace_trigraph())
break;
- put_back(this, '?');
+ put_back('?');
/* we don't put back the 2nd ? as the comment text is discarded
* anyway */
break;
case '\\':
- next_char(this);
- if(this->c == '\n') {
- next_char(this);
- this->source_position.linenr++;
+ next_char();
+ if(c == '\n') {
+ next_char();
+ lexer_token.source_position.linenr++;
}
break;
case EOF:
+ case '\r':
case '\n':
return;
+
default:
- next_char(this);
+ next_char();
break;
}
}
}
-#if 0
+static token_t pp_token;
+
+static inline
+void next_pp_token(void)
+{
+ lexer_next_preprocessing_token();
+ pp_token = lexer_token;
+}
+
static
-void eat_until_newline(lexer_t *this)
+void eat_until_newline(void)
{
- while(this->c != '\n') {
- next_char(this);
+ while(pp_token.type != '\n' && pp_token.type != T_EOF) {
+ next_pp_token();
}
}
-#endif
-#if 0
static
-void parse_preprocessor_directive(lexer_t *this, token_t *result_token)
+void error_directive(void)
{
- (void) result_token;
- /* skip whitespaces */
- while(this->c == ' ' || this->c == '\t' || this->c == '\r') {
- next_char(this);
- }
+ error_prefix();
+ fprintf(stderr, "#error directive: \n");
+
+ /* parse pp-tokens until new-line */
}
-#endif
-void preprocessor_next_token(lexer_t *this, token_t *token)
+static
+void define_directive(void)
{
- /* skip whitespaces */
- while(this->c == ' ' || this->c == '\t' || this->c == '\r') {
- next_char(this);
+ lexer_next_preprocessing_token();
+ if(lexer_token.type != T_IDENTIFIER) {
+ parse_error("expected identifier after #define\n");
+ eat_until_newline();
}
+}
- switch(this->c) {
- case 'A' ... 'Z':
- case 'a' ... 'z':
- case '_':
- parse_symbol(this, token);
- }
+static
+void ifdef_directive(int is_ifndef)
+{
+ (void) is_ifndef;
+ lexer_next_preprocessing_token();
+ //expect_identifier();
+ //extect_newline();
}
-void lexer_next_token(lexer_t *this, token_t *token)
+static
+void endif_directive(void)
{
- int line_begin = 0;
+ //expect_newline();
+}
- /* skip whitespaces */
- while(this->c == ' ' || this->c == '\t' || this->c == '\n'
- || this->c == '\r') {
- if(this->c == '\n') {
- line_begin = 1;
- this->source_position.linenr++;
- }
- next_char(this);
+static
+void parse_line_directive(void)
+{
+ if(pp_token.type != T_INTEGER) {
+ parse_error("expected integer");
+ } else {
+ lexer_token.source_position.linenr = pp_token.v.intvalue - 1;
+ next_pp_token();
+ }
+ if(pp_token.type == T_STRING_LITERAL) {
+ lexer_token.source_position.input_name = pp_token.v.string;
+ next_pp_token();
}
- switch(this->c) {
- case 'A' ... 'Z':
- case 'a' ... 'z':
- case '_':
- parse_symbol(this, token);
- break;
+ eat_until_newline();
+}
- case '0' ... '9':
- parse_number(this, token);
- break;
+static
+void parse_preprocessor_identifier(void)
+{
+ assert(pp_token.type == T_IDENTIFIER);
+ symbol_t *symbol = pp_token.v.symbol;
- case '"':
- parse_string_literal(this, token);
+ switch(symbol->pp_ID) {
+ case TP_include:
+ printf("include - enable header name parsing!\n");
+ break;
+ case TP_define:
+ define_directive();
+ break;
+ case TP_ifdef:
+ ifdef_directive(0);
+ break;
+ case TP_ifndef:
+ ifdef_directive(1);
+ break;
+ case TP_endif:
+ endif_directive();
break;
+ case TP_line:
+ next_pp_token();
+ parse_line_directive();
+ break;
+ case TP_if:
+ case TP_else:
+ case TP_elif:
+ case TP_undef:
+ case TP_error:
+ error_directive();
+ break;
+ case TP_pragma:
+ break;
+ }
+}
- case '\'':
- next_char(this);
- if(this->c == '\\') {
- next_char(this);
- token->type = T_INTEGER;
- token->v.intvalue = parse_escape_sequence(this);
- } else {
- if(this->c == '\n') {
- parse_error(this, "newline while parsing character constant");
- this->source_position.linenr++;
- }
- token->type = T_INTEGER;
- token->v.intvalue = this->c;
- next_char(this);
- }
- if(this->c != '\'') {
- parse_error(this, "multibyte character constant");
- token->type = T_ERROR;
- } else {
- next_char(this);
- }
+static
+void parse_preprocessor_directive()
+{
+ next_pp_token();
+
+ switch(pp_token.type) {
+ case T_IDENTIFIER:
+ parse_preprocessor_identifier();
break;
+ case T_INTEGER:
+ parse_line_directive();
+ break;
+ default:
+ parse_error("invalid preprocessor directive");
+ eat_until_newline();
+ break;
+ }
+}
+
+#define MAYBE_PROLOG \
+ next_char(); \
+ while(1) { \
+ switch(c) {
- case '\\':
- next_char(this);
- if(this->c == '\n') {
- next_char(this);
- this->source_position.linenr++;
- lexer_next_token(this, token);
+#define MAYBE(ch, set_type) \
+ case ch: \
+ next_char(); \
+ lexer_token.type = set_type; \
+ return;
+
+#define ELSE_CODE(code) \
+ SKIP_TRIGRAPHS(, \
+ code; \
+ ) \
+ \
+ case '\\': \
+ next_char(); \
+ EAT_NEWLINE(break;) \
+ /* fallthrough */ \
+ default: \
+ code; \
+ } \
+ } /* end of while(1) */ \
+ break;
+
+#define ELSE(set_type) \
+ ELSE_CODE( \
+ lexer_token.type = set_type; \
+ return; \
+ )
+
+void lexer_next_preprocessing_token(void)
+{
+ while(1) {
+ switch(c) {
+ case ' ':
+ case '\t':
+ next_char();
+ break;
+
+ MATCH_NEWLINE(
+ lexer_token.type = '\n';
return;
- } else {
- parse_error(this, "unexpected '\\' found");
- token->type = T_ERROR;
- }
- break;
+ )
-#define MAYBE1(ch, set_type) \
- next_char(this); \
- while(1) { \
- switch(this->c) { \
- case ch: \
- next_char(this); \
- token->type = set_type; \
- return; \
-
-#define MAYBE(ch, set_type) \
- case ch: \
- next_char(this); \
- token->type = set_type; \
- return;
+ SYMBOL_CHARS
+ parse_symbol();
+ return;
-#define ELSE(set_type) \
- case '?': \
- next_char(this); \
- if(this->c != '?') { \
- put_back(this, this->c); \
- this->c = '?'; \
- token->type = set_type; \
- return; \
- } \
- next_char(this); \
- if(replace_trigraph(this)) \
- break; \
- put_back(this, '?'); \
- put_back(this, this->c); \
- this->c = '?'; \
- token->type = set_type; \
- return; \
- \
- case '\\': \
- next_char(this); \
- if(this->c == '\n') { \
- next_char(this); \
- this->source_position.linenr++; \
- break; \
- } \
- /* fallthrough */ \
- default: \
- token->type = set_type; \
- return; \
- } \
- } /* end of while(1) */ \
- break;
+ DIGITS
+ parse_number();
+ return;
+
+ case '"':
+ parse_string_literal();
+ return;
- case '.':
- next_char(this);
- if(this->c == '.') {
- next_char(this);
- if(this->c == '.') {
- next_char(this);
- token->type = T_DOTDOTDOT;
+ case '\'':
+ parse_character_constant();
+ return;
+
+ case '\\':
+ next_char();
+ if(c == '\n') {
+ next_char();
+ lexer_token.source_position.linenr++;
+ break;
} else {
- put_back(this, '.');
- token->type = '.';
+ parse_error("unexpected '\\' found");
+ lexer_token.type = T_ERROR;
}
- } else {
- token->type = '.';
- }
- break;
- case '&':
- MAYBE1('&', T_ANDAND)
- MAYBE('=', T_ANDEQUAL)
- ELSE('&')
- case '*':
- MAYBE1('=', T_ASTERISKEQUAL)
- ELSE('*')
- case '+':
- MAYBE1('+', T_PLUSPLUS)
- MAYBE('=', T_PLUSEQUAL)
- ELSE('+')
- case '-':
- MAYBE1('-', T_MINUSMINUS)
- MAYBE('=', T_MINUSEQUAL)
- ELSE('-')
- case '!':
- MAYBE1('=', T_EXCLAMATIONMARKEQUAL)
- ELSE('!')
- case '/':
- MAYBE1('=', T_SLASHEQUAL)
- case '*':
- next_char(this);
- skip_multiline_comment(this);
- lexer_next_token(this, token);
- return;
- case '/':
- next_char(this);
- skip_line_comment(this);
- lexer_next_token(this, token);
- return;
- ELSE('/')
- case '%':
- MAYBE1('=', T_PERCENTEQUAL)
- case ':':
- /* TODO find trigraphs... */
- next_char(this);
- if(this->c == '%') {
- next_char(this);
- if(this->c == ':') {
- next_char(this);
- token->type = T_PERCENTCOLONPERCENTCOLON;
- } else {
- put_back(this, '%');
- token->type = T_PERCENTCOLON;
- }
+ return;
+
+ case '.':
+ MAYBE_PROLOG
+ case '.':
+ MAYBE_PROLOG
+ MAYBE('.', T_DOTDOTDOT)
+ ELSE_CODE(
+ put_back(c);
+ c = '.';
+ lexer_token.type = '.';
+ return;
+ )
+ ELSE('.')
+ case '&':
+ MAYBE_PROLOG
+ MAYBE('&', T_ANDAND)
+ MAYBE('=', T_ANDEQUAL)
+ ELSE('&')
+ case '*':
+ MAYBE_PROLOG
+ MAYBE('=', T_ASTERISKEQUAL)
+ ELSE('*')
+ case '+':
+ MAYBE_PROLOG
+ MAYBE('+', T_PLUSPLUS)
+ MAYBE('=', T_PLUSEQUAL)
+ ELSE('+')
+ case '-':
+ MAYBE_PROLOG
+ MAYBE('-', T_MINUSMINUS)
+ MAYBE('=', T_MINUSEQUAL)
+ ELSE('-')
+ case '!':
+ MAYBE_PROLOG
+ MAYBE('=', T_EXCLAMATIONMARKEQUAL)
+ ELSE('!')
+ case '/':
+ MAYBE_PROLOG
+ MAYBE('=', T_SLASHEQUAL)
+ case '*':
+ next_char();
+ skip_multiline_comment();
+ lexer_next_preprocessing_token();
return;
- }
- token->type = T_PERCENTCOLON;
- return;
- MAYBE('>', T_PERCENTGREATER)
- ELSE('%')
- case '<':
- MAYBE1(':', T_LESSCOLON)
- MAYBE('%', T_LESSPERCENT)
- case '<':
- /* TODO trigraphs... */
- next_char(this);
- if(this->c == '<') {
- next_char(this);
- if(this->c == '=') {
- next_char(this);
- token->type = T_LESSLESSEQUAL;
- } else {
- token->type = T_LESSLESS;
- }
- } else {
- token->type = T_LESS;
- }
+ case '/':
+ next_char();
+ skip_line_comment();
+ lexer_next_preprocessing_token();
+ return;
+ ELSE('/')
+ case '%':
+ MAYBE_PROLOG
+ MAYBE('>', T_PERCENTGREATER)
+ MAYBE('=', T_PERCENTEQUAL)
+ case ':':
+ MAYBE_PROLOG
+ case '%':
+ MAYBE_PROLOG
+ MAYBE(':', T_PERCENTCOLONPERCENTCOLON)
+ ELSE_CODE(
+ put_back(c);
+ c = '%';
+ lexer_token.type = T_PERCENTCOLON;
+ return;
+ )
+ ELSE(T_PERCENTCOLON)
+ ELSE('%')
+ case '<':
+ MAYBE_PROLOG
+ MAYBE(':', T_LESSCOLON)
+ MAYBE('%', T_LESSPERCENT)
+ case '<':
+ MAYBE_PROLOG
+ MAYBE('=', T_LESSLESSEQUAL)
+ ELSE(T_LESSLESS)
+ ELSE('<')
+ case '>':
+ MAYBE_PROLOG
+ case '>':
+ MAYBE_PROLOG
+ MAYBE('=', T_GREATERGREATEREQUAL)
+ ELSE(T_GREATERGREATER)
+ ELSE('>')
+ case '^':
+ MAYBE_PROLOG
+ MAYBE('=', T_CARETEQUAL)
+ ELSE('^')
+ case '|':
+ MAYBE_PROLOG
+ MAYBE('=', T_PIPEEQUAL)
+ MAYBE('|', T_PIPEPIPE)
+ ELSE('|')
+ case ':':
+ MAYBE_PROLOG
+ MAYBE('>', T_COLONGREATER)
+ ELSE(':')
+ case '=':
+ MAYBE_PROLOG
+ MAYBE('=', T_EQUALEQUAL)
+ ELSE('=')
+ case '#':
+ MAYBE_PROLOG
+ MAYBE('#', T_HASHHASH)
+ ELSE('#')
+
+ case '?':
+ next_char();
+ /* just a simple ? */
+ if(c != '?') {
+ lexer_token.type = '?';
return;
- ELSE('<')
- case '>':
- next_char(this);
- while(1) {
- switch(this->c) {
- case '>':
- next_char(this);
- /* TODO trigraphs */
- if(this->c == '=') {
- next_char(this);
- token->type = T_GREATERGREATEREQUAL;
- } else {
- token->type = T_GREATERGREATER;
- }
+ }
+ /* might be a trigraph */
+ next_char();
+ if(replace_trigraph()) {
break;
- ELSE('>')
- case '^':
- MAYBE1('=', T_CARETEQUAL)
- ELSE('^')
- case '|':
- MAYBE1('=', T_PIPEEQUAL)
- MAYBE('|', T_PIPEPIPE)
- ELSE('|')
- case ':':
- MAYBE1('>', T_COLONGREATER)
- ELSE(':')
- case '=':
- MAYBE1('=', T_EQUALEQUAL)
- ELSE('=')
- case '#':
- MAYBE1('#', T_HASHHASH)
-#if 0
- else {
- if(line_begin) {
- parse_preprocessor_directive(this, token);
- return;
- } else {
- token->type = '#';
}
-#else
- ELSE('#')
-#endif
+ put_back(c);
+ c = '?';
+ lexer_token.type = '?';
+ return;
- case '?':
- next_char(this);
- /* just a simple ? */
- if(this->c != '?') {
- token->type = '?';
- break;
- }
- /* might be a trigraph */
- next_char(this);
- if(replace_trigraph(this)) {
- lexer_next_token(this, token);
+ case '[':
+ case ']':
+ case '(':
+ case ')':
+ case '{':
+ case '}':
+ case '~':
+ case ';':
+ case ',':
+ lexer_token.type = c;
+ next_char();
+ return;
+
+ case EOF:
+ lexer_token.type = T_EOF;
+ return;
+
+ default:
+ next_char();
+ error_prefix();
+ fprintf(stderr, "unknown character '%c' found\n", c);
+ lexer_token.type = T_ERROR;
return;
}
- put_back(this, this->c);
- this->c = '?';
- token->type = '?';
- break;
+ }
+}
- case '[':
- case ']':
- case '(':
- case ')':
- case '{':
- case '}':
- case '~':
- case ';':
- case ',':
- token->type = this->c;
- next_char(this);
- break;
+void lexer_next_token(void)
+{
+ lexer_next_preprocessing_token();
+ if(lexer_token.type != '\n')
+ return;
- case EOF:
- token->type = T_EOF;
- break;
+newline_found:
+ do {
+ lexer_next_preprocessing_token();
+ } while(lexer_token.type == '\n');
- default:
- error_prefix(this);
- fprintf(stderr, "unknown character '%c' found\n", this->c);
- token->type = T_ERROR;
- next_char(this);
- break;
+ if(lexer_token.type == '#') {
+ parse_preprocessor_directive();
+ goto newline_found;
}
}
-void lexer_init(lexer_t *this, FILE *stream, const char *input_name)
+void init_lexer(void)
{
- memset(this, 0, sizeof(this[0]));
-
- this->input = stream;
+ strset_init(&stringset);
+}
- this->source_position.linenr = 0;
- this->source_position.input_name = input_name;
- strset_init(&this->stringset);
+void lexer_open_stream(FILE *stream, const char *input_name)
+{
+ input = stream;
+ lexer_token.source_position.linenr = 0;
+ lexer_token.source_position.input_name = input_name;
/* we place a virtual '\n' at the beginning so the lexer knows we're at the
* beginning of a line */
- this->c = '\n';
+ c = '\n';
}
-void lexer_destroy(lexer_t *this)
+void exit_lexer(void)
{
- (void) this;
+ strset_destroy(&stringset);
}
static __attribute__((unused))
void dbg_pos(const source_position_t source_position)
{
- fprintf(stdout, "%s:%d\n", source_position.input_name, source_position.linenr);
+ fprintf(stdout, "%s:%d\n", source_position.input_name,
+ source_position.linenr);
fflush(stdout);
}