+/*
+ * Copyright (C) 2012
+ * Wolfgang Bumiller
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy of
+ * this software and associated documentation files (the "Software"), to deal in
+ * the Software without restriction, including without limitation the rights to
+ * use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
+ * of the Software, and to permit persons to whom the Software is furnished to do
+ * so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "gmqcc.h"
#include "lexer.h"
+/*
+ * List of Keywords
+ */
+
+/* original */
+static const char *keywords_qc[] = {
+ "for", "do", "while",
+ "if", "else",
+ "local",
+ "return",
+ "const"
+};
+static size_t num_keywords_qc = sizeof(keywords_qc) / sizeof(keywords_qc[0]);
+
+/* For fte/gmgqcc */
+static const char *keywords_fg[] = {
+ "switch", "case", "default",
+ "struct", "union",
+ "break", "continue",
+ "typedef",
+ "goto",
+
+ "noreturn",
+
+ "__builtin_debug_printtype"
+};
+static size_t num_keywords_fg = sizeof(keywords_fg) / sizeof(keywords_fg[0]);
+
+/*
+ * Lexer code
+ */
+
char* *lex_filenames;
void lexerror(lex_file *lex, const char *fmt, ...)
{
- va_list ap;
+ va_list ap;
- va_start(ap, fmt);
- if (lex)
+ va_start(ap, fmt);
+ if (lex)
con_vprintmsg(LVL_ERROR, lex->name, lex->sline, "parse error", fmt, ap);
else
con_vprintmsg(LVL_ERROR, "", 0, "parse error", fmt, ap);
- va_end(ap);
+ va_end(ap);
}
bool lexwarn(lex_file *lex, int warntype, const char *fmt, ...)
{
- va_list ap;
- int lvl = LVL_WARNING;
-
- if (!OPTS_WARN(warntype))
- return false;
+ bool r;
+ lex_ctx ctx;
+ va_list ap;
- if (opts_werror)
- lvl = LVL_ERROR;
+ ctx.file = lex->name;
+ ctx.line = lex->sline;
- va_start(ap, fmt);
- con_vprintmsg(lvl, lex->name, lex->sline, "warning", fmt, ap);
- va_end(ap);
-
- return opts_werror;
+ va_start(ap, fmt);
+ r = vcompile_warning(ctx, warntype, fmt, ap);
+ va_end(ap);
+ return r;
}
{
int c2;
c2 = lex_fgetc(lex);
+ /* we just used fgetc() so count lines
+ * need to offset a \n the ungetch would recognize
+ */
+ if (!lex->push_line && c2 == '\n')
+ lex->line++;
if (ch == '<' && c2 == ':')
return '[';
else if (ch == ':' && c2 == '>')
goto unroll;
}
- for (ch = lex_getch(lex); vec_size(param) < 32 && ch != ')' && ch != '\n'; ch = lex_getch(lex))
+ for (ch = lex_getch(lex); vec_size(param) < 1024 && ch != ')' && ch != '\n'; ch = lex_getch(lex))
vec_push(param, ch);
vec_push(param, 0);
if (!strcmp(command, "push")) {
if (!strcmp(param, "line")) {
lex->push_line++;
- --line;
+ if (lex->push_line == 1)
+ --line;
}
else
goto unroll;
if (!strcmp(param, "line")) {
if (lex->push_line)
lex->push_line--;
- --line;
+ if (lex->push_line == 0)
+ --line;
}
else
goto unroll;
goto unroll;
lex->line = line;
- while (ch != '\n')
+ while (ch != '\n' && ch != EOF)
ch = lex_getch(lex);
return true;
vec_pop(command);
}
vec_free(command);
+ lex_ungetch(lex, ' ');
}
if (command) {
vec_pop(command);
vec_pop(command);
}
vec_free(command);
+ lex_ungetch(lex, ' ');
}
if (pragma) {
vec_pop(pragma);
* here is to store the line of the first character after skipping
* the initial whitespace in lex->sline, this happens in lex_do.
*/
-static int lex_skipwhite(lex_file *lex)
+static int lex_skipwhite(lex_file *lex, bool hadwhite)
{
int ch = 0;
- bool haswhite = false;
+ bool haswhite = hadwhite;
do
{
}
break;
}
+ lex_ungetch(lex, ch);
}
if (lex->flags.preprocessing) {
- lex_tokench(lex, ' '); /* ch); */
+ if (ch == '\n')
+ lex_tokench(lex, '\n');
+ else
+ lex_tokench(lex, ' '); /* ch); */
}
}
ch = ' '; /* cause TRUE in the isspace check */
static int GMQCC_WARN lex_finish_string(lex_file *lex, int quote)
{
int ch = 0;
+ int nextch;
while (ch != EOF)
{
case 't': ch = '\t'; break;
case 'f': ch = '\f'; break;
case 'v': ch = '\v'; break;
+ case 'x':
+ case 'X':
+ /* same procedure as in fteqcc */
+ ch = 0;
+ nextch = lex_getch(lex);
+ if (nextch >= '0' && nextch <= '9')
+ ch += nextch - '0';
+ else if (nextch >= 'a' && nextch <= 'f')
+ ch += nextch - 'a' + 10;
+ else if (nextch >= 'A' && nextch <= 'F')
+ ch += nextch - 'A' + 10;
+ else {
+ lexerror(lex, "bad character code");
+ lex_ungetch(lex, nextch);
+ return (lex->tok.ttype = TOKEN_ERROR);
+ }
+
+ ch *= 0x10;
+ nextch = lex_getch(lex);
+ if (nextch >= '0' && nextch <= '9')
+ ch += nextch - '0';
+ else if (nextch >= 'a' && nextch <= 'f')
+ ch += nextch - 'a' + 10;
+ else if (nextch >= 'A' && nextch <= 'F')
+ ch += nextch - 'A' + 10;
+ else {
+ lexerror(lex, "bad character code");
+ lex_ungetch(lex, nextch);
+ return (lex->tok.ttype = TOKEN_ERROR);
+ }
+ break;
+
+ /* fteqcc support */
+ case '0': case '1': case '2': case '3':
+ case '4': case '5': case '6': case '7':
+ case '8': case '9':
+ ch = 18 + ch - '0';
+ break;
+ case '<': ch = 29; break;
+ case '-': ch = 30; break;
+ case '>': ch = 31; break;
+ case '[': ch = 16; break;
+ case ']': ch = 17; break;
+ case '{':
+ ch = 0;
+ for (nextch = lex_getch(lex); nextch != '}'; nextch = lex_getch(lex)) {
+ ch = ch * 10 + nextch - '0';
+ if (nextch < '0' || nextch > '9' || ch > 255) {
+ lexerror(lex, "bad character code");
+ return (lex->tok.ttype = TOKEN_ERROR);
+ }
+ }
+ break;
+ case '\n': ch = '\n'; break;
+
default:
lexwarn(lex, WARN_UNKNOWN_CONTROL_SEQUENCE, "unrecognized control sequence: \\%c", ch);
/* so we just add the character plus backslash no matter what it actually is */
int ch = lastch;
/* parse a number... */
- lex->tok.ttype = TOKEN_INTCONST;
+ if (ch == '.')
+ lex->tok.ttype = TOKEN_FLOATCONST;
+ else
+ lex->tok.ttype = TOKEN_INTCONST;
lex_tokench(lex, ch);
}
}
/* NOT else, '.' can come from above as well */
- if (ch == '.' && !ishex)
+ if (lex->tok.ttype != TOKEN_FLOATCONST && ch == '.' && !ishex)
{
/* Allow floating comma in non-hex mode */
lex->tok.ttype = TOKEN_FLOATCONST;
int lex_do(lex_file *lex)
{
- int ch, nextch;
+ int ch, nextch, thirdch;
+ bool hadwhite = false;
lex_token_new(lex);
#if 0
#endif
while (true) {
- ch = lex_skipwhite(lex);
+ ch = lex_skipwhite(lex, hadwhite);
+ hadwhite = true;
if (!lex->flags.mergelines || ch != '\\')
break;
ch = lex_getch(lex);
continue;
}
- lex->sline = lex->line;
- lex->tok.ctx.line = lex->sline;
- lex->tok.ctx.file = lex->name;
-
if (lex->flags.preprocessing && (ch == TOKEN_WHITE || ch == TOKEN_EOL || ch == TOKEN_FATAL)) {
return (lex->tok.ttype = ch);
}
+ lex->sline = lex->line;
+ lex->tok.ctx.line = lex->sline;
+ lex->tok.ctx.file = lex->name;
+
if (lex->eof)
return (lex->tok.ttype = TOKEN_FATAL);
}
/* modelgen / spiritgen commands */
- if (ch == '$') {
+ if (ch == '$' && !lex->flags.preprocessing) {
const char *v;
size_t frame;
if (rc < 0)
return (lex->tok.ttype = TOKEN_FATAL);
- v = lex->tok.value;
if (lex->modelname) {
frame_macro m;
m.value = lex->framevalue;
if (!strcmp(v, "flush"))
{
- size_t frame;
- for (frame = 0; frame < vec_size(lex->frames); ++frame)
- mem_d(lex->frames[frame].name);
+ size_t fi;
+ for (fi = 0; fi < vec_size(lex->frames); ++fi)
+ mem_d(lex->frames[fi].name);
vec_free(lex->frames);
/* skip line (fteqcc does it too) */
ch = lex_getch(lex);
switch (ch)
{
case '[':
+ nextch = lex_getch(lex);
+ if (nextch == '[') {
+ lex_tokench(lex, nextch);
+ lex_endtoken(lex);
+ return (lex->tok.ttype = TOKEN_ATTRIBUTE_OPEN);
+ }
+ lex_ungetch(lex, nextch);
+ /* FALL THROUGH */
case '(':
+ case ':':
+ case '?':
lex_tokench(lex, ch);
lex_endtoken(lex);
if (lex->flags.noops)
return (lex->tok.ttype = ch);
else
return (lex->tok.ttype = TOKEN_OPERATOR);
+
+ case ']':
+ nextch = lex_getch(lex);
+ if (nextch == ']') {
+ lex_tokench(lex, nextch);
+ lex_endtoken(lex);
+ return (lex->tok.ttype = TOKEN_ATTRIBUTE_CLOSE);
+ }
+ lex_ungetch(lex, nextch);
+ /* FALL THROUGH */
case ')':
case ';':
- case ':':
case '{':
case '}':
- case ']':
case '#':
lex_tokench(lex, ch);
break;
}
+ if (ch == '.') {
+ nextch = lex_getch(lex);
+ /* digits starting with a dot */
+ if (isdigit(nextch)) {
+ lex_ungetch(lex, nextch);
+ lex->tok.ttype = lex_finish_digit(lex, ch);
+ lex_endtoken(lex);
+ return lex->tok.ttype;
+ }
+ lex_ungetch(lex, nextch);
+ }
+
if (lex->flags.noops)
{
/* Detect characters early which are normally
*/
switch (ch)
{
+ /*
case '+':
case '-':
+ */
case '*':
case '/':
case '<':
nextch = lex_getch(lex);
if (nextch != '.') {
lex_ungetch(lex, nextch);
- lex_ungetch(lex, nextch);
+ lex_ungetch(lex, '.');
lex_endtoken(lex);
return (lex->tok.ttype = ch);
}
lex_tokench(lex, ch);
nextch = lex_getch(lex);
- if (nextch == ch || nextch == '=') {
+ if (nextch == '=' || (nextch == ch && ch != '!')) {
lex_tokench(lex, nextch);
} else if (ch == '-' && nextch == '>') {
lex_tokench(lex, nextch);
+ } else if (ch == '&' && nextch == '~') {
+ thirdch = lex_getch(lex);
+ if (thirdch != '=') {
+ lex_ungetch(lex, thirdch);
+ lex_ungetch(lex, nextch);
+ }
+ else {
+ lex_tokench(lex, nextch);
+ lex_tokench(lex, thirdch);
+ }
} else
lex_ungetch(lex, nextch);
} else if (!strcmp(v, "vector")) {
lex->tok.ttype = TOKEN_TYPENAME;
lex->tok.constval.t = TYPE_VECTOR;
- } else if (!strcmp(v, "for") ||
- !strcmp(v, "while") ||
- !strcmp(v, "do") ||
- !strcmp(v, "if") ||
- !strcmp(v, "else") ||
- !strcmp(v, "local") ||
- !strcmp(v, "return") ||
- !strcmp(v, "not") ||
- !strcmp(v, "const"))
- {
- lex->tok.ttype = TOKEN_KEYWORD;
- }
- else if (opts_standard != COMPILER_QCC)
- {
- /* other standards reserve these keywords */
- if (!strcmp(v, "switch") ||
- !strcmp(v, "struct") ||
- !strcmp(v, "union") ||
- !strcmp(v, "break") ||
- !strcmp(v, "continue") ||
- !strcmp(v, "var"))
- {
- lex->tok.ttype = TOKEN_KEYWORD;
+ } else {
+ size_t kw;
+ for (kw = 0; kw < num_keywords_qc; ++kw) {
+ if (!strcmp(v, keywords_qc[kw]))
+ return (lex->tok.ttype = TOKEN_KEYWORD);
+ }
+ if (opts.standard != COMPILER_QCC) {
+ for (kw = 0; kw < num_keywords_fg; ++kw) {
+ if (!strcmp(v, keywords_fg[kw]))
+ return (lex->tok.ttype = TOKEN_KEYWORD);
+ }
}
}
while (!lex->flags.preprocessing && lex->tok.ttype == TOKEN_STRINGCONST)
{
/* Allow c style "string" "continuation" */
- ch = lex_skipwhite(lex);
+ ch = lex_skipwhite(lex, false);
if (ch != '"') {
lex_ungetch(lex, ch);
break;
lex_tokench(lex, ch);
lex_endtoken(lex);
+ lex->tok.ttype = TOKEN_CHARCONST;
/* It's a vector if we can successfully scan 3 floats */
-#ifdef WIN32
+#ifdef _MSC_VER
if (sscanf_s(lex->tok.value, " %f %f %f ",
&lex->tok.constval.v.x, &lex->tok.constval.v.y, &lex->tok.constval.v.z) == 3)
#else
{
lex->tok.ttype = TOKEN_VECTORCONST;
}
+ else
+ {
+ if (!lex->flags.preprocessing && strlen(lex->tok.value) > 1) {
+ if (lexwarn(lex, WARN_MULTIBYTE_CHARACTER, "multibyte character: `%s`", lex->tok.value))
+ return (lex->tok.ttype = TOKEN_ERROR);
+ }
+ lex->tok.constval.i = lex->tok.value[0];
+ }
return lex->tok.ttype;
}
return lex->tok.ttype;
}
+ if (lex->flags.preprocessing) {
+ lex_tokench(lex, ch);
+ lex_endtoken(lex);
+ return (lex->tok.ttype = ch);
+ }
+
lexerror(lex, "unknown token");
return (lex->tok.ttype = TOKEN_ERROR);
}