int
lexpeekc(LexStream *L)
{
int c = lexgetc(L);
lexungetc(L, c);
return c;
}
static void
testFloat(const char *str, float fval)
{
AST *ast;
LexStream L;
Token t;
int c;
uint32_t val;
union {
uint32_t i;
float f;
} v;
v.f = fval;
val = v.i;
printf("testing number[%s]...", str);
strToLex(&L, str, NULL);
t = getToken(&L, &ast);
EXPECTEQ(t, T_FLOATNUM);
c = lexgetc(&L);
EXPECTEQ(c, T_EOF);
assert(ast != NULL);
assert(ast->kind == AST_FLOAT);
EXPECTEQ(ast->d.ival, val);
printf("passed\n");
}
inclusion()
{ int i,c;
while (iswhite(c=lexgetc()));
if (c != '<' && c != '"') {
my_yyerror("< or \" expected after an include statement\n",
"cannot continue compilation"); }
if(c=='"'){ for(s=(char *)buffer; (c=lexgetc())!='"'; *s++ = c ){
if (c==EOF) { my_yyerror("reached EOF and have not found \" to close string of include\n","cannot continue compilation");
my_exit(1); }
};
*s = '\0';
while (iswhite(c=lexgetc()));
if (c != ';') { my_yyerror("; expected after string",
"cannot continue compilation");
my_exit(1); }
}else{
s = (char *)buffer;
t = incdir;
while(*s++ = *t++);
s--;
*s++ = '/';
while(iswhite(c=lexgetc()));
if (isalpha(c)) {
while(isalpha(c) ||isdigit(c) ||c=='.')
{ *s=c; c=lexgetc(); s++;
if (c==EOF) { my_yyerror("reached EOF and have not found >","cannot continue compilation");
my_exit(1); }
}
*s = '\0';
while (iswhite(c)) c=lexgetc();
if ( c != '>' ) { my_yyerror("> needed to close include statement","cannot continue compilation"); my_exit(1); }
while (iswhite(c=lexgetc()));
if (c != ';') { my_yyerror("; expected after include <filename>",
"cannot continue compilation");
my_exit(1); }
} else { my_yyerror("after < a file name is required",
"cannot continue compilation"); my_exit(1); }
}
if((lexin=fopen(buffer,"r"))==NULL){
my_yyerror( "cannot open include file"," ");
fprintf(stderr,"'%s` ",buffer);
my_exit(1);
}
in_index++;
in_files[in_index].in_fdes = lexin;
in_files[in_index].in_name = strsave(buffer);
in_files[in_index].in_line = 1;
i = 0;
while (((int)c=getc(lexin))=='\n') { }
while ( c !='\n' && c != EOF){
linebuf[i]=c;
c = getc(lexin);
i++; }
linebuf[i] = c;
curr_index = 0;
curr_length = i;
}
/* parse a string */
static int
parseString(LexStream *L, AST **ast_ptr)
{
int c;
struct flexbuf fb;
AST *ast;
flexbuf_init(&fb, INCSTR);
c = lexgetc(L);
while (c != '"' && c > 0 && c < 256) {
flexbuf_addchar(&fb, c);
c = lexgetc(L);
}
flexbuf_addchar(&fb, '\0');
ast = NewAST(AST_STRING, NULL, NULL);
ast->d.string = flexbuf_get(&fb);
*ast_ptr = ast;
return T_STRING;
}
/*
* Read a number for the lexical analyzer.
* Input is the first character of the number.
* Return value is the lexical type.
*/
static int
lexnumber(wint_t c)
{
wchar_t *cp;
int dotfound = 0;
int efound = 0;
INT number;
cp = linebuf;
do {
if (iswdigit(c))
;
else if (c == '.') {
if (dotfound++)
break;
} else if (c == 'e' || c == 'E') {
if ((c = lexgetc()) != '-' && c != '+') {
lexungetc(c);
c = 'e';
} else
*cp++ = 'e';
if (efound++)
break;
} else
break;
*cp++ = c;
} while ((c = lexgetc()) != WEOF);
*cp = '\0';
if (dotfound && cp == linebuf+1)
return (DOT);
lexungetc(c);
errno = 0;
if (!dotfound && !efound &&
((number = wcstol(linebuf, (wchar_t **)0, 10)), errno != ERANGE))
yylval.node = intnode(number);
else
yylval.node = realnode((REAL)wcstod(linebuf, (wchar_t **)0));
return (CONSTANT);
}
yylex()
{
int k;
STRING strsave();
STRING getstring(),getword();
float getnum();
while(iswhite(c=lexgetc()));
if ( (isalpha(c)||c=='@') && c!=EOF ) { s = getword(c);
if ((k=keyfind(s))!=NKEYWORDS && cconst)
{ yylval.strg = keywords[k].keyname;
cconst=false;
return(keywords[k].keyret); }
yylval.strg = s;
return(WORD); }
if ( isdigit(c)||c=='~' ){ yylval.numb=(float)getnum(c);
return(NUMB);}
if ( c=='`') { c = lexgetc();
yylval.strg=getstring(c);
return(STRING_QUOTED); }
if ( c == '[' ) cconst=true;
return(c);
}
static void
testNumber(const char *str, uint32_t val)
{
AST *ast;
LexStream L;
Token t;
int c;
printf("testing number[%s]...", str);
strToLex(&L, str, NULL);
t = getToken(&L, &ast);
EXPECTEQ(t, T_NUM);
c = lexgetc(&L);
EXPECTEQ(c, T_EOF);
assert(ast != NULL);
assert(ast->kind == AST_INTEGER);
EXPECTEQ(ast->d.ival, val);
printf("passed\n");
}
/*
* Read a string or regular expression, terminated by ``endc'',
* for lexical analyzer, processing escape sequences.
* Return string length.
*/
static size_t
lexescape(wint_t endc, int regx, int cmd_line_operand)
{
static char nlre[256];
static char nlstr[256];
static char eofre[256];
static char eofstr[256];
int first_time = 1;
wint_t c;
wchar_t *cp;
int n, max;
if (first_time == 1) {
(void) strcpy(nlre, gettext("Newline in regular expression\n"));
(void) strcpy(nlstr, gettext("Newline in string\n"));
(void) strcpy(eofre, gettext("EOF in regular expression\n"));
(void) strcpy(eofstr, gettext("EOF in string\n"));
first_time = 0;
}
cp = linebuf;
while ((c = lexgetc()) != endc) {
if (c == '\n')
awkerr(regx ? nlre : nlstr);
if (c == '\\') {
switch (c = lexgetc(), c) {
case '\\':
if (regx)
*cp++ = '\\';
break;
case '/':
c = '/';
break;
case 'n':
c = '\n';
break;
case 'b':
c = '\b';
break;
case 't':
c = '\t';
break;
case 'r':
c = '\r';
break;
case 'f':
c = '\f';
break;
case 'v':
c = '\v';
break;
case 'a':
c = (char)0x07;
break;
case 'x':
n = 0;
while (iswxdigit(c = lexgetc())) {
if (iswdigit(c))
c -= '0';
else if (iswupper(c))
c -= 'A'-10;
else
c -= 'a'-10;
n = (n<<4) + c;
}
lexungetc(c);
c = n;
break;
case '0':
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
#if 0
/*
* Posix.2 draft 10 disallows the use of back-referencing - it explicitly
* requires processing of the octal escapes both in strings and
* regular expressions. The following code is disabled instead of
* removed as back-referencing may be reintroduced in a future draft
* of the standard.
*/
/*
* For regular expressions, we disallow
* \ooo to mean octal character, in favour
* of back referencing.
*/
if (regx) {
*cp++ = '\\';
break;
}
#endif
max = 3;
n = 0;
do {
n = (n<<3) + c-'0';
if ((c = lexgetc()) > '7' || c < '0')
break;
} while (--max);
lexungetc(c);
/*
* an octal escape sequence must have at least
* 2 digits after the backslash, otherwise
* it gets passed straight thru for possible
* use in backreferencing.
*/
if (max == 3) {
*cp++ = '\\';
n += '0';
}
c = n;
break;
case '\n':
continue;
default:
if (c != endc || cmd_line_operand) {
*cp++ = '\\';
if (c == endc)
lexungetc(c);
}
}
}
if (c == WEOF)
awkerr(regx ? eofre : eofstr);
*cp++ = c;
}
*cp = '\0';
return (cp - linebuf);
}
/*
* Read an identifier.
* Input is first character of identifier.
* Return VAR.
*/
static int
lexid(wint_t c)
{
wchar_t *cp;
size_t i;
NODE *np;
cp = linebuf;
do {
*cp++ = c;
c = lexgetc();
} while (iswalpha(c) || iswdigit(c) || c == '_');
*cp = '\0';
lexungetc(c);
yylval.node = np = vlook(linebuf);
switch (np->n_type) {
case KEYWORD:
switch (np->n_keywtype) {
case PRINT:
case PRINTF:
++inprint;
/* FALLTHROUGH */
default:
return ((int)np->n_keywtype);
}
/* NOTREACHED */
case ARRAY:
case VAR:
/*
* If reading the argument list, create a dummy node
* for the duration of that function. These variables
* can be removed from the symbol table at function end
* but they must still exist because the execution tree
* knows about them.
*/
if (funparm) {
do_funparm:
np = emptynode(PARM, i = (cp-linebuf));
np->n_flags = FSTRING;
np->n_string = _null;
np->n_strlen = 0;
(void) memcpy(np->n_name, linebuf,
(i+1) * sizeof (wchar_t));
addsymtab(np);
yylval.node = np;
} else if (np == varNF || (np == varFS &&
(!doing_begin || begin_getline))) {
/*
* If the user program references NF or sets
* FS either outside of a begin block or
* in a begin block after a getline then the
* input line will be split immediately upon read
* rather than when a field is first referenced.
*/
needsplit = 1;
} else if (np == varENVIRON)
needenviron = 1;
/* FALLTHROUGH */
case PARM:
return (VAR);
case UFUNC:
/*
* It is ok to redefine functions as parameters
*/
if (funparm) goto do_funparm;
/* FALLTHROUGH */
case FUNC:
case GETLINE:
/*
* When a getline is encountered, clear the 'doing_begin' flag.
* This will force the 'needsplit' flag to be set, even inside
* a begin block, if FS is altered. (See VAR case above)
*/
if (doing_begin)
begin_getline = 1;
return (np->n_type);
}
/* NOTREACHED */
return (0);
}
/*
* The lexical analyzer.
*/
int
yylex()
{
wint_t c, c1;
int i;
static int savetoken = 0;
static int wasfield;
static int isfuncdef;
static int nbrace, nparen, nbracket;
static struct ctosymstruct {
wint_t c, sym;
} ctosym[] = {
{ '|', BAR }, { '^', CARAT },
{ '~', TILDE }, { '<', LANGLE },
{ '>', RANGLE }, { '+', PLUSC },
{ '-', HYPHEN }, { '*', STAR },
{ '/', SLASH }, { '%', PERCENT },
{ '!', EXCLAMATION }, { '$', DOLLAR },
{ '[', LSQUARE }, { ']', RSQUARE },
{ '(', LPAREN }, { ')', RPAREN },
{ ';', SEMI }, { '{', LBRACE },
{ '}', RBRACE }, { 0, 0 }
};
if (savetoken) {
c = savetoken;
savetoken = 0;
} else if (redelim != '\0') {
c = redelim;
redelim = 0;
catterm = 0;
savetoken = c;
c = lexlast = lexregexp(c);
goto out;
} else while ((c = lexgetc()) != WEOF) {
if (iswalpha(c) || c == '_') {
c = lexid(c);
} else if (iswdigit(c) || c == '.') {
c = lexnumber(c);
} else if (isWblank(c)) {
continue;
} else switch (c) {
#if DOS || OS2
case 032: /* ^Z */
continue;
#endif
case '"':
c = lexstring(c);
break;
case '#':
while ((c = lexgetc()) != '\n' && c != WEOF)
;
lexungetc(c);
continue;
case '+':
if ((c1 = lexgetc()) == '+')
c = INC;
else if (c1 == '=')
c = AADD;
else
lexungetc(c1);
break;
case '-':
if ((c1 = lexgetc()) == '-')
c = DEC;
else if (c1 == '=')
c = ASUB;
else
lexungetc(c1);
break;
case '*':
if ((c1 = lexgetc()) == '=')
c = AMUL;
else if (c1 == '*') {
if ((c1 = lexgetc()) == '=')
c = AEXP;
else {
c = EXP;
lexungetc(c1);
}
} else
lexungetc(c1);
break;
case '^':
if ((c1 = lexgetc()) == '=') {
c = AEXP;
} else {
c = EXP;
lexungetc(c1);
}
break;
case '/':
if ((c1 = lexgetc()) == '=' &&
lexlast != RE && lexlast != NRE &&
lexlast != ';' && lexlast != '\n' &&
lexlast != ',' && lexlast != '(')
c = ADIV;
else
lexungetc(c1);
break;
case '%':
if ((c1 = lexgetc()) == '=')
c = AREM;
else
lexungetc(c1);
break;
case '&':
if ((c1 = lexgetc()) == '&')
c = AND;
else
lexungetc(c1);
break;
case '|':
if ((c1 = lexgetc()) == '|')
c = OR;
else {
lexungetc(c1);
if (inprint)
c = PIPE;
}
break;
case '>':
if ((c1 = lexgetc()) == '=')
c = GE;
else if (c1 == '>')
c = APPEND;
else {
lexungetc(c1);
if (nparen == 0 && inprint)
c = WRITE;
}
break;
case '<':
if ((c1 = lexgetc()) == '=')
c = LE;
else
lexungetc(c1);
break;
case '!':
if ((c1 = lexgetc()) == '=')
c = NE;
else if (c1 == '~')
c = NRE;
else
lexungetc(c1);
break;
case '=':
if ((c1 = lexgetc()) == '=')
c = EQ;
else {
lexungetc(c1);
c = ASG;
}
break;
case '\n':
switch (lexlast) {
case ')':
if (catterm || inprint) {
c = ';';
break;
}
/* FALLTHROUGH */
case AND:
case OR:
case COMMA:
case '{':
case ELSE:
case ';':
case DO:
continue;
case '}':
if (nbrace != 0)
continue;
/* FALLTHROUGH */
default:
c = ';';
break;
}
break;
case ELSE:
if (lexlast != ';') {
savetoken = ELSE;
c = ';';
}
break;
case '(':
++nparen;
break;
case ')':
if (--nparen < 0)
awkerr(unbal, "()");
break;
case '{':
nbrace++;
break;
case '}':
if (--nbrace < 0) {
char brk[3];
brk[0] = '{';
brk[1] = '}';
brk[2] = '\0';
awkerr(unbal, brk);
}
if (lexlast != ';') {
savetoken = c;
c = ';';
}
break;
case '[':
++nbracket;
break;
case ']':
if (--nbracket < 0) {
char brk[3];
brk[0] = '[';
brk[1] = ']';
brk[2] = '\0';
awkerr(unbal, brk);
}
break;
case '\\':
if ((c1 = lexgetc()) == '\n')
continue;
lexungetc(c1);
break;
case ',':
c = COMMA;
break;
case '?':
c = QUEST;
break;
case ':':
c = COLON;
break;
default:
if (!iswprint(c))
awkerr(
gettext("invalid character \"%s\""),
toprint(c));
break;
}
break;
}
switch (c) {
case ']':
++catterm;
break;
case VAR:
if (catterm) {
savetoken = c;
c = CONCAT;
catterm = 0;
} else if (!isfuncdef) {
if ((c1 = lexgetc()) != '(')
++catterm;
lexungetc(c1);
}
isfuncdef = 0;
break;
case PARM:
case CONSTANT:
if (catterm) {
savetoken = c;
c = CONCAT;
catterm = 0;
} else {
if (lexlast == '$')
wasfield = 2;
++catterm;
}
break;
case INC:
case DEC:
if (!catterm || lexlast != CONSTANT || wasfield)
break;
/* FALLTHROUGH */
case UFUNC:
case FUNC:
case GETLINE:
case '!':
case '$':
case '(':
if (catterm) {
savetoken = c;
c = CONCAT;
catterm = 0;
}
break;
case '}':
if (nbrace == 0)
savetoken = ';';
/* FALLTHROUGH */
case ';':
inprint = 0;
/* FALLTHROUGH */
default:
if (c == DEFFUNC)
isfuncdef = 1;
catterm = 0;
}
lexlast = c;
if (wasfield)
wasfield--;
/*
* Map character constants to symbolic names.
*/
for (i = 0; ctosym[i].c != 0; i++)
if (c == ctosym[i].c) {
c = ctosym[i].sym;
break;
}
out:
#ifdef DEBUG
if (dflag)
(void) printf("%d\n", (int)c);
#endif
return ((int)c);
}
int
getToken(LexStream *L, AST **ast_ptr)
{
// int base = 10;
int c;
AST *ast = NULL;
int at_startofline = (L->eoln == 1);
int peekc;
c = skipSpace(L, &ast);
if (c >= 127) {
*ast_ptr = last_ast = ast;
return c;
} else if (safe_isdigit(c)) {
lexungetc(L,c);
ast = NewAST(AST_INTEGER, NULL, NULL);
c = parseNumber(L, 10, &ast->d.ival);
if (c == T_FLOATNUM)
ast->kind = AST_FLOAT;
} else if (c == '$') {
ast = NewAST(AST_INTEGER, NULL, NULL);
c = parseNumber(L, 16, &ast->d.ival);
} else if (c == '%') {
ast = NewAST(AST_INTEGER, NULL, NULL);
c = lexgetc(L);
if (c == '%') {
c = parseNumber(L, 4, &ast->d.ival);
} else {
lexungetc(L, c);
c = parseNumber(L, 2, &ast->d.ival);
}
} else if (isIdentifierStart(c)) {
lexungetc(L, c);
c = parseIdentifier(L, &ast, NULL);
/* if in pasm, and at start of line, restart temporary
labels */
if (c == T_IDENTIFIER && InDatBlock(L) && at_startofline) {
L->lastGlobal = ast->d.string;
}
} else if (c == ':') {
peekc = lexgetc(L);
if (peekc == '=') {
c = T_ASSIGN;
} else if (!gl_p2 && isIdentifierStart(peekc) && InDatBlock(L)) {
lexungetc(L, peekc);
c = parseIdentifier(L, &ast, L->lastGlobal ? L->lastGlobal : "");
} else {
lexungetc(L, peekc);
}
} else if (gl_p2 && c == '.' && isIdentifierStart(lexpeekc(L)) && InDatBlock(L)) {
c = parseIdentifier(L, &ast, L->lastGlobal ? L->lastGlobal : "");
} else if (strchr(operator_chars, c) != NULL) {
char op[6];
int i;
int token;
Symbol *sym = NULL;
op[0] = token = c;
for (i = 1; i < sizeof(op)-1; i++) {
c = lexgetc(L);
if (c >= 128 || strchr(operator_chars, c) == NULL) {
lexungetc(L, c);
break;
}
op[i] = c;
op[i+1] = 0;
sym = FindSymbol(&reservedWords, op);
if (sym) {
token = INTVAL(sym);
} else {
lexungetc(L, c);
break;
}
}
c = token;
} else if (c == '"') {
c = parseString(L, &ast);
}
*ast_ptr = last_ast = ast;
return c;
}
int
skipSpace(LexStream *L, AST **ast_ptr)
{
int c;
int commentNest;
int start_indent;
struct flexbuf cb;
AST *ast;
int startcol = 0;
int startline = 0;
flexbuf_init(&cb, INCSTR);
c = lexgetc(L);
again:
while (c == ' ' || c == '\t') {
c = lexgetc(L);
}
/* ignore completely empty lines or ones with just comments */
if (c == '\'') {
while (c == '\'') c = lexgetc(L);
while (c != '\n' && c != T_EOF) {
flexbuf_addchar(&cb, c);
c = lexgetc(L);
}
flexbuf_addchar(&cb, '\n');
flexbuf_addchar(&cb, 0);
ast = NewAST(AST_COMMENT, NULL, NULL);
ast->d.string = flexbuf_get(&cb);
comment_chain = AddToList(comment_chain, ast);
}
if (c == '{') {
struct flexbuf anno;
int annotate = 0;
int directive = 0;
int doccomment = 0;
startcol = L->colCounter;
startline = L->lineCounter;
flexbuf_init(&anno, INCSTR);
commentNest = 1;
/* check for special comments {++... } which indicate
inline C code
We also set up the preprocessor to emit {#line xx} directives when
doing #include
*/
c = lexgetc(L);
if (c == '+') {
c = lexgetc(L);
if (c == '+') {
annotate = 1;
c = lexgetc(L);
}
} else if (c == '#') {
c = lexgetc(L);
directive = 1;
} else if (c == '{') {
c = lexgetc(L);
doccomment = 1;
}
lexungetc(L, c);
for(;;) {
c = lexgetc(L);
if (c == '{' && !doccomment)
commentNest++;
else if (c == '}') {
if (doccomment) {
int peekc;
peekc = lexgetc(L);
if (peekc == '}') {
commentNest = 0;
} else {
lexungetc(L, peekc);
}
} else {
--commentNest;
}
}
if (commentNest <= 0 || c == T_EOF) {
break;
}
if (annotate || directive) {
flexbuf_addchar(&anno, c);
} else {
flexbuf_addchar(&cb, c);
}
}
if (c == T_EOF) {
if (commentNest > 0)
fprintf(stderr, "WARNING: EOF seen inside comment\n");
return c;
}
if (annotate) {
AST *ast = NewAST(AST_ANNOTATION, NULL, NULL);
flexbuf_addchar(&anno, '\0');
ast->d.string = flexbuf_get(&anno);
*ast_ptr = ast;
// if this is indented and inside a PUB or PRI,
// then treat it as inline C code
if (startcol > 1 && startline > L->block_firstline && (L->in_block == T_PUB || L->in_block == T_PRI)) {
return T_INLINECCODE;
}
return T_ANNOTATION;
} else if (directive) {
char *dir;
flexbuf_addchar(&anno, '\0');
dir = flexbuf_get(&anno);
if (!strncmp(dir, "line ", 5)) {
char *ptr = dir+5;
int lineno;
lineno = strtol(ptr, &ptr, 10);
if (lineno > 0) {
if (*ptr == ' ') ptr++;
L->fileName = strdup(ptr);
L->lineCounter = lineno;
}
}
free(dir);
} else {
flexbuf_addchar(&cb, '\0');
ast = NewAST(AST_COMMENT, NULL, NULL);
ast->d.string = flexbuf_get(&cb);
comment_chain = AddToList(comment_chain, ast);
}
c = lexgetc(L);
goto again;
}
if (L->eoln && (L->in_block == T_PUB || L->in_block == T_PRI)) {
if (c == '\n') {
c = lexgetc(L);
goto again;
}
/* if there is a pending indent, send it back */
if (L->pending_indent) {
lexungetc(L, c);
--L->pending_indent;
return T_INDENT;
}
/* on EOF send as many OUTDENTS as we need */
if (c == T_EOF) {
if (L->indentsp > 0) {
lexungetc(L, c);
--L->indentsp;
return T_OUTDENT;
}
}
/* if our indentation is <= the start value, send back an outdent */
start_indent = L->colCounter-1;
if (start_indent <= L->indent[L->indentsp] && L->indentsp > 0) {
lexungetc(L, c);
--L->indentsp;
return T_OUTDENT;
}
}
// force an end-of line at EOF
if (c == T_EOF && !L->eoln && !L->eof) {
L->eof = L->eoln = 1;
return T_EOLN;
}
if (L->eoln) {
L->eoln = 0;
L->firstNonBlank = L->colCounter-1;
}
if (c == '\n') {
L->eoln = 1;
return T_EOLN;
}
if (current && !current->sawToken) {
current->sawToken = 1;
current->topcomment = GetComments();
}
return c;
}
/* parse an identifier */
static int
parseIdentifier(LexStream *L, AST **ast_ptr, const char *prefix)
{
int c;
struct flexbuf fb;
Symbol *sym;
AST *ast = NULL;
int startColumn = L->colCounter - 1;
char *idstr;
flexbuf_init(&fb, INCSTR);
if (prefix) {
flexbuf_addmem(&fb, prefix, strlen(prefix));
if (gl_gas_dat) {
flexbuf_addchar(&fb, '.');
} else {
flexbuf_addchar(&fb, ':');
}
}
c = lexgetc(L);
while (isIdentifierChar(c)) {
//flexbuf_addchar(&fb, tolower(c));
flexbuf_addchar(&fb, c);
c = lexgetc(L);
}
// add a trailing 0, and make sure there is room for an extra
// character in case the name mangling needs it
flexbuf_addchar(&fb, '\0');
flexbuf_addchar(&fb, '\0');
idstr = flexbuf_get(&fb);
lexungetc(L, c);
/* check for reserved words */
if (InDatBlock(L)) {
sym = FindSymbol(&pasmWords, idstr);
if (sym) {
free(idstr);
if (sym->type == SYM_INSTR) {
ast = NewAST(AST_INSTR, NULL, NULL);
ast->d.ptr = sym->val;
*ast_ptr = ast;
return T_INSTR;
}
if (sym->type == SYM_INSTRMODIFIER) {
ast = NewAST(AST_INSTRMODIFIER, NULL, NULL);
ast->d.ptr = sym->val;
*ast_ptr = ast;
return T_INSTRMODIFIER;
}
fprintf(stderr, "Internal error: Unknown pasm symbol type %d\n", sym->type);
}
}
sym = FindSymbol(&reservedWords, idstr);
if (sym != NULL) {
if (sym->type == SYM_BUILTIN)
{
/* run any parse hooks */
Builtin *b = (Builtin *)sym->val;
if (b && b->parsehook) {
(*b->parsehook)(b);
}
goto is_identifier;
}
if (sym->type == SYM_CONSTANT
|| sym->type == SYM_FLOAT_CONSTANT)
{
goto is_identifier;
}
free(idstr);
if (sym->type == SYM_RESERVED) {
c = INTVAL(sym);
/* check for special handling */
switch(c) {
case T_PUB:
case T_PRI:
case T_DAT:
case T_OBJ:
case T_VAR:
case T_CON:
L->in_block = c;
L->block_firstline = L->lineCounter;
//EstablishIndent(L, 1);
break;
case T_ASM:
if (L->in_block == T_ASM) {
fprintf(stderr, "WARNING: ignoring nested asm\n");
} else {
L->save_block = L->in_block;
}
L->in_block = c;
break;
case T_ENDASM:
L->in_block = L->save_block;
break;
case T_IF:
case T_IFNOT:
case T_ELSE:
case T_ELSEIF:
case T_ELSEIFNOT:
case T_REPEAT:
case T_CASE:
EstablishIndent(L, startColumn);
break;
default:
break;
}
if (!ast)
ast = GetComments();
*ast_ptr = ast;
return c;
}
if (sym->type == SYM_HWREG) {
ast = NewAST(AST_HWREG, NULL, NULL);
ast->d.ptr = sym->val;
*ast_ptr = ast;
return T_HWREG;
}
fprintf(stderr, "Internal error: Unknown symbol type %d\n", sym->type);
}
is_identifier:
ast = NewAST(AST_IDENTIFIER, NULL, NULL);
/* make sure identifiers do not conflict with C keywords */
if (gl_normalizeIdents || Is_C_Reserved(idstr)) {
NormalizeIdentifier(idstr);
}
ast->d.string = idstr;
*ast_ptr = ast;
return T_IDENTIFIER;
}
/*
* actual parsing functions
*/
static int
parseNumber(LexStream *L, unsigned int base, uint32_t *num)
{
unsigned long uval, digit;
unsigned int c;
int sawdigit = 0;
int kind = T_NUM;
uval = 0;
for(;;) {
c = lexgetc(L);
if (c == '_')
continue;
else if (c >= 'A' && c <= 'Z')
digit = 10 + c - 'A';
else if (c >= 'a' && c <= 'z')
digit = 10 + c - 'a';
else if (c >= '0' && c <= '9') {
digit = (c - '0');
} else {
break;
}
if (digit < base) {
uval = base * uval + digit;
sawdigit = 1;
} else {
break;
}
}
if ( base == 10 && (c == '.' || c == 'e' || c == 'E') ) {
/* potential floating point number */
float f = (float)uval;
float ff = 0.0;
static float divby[45] = {
1e-1f, 1e-2f, 1e-3f, 1e-4f, 1e-5f,
1e-6f, 1e-7f, 1e-8f, 1e-9f, 1e-10f,
1e-11f, 1e-12f, 1e-13f, 1e-14f, 1e-15f,
1e-16f, 1e-17f, 1e-18f, 1e-19f, 1e-20f,
1e-21f, 1e-22f, 1e-23f, 1e-24f, 1e-25f,
1e-26f, 1e-27f, 1e-28f, 1e-29f, 1e-30f,
1e-31f, 1e-32f, 1e-33f, 1e-34f, 1e-35f,
1e-36f, 1e-37f, 1e-38f, 1e-39f, 1e-40f,
1e-41f, 1e-42f, 1e-43f, 1e-44f, 1e-45f,
};
int counter = 0;
int exponent = 0;
if (c == '.') {
c = lexgetc(L);
if ( c != 'e' && c != 'E' && (c < '0' || c > '9')) {
lexungetc(L, c);
c = '.';
goto donefloat;
}
}
while (c >= '0' && c <= '9') {
ff = ff + divby[counter]*(float)(c-'0');
c = lexgetc(L);
counter++;
}
if (c == 'e' || c == 'E') {
int expval = 0;
int neg = 1;
c = lexgetc(L);
if (c == '+') {
c = lexgetc(L);
} else if (c == '-') {
c = lexgetc(L);
neg = -neg;
}
while (c >= '0' && c <= '9') {
expval = 10*expval + (c - '0');
c = lexgetc(L);
}
if (neg < 0)
expval = -expval;
exponent += expval;
}
f = f + ff;
if (exponent < 0 && exponent >= -45) {
f *= divby[-(exponent+1)];
} else if (exponent != 0) {
f *= powf(10.0f, (float)exponent);
}
uval = floatAsInt(f);
kind = T_FLOATNUM;
}
donefloat:
lexungetc(L, c);
*num = uval;
return sawdigit ? kind : ((base == 16) ? T_HERE : '%');
}
yylex()
{
int k;
STRING synfind(),strsave(),getident(),getword2(),getstring(),getword();
float getnum();
while(iswhite(c=lexgetc()));
startoflex = curr_index;
if((isalpha(c)||c=='@' )&& c != EOF) {
s = getident(c);
if(strcmp("include",s)==0){
inclusion();
return(yylex());
}else if((k=keyfind(synfind(s)))!=NKEYWORDS && const_list == 0){
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg = keywords[k].keyname;
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len = startoflex;
return(keywords[k].keyret);
}else{ if (const_list != 0)
{
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg = s;
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD);
}
else
{
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg = s;
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(IDENT);
}
}
}else if((isdigit(c) || c == '~') && c !=EOF) {
yylval.numb = (float) getnum(c);
return(CONST);
}else{
switch(c){
case '+':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
yylval.resptr->strg = "plus";
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(ADDOP);
case '-':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
yylval.resptr->strg = "minus";
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(ADDOP);
case '*':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
c=lexgetc();
yylval.resptr->line = in_files[in_index].in_line;
if (c=='*') { yylval.resptr->strg = "exp";
yylval.resptr->len = startoflex;
return(EXP); }
peekc = c;
yylval.resptr->strg = "times";
yylval.resptr->len = startoflex;
return(MULOP);
case '/': c = lexgetc();
if (const_list!=0) {
peekc = c;
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg=getword2('/');
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
if (c=='/'){
while((c=lexgetc())!='\n');
peekc = c;
return(yylex()); }
peekc = c;
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg = "fdiv";
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(MULOP);
case '^':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
yylval.resptr->strg ="strconc";
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(STRCONC);
case '#':
if (const_list!=0) {
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->strg=getword2(c);
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len =
startoflex;
return(WORD); }
return(c);
case '"':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->len =
startoflex;
return(WORD); }
c = lexgetc() ;
yylval.resptr->strg = getword(c);
yylval.resptr->len =
startoflex;
return(WORD);
case '`':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
c = lexgetc();
yylval.resptr->strg = getstring(c);
yylval.resptr->len =
startoflex;
return(STRING_QUOTED);
case ':':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len = startoflex;
if (const_list!=0) {
yylval.resptr->strg = getword(c);
return(WORD); }
c = lexgetc();
if (c==':') { yylval.resptr->strg = "cons";
return(CONS); } else {
peekc=c;
return(COLON);}
case '[': c = lexgetc();
if (c=='%' && const_list!=0) {
peekc = c;
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->strg=getword2('[');
yylval.resptr->len =
startoflex;
return(WORD); }
if (c == '%') { return(CLSTART);
}
else
{ peekc = c;
const_list++;
return(LSTART);
}
case '%':c = lexgetc();
if ((c==']' || c==')') && const_list!=0) {
peekc = c;
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->strg=getword2('%');
yylval.resptr->len =
startoflex;
return(WORD); }
if (c == ']')
{ return(CLEND);
}
else
{
my_yyerror("incorrect use of %","cannot continue compliation ");
my_exit(1);
}
case ']':
{ const_list--;
return(LEND);
}
case '.':
if (const_list!=0) {
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->strg=getword2(c);
yylval.resptr->len =
startoflex;
return(WORD); }
return(DOT);
case '<':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->len =
startoflex;
return(WORD); }
c=lexgetc();
if (c=='=') {
yylval.resptr->strg = "le";
yylval.resptr->len =
startoflex;
return(RELOP); } else
if (c =='>') {
yylval.resptr->strg = "append";
yylval.resptr->len =
startoflex;
return(APPEND); } else {
peekc=c;
yylval.resptr->strg = "lt";
yylval.resptr->len =
startoflex;
return(RELOP); }
case '>':
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
if (const_list!=0) {
yylval.resptr->strg=getword2(c);
yylval.resptr->len =
startoflex;
return(WORD); }
c=lexgetc();
if (c=='=') { yylval.resptr->strg = "ge";
yylval.resptr->len =
startoflex;
return(RELOP); } else {
yylval.resptr->strg = "gt";
yylval.resptr->len =
startoflex;
peekc=c;
return(RELOP); }
case '&':
case ',':
case ';':
case '=':
case '$':
case '(':
case ')':
if (const_list!=0) {
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->strg=getword2(c);
yylval.resptr->len =
startoflex;
return(WORD); }
if (c==',') {
yylval.resptr = (RESPTR) calloc(1,sizeof(RES));
yylval.resptr->line = in_files[in_index].in_line;
yylval.resptr->len = startoflex;
return(COMMA); }
default:
return(c);
}
}
}