Merge branch 'cooking' of git://github.com/graphitemaster/gmqcc into cooking

[xonotic/gmqcc.git] / lexer.c

diff --git a/lexer.c b/lexer.c
index 64cdb3d3cff06b7212c7caf4cc44472ebe700d1e..8289057e1f0f20c720aa506241c1afe17818b3d5 100644 (file)
--- a/lexer.c
+++ b/lexer.c
@@ -20,14 +20,12 @@
  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
-#include <stdio.h>
-#include <stdlib.h>
+#include <ctype.h>
 #include <string.h>
-#include <stdarg.h>
+#include <stdlib.h>
 
 #include "gmqcc.h"
 #include "lexer.h"
-
 /*
  * List of Keywords
  */
@@ -58,21 +56,21 @@ static size_t num_keywords_fg = sizeof(keywords_fg) / sizeof(keywords_fg[0]);
  * Lexer code
  */
 
-char* *lex_filenames;
+static char* *lex_filenames;
 
-void lexerror(lex_file *lex, const char *fmt, ...)
+static void lexerror(lex_file *lex, const char *fmt, ...)
 {
     va_list ap;
 
     va_start(ap, fmt);
     if (lex)
-        con_vprintmsg(LVL_ERROR, lex->name, lex->sline, "parse error", fmt, ap);
+        con_vprintmsg(LVL_ERROR, lex->name, lex->sline, lex->column, "parse error", fmt, ap);
     else
-        con_vprintmsg(LVL_ERROR, "", 0, "parse error", fmt, ap);
+        con_vprintmsg(LVL_ERROR, "", 0, 0, "parse error", fmt, ap);
     va_end(ap);
 }
 
-bool lexwarn(lex_file *lex, int warntype, const char *fmt, ...)
+static bool lexwarn(lex_file *lex, int warntype, const char *fmt, ...)
 {
     bool    r;
     lex_ctx ctx;
@@ -174,9 +172,11 @@ static void lex_token_new(lex_file *lex)
 #else
     if (lex->tok.value)
         vec_shrinkto(lex->tok.value, 0);
+        
     lex->tok.constval.t  = 0;
-    lex->tok.ctx.line = lex->sline;
-    lex->tok.ctx.file = lex->name;
+    lex->tok.ctx.line    = lex->sline;
+    lex->tok.ctx.file    = lex->name;
+    lex->tok.ctx.column  = lex->column;
 #endif
 }
 #endif
@@ -184,7 +184,7 @@ static void lex_token_new(lex_file *lex)
 lex_file* lex_open(const char *file)
 {
     lex_file *lex;
-    FILE *in = file_open(file, "rb");
+    FILE *in = fs_file_open(file, "rb");
 
     if (!in) {
         lexerror(NULL, "open failed: '%s'\n", file);
@@ -193,19 +193,19 @@ lex_file* lex_open(const char *file)
 
     lex = (lex_file*)mem_a(sizeof(*lex));
     if (!lex) {
-        file_close(in);
+        fs_file_close(in);
         lexerror(NULL, "out of memory\n");
         return NULL;
     }
 
     memset(lex, 0, sizeof(*lex));
 
-    lex->file = in;
-    lex->name = util_strdup(file);
-    lex->line = 1; /* we start counting at 1 */
-
+    lex->file    = in;
+    lex->name    = util_strdup(file);
+    lex->line    = 1; /* we start counting at 1 */
+    lex->column  = 0;
     lex->peekpos = 0;
-    lex->eof = false;
+    lex->eof     = false;
 
     vec_push(lex_filenames, lex->name);
     return lex;
@@ -228,11 +228,11 @@ lex_file* lex_open_string(const char *str, size_t len, const char *name)
     lex->open_string_length = len;
     lex->open_string_pos    = 0;
 
-    lex->name = util_strdup(name ? name : "<string-source>");
-    lex->line = 1; /* we start counting at 1 */
-
+    lex->name    = util_strdup(name ? name : "<string-source>");
+    lex->line    = 1; /* we start counting at 1 */
     lex->peekpos = 0;
-    lex->eof = false;
+    lex->eof     = false;
+    lex->column  = 0;
 
     vec_push(lex_filenames, lex->name);
 
@@ -258,7 +258,7 @@ void lex_close(lex_file *lex)
         vec_free(lex->modelname);
 
     if (lex->file)
-        file_close(lex->file);
+        fs_file_close(lex->file);
 #if 0
     if (lex->tok)
         token_delete(lex->tok);
@@ -271,11 +271,14 @@ void lex_close(lex_file *lex)
 
 static int lex_fgetc(lex_file *lex)
 {
-    if (lex->file)
-        return fgetc(lex->file);
+    if (lex->file) {
+        lex->column++;
+        return fs_file_getc(lex->file);
+    }
     if (lex->open_string) {
         if (lex->open_string_pos >= lex->open_string_length)
             return EOF;
+        lex->column++;
         return lex->open_string[lex->open_string_pos++];
     }
     return EOF;
@@ -291,12 +294,22 @@ static int lex_try_trigraph(lex_file *lex, int old)
 {
     int c2, c3;
     c2 = lex_fgetc(lex);
+    if (!lex->push_line && c2 == '\n') {
+        lex->line++;
+        lex->column = 0;
+    }
+    
     if (c2 != '?') {
         lex_ungetch(lex, c2);
         return old;
     }
 
     c3 = lex_fgetc(lex);
+    if (!lex->push_line && c3 == '\n') {
+        lex->line++;
+        lex->column = 0;
+    }
+    
     switch (c3) {
         case '=': return '#';
         case '/': return '\\';
@@ -361,8 +374,11 @@ static int lex_getch(lex_file *lex)
 static void lex_ungetch(lex_file *lex, int ch)
 {
     lex->peek[lex->peekpos++] = ch;
-    if (!lex->push_line && ch == '\n')
+    lex->column--;
+    if (!lex->push_line && ch == '\n') {
         lex->line--;
+        lex->column = 0;
+    }
 }
 
 /* classify characters
@@ -479,6 +495,9 @@ static bool lex_try_pragma(lex_file *lex)
     lex->line = line;
     while (ch != '\n' && ch != EOF)
         ch = lex_getch(lex);
+    vec_free(command);
+    vec_free(param);
+    vec_free(pragma);
     return true;
 
 unroll:
@@ -491,13 +510,13 @@ unroll:
         vec_free(command);
         lex_ungetch(lex, ' ');
     }
-    if (command) {
-        vec_pop(command);
-        while (vec_size(command)) {
-            lex_ungetch(lex, (unsigned char)vec_last(command));
-            vec_pop(command);
+    if (param) {
+        vec_pop(param);
+        while (vec_size(param)) {
+            lex_ungetch(lex, (unsigned char)vec_last(param));
+            vec_pop(param);
         }
-        vec_free(command);
+        vec_free(param);
         lex_ungetch(lex, ' ');
     }
     if (pragma) {
@@ -865,6 +884,7 @@ static int GMQCC_WARN lex_finish_string(lex_file *lex, int quote)
                         ch = 0;
                     else {
                         --u8len;
+                        lex->column += u8len;
                         for (uc = 0; uc < u8len; ++uc)
                             lex_tokench(lex, u8buf[uc]);
                         /* the last character will be inserted with the tokench() call
@@ -988,6 +1008,8 @@ int lex_do(lex_file *lex)
         if (!lex->flags.mergelines || ch != '\\')
             break;
         ch = lex_getch(lex);
+        if (ch == '\r')
+            ch = lex_getch(lex);
         if (ch != '\n') {
             lex_ungetch(lex, ch);
             ch = '\\';
@@ -1244,31 +1266,37 @@ int lex_do(lex_file *lex)
             default:
                 break;
         }
+    }
 
-        if (ch == '.')
-        {
-            lex_tokench(lex, ch);
-            /* peak ahead once */
-            nextch = lex_getch(lex);
-            if (nextch != '.') {
-                lex_ungetch(lex, nextch);
-                lex_endtoken(lex);
-                return (lex->tok.ttype = ch);
-            }
-            /* peak ahead again */
-            nextch = lex_getch(lex);
-            if (nextch != '.') {
-                lex_ungetch(lex, nextch);
-                lex_ungetch(lex, '.');
-                lex_endtoken(lex);
+    if (ch == '.')
+    {
+        lex_tokench(lex, ch);
+        /* peak ahead once */
+        nextch = lex_getch(lex);
+        if (nextch != '.') {
+            lex_ungetch(lex, nextch);
+            lex_endtoken(lex);
+            if (lex->flags.noops)
                 return (lex->tok.ttype = ch);
-            }
-            /* fill the token to be "..." */
-            lex_tokench(lex, ch);
-            lex_tokench(lex, ch);
+            else
+                return (lex->tok.ttype = TOKEN_OPERATOR);
+        }
+        /* peak ahead again */
+        nextch = lex_getch(lex);
+        if (nextch != '.') {
+            lex_ungetch(lex, nextch);
+            lex_ungetch(lex, '.');
             lex_endtoken(lex);
-            return (lex->tok.ttype = TOKEN_DOTS);
+            if (lex->flags.noops)
+                return (lex->tok.ttype = ch);
+            else
+                return (lex->tok.ttype = TOKEN_OPERATOR);
         }
+        /* fill the token to be "..." */
+        lex_tokench(lex, ch);
+        lex_tokench(lex, ch);
+        lex_endtoken(lex);
+        return (lex->tok.ttype = TOKEN_DOTS);
     }
 
     if (ch == ',' || ch == '.') {
@@ -1278,15 +1306,23 @@ int lex_do(lex_file *lex)
     }
 
     if (ch == '+' || ch == '-' || /* ++, --, +=, -=  and -> as well! */
-        ch == '>' || ch == '<' || /* <<, >>, <=, >= */
-        ch == '=' || ch == '!' || /* ==, != */
-        ch == '&' || ch == '|')   /* &&, ||, &=, |= */
-    {
+        ch == '>' || ch == '<' || /* <<, >>, <=, >=                  */
+        ch == '=' || ch == '!' || /* <=>, ==, !=                     */
+        ch == '&' || ch == '|' || /* &&, ||, &=, |=                  */
+        ch == '~'                 /* ~=, ~                           */
+    )  {
         lex_tokench(lex, ch);
 
         nextch = lex_getch(lex);
-        if (nextch == '=' || (nextch == ch && ch != '!')) {
+        if ((nextch == '=' && ch != '<') || (nextch == ch && ch != '!')) {
+            lex_tokench(lex, nextch);
+        } else if (ch == '<' && nextch == '=') {
             lex_tokench(lex, nextch);
+            if ((thirdch = lex_getch(lex)) == '>')
+                lex_tokench(lex, thirdch);
+            else
+                lex_ungetch(lex, thirdch);
+
         } else if (ch == '-' && nextch == '>') {
             lex_tokench(lex, nextch);
         } else if (ch == '&' && nextch == '~') {
@@ -1310,8 +1346,9 @@ int lex_do(lex_file *lex)
                 lex->tok.constval.f = -lex->tok.constval.f;
             lex_endtoken(lex);
             return lex->tok.ttype;
-        } else
+        } else {
             lex_ungetch(lex, nextch);
+        }
 
         lex_endtoken(lex);
         return (lex->tok.ttype = TOKEN_OPERATOR);
@@ -1331,7 +1368,7 @@ int lex_do(lex_file *lex)
         lex_tokench(lex, ch);
 
         nextch = lex_getch(lex);
-        if (nextch == '=') {
+        if (nextch == '=' || nextch == '*') {
             lex_tokench(lex, nextch);
         } else
             lex_ungetch(lex, nextch);
@@ -1340,6 +1377,12 @@ int lex_do(lex_file *lex)
         return (lex->tok.ttype = TOKEN_OPERATOR);
     }
 
+    if (ch == '%') {
+        lex_tokench(lex, ch);
+        lex_endtoken(lex);
+        return (lex->tok.ttype = TOKEN_OPERATOR);
+    }
+
     if (isident_start(ch))
     {
         const char *v;
@@ -1377,7 +1420,7 @@ int lex_do(lex_file *lex)
                 if (!strcmp(v, keywords_qc[kw]))
                     return (lex->tok.ttype = TOKEN_KEYWORD);
             }
-            if (opts.standard != COMPILER_QCC) {
+            if (OPTS_OPTION_U32(OPTION_STANDARD) != COMPILER_QCC) {
                 for (kw = 0; kw < num_keywords_fg; ++kw) {
                     if (!strcmp(v, keywords_fg[kw]))
                         return (lex->tok.ttype = TOKEN_KEYWORD);
@@ -1474,6 +1517,6 @@ int lex_do(lex_file *lex)
         return (lex->tok.ttype = ch);
     }
 
-    lexerror(lex, "unknown token");
+    lexerror(lex, "unknown token: `%s`", lex->tok.value);
     return (lex->tok.ttype = TOKEN_ERROR);
 }