Fix memory management issues with expanded %include

Ownership of the filename string was a bit fuzzy, with the result that
we were freeing it even though it was retained for use by __FILE__.
Clean up a number of other memory management issues with the new
quoting code, and change the stdscan implementation to one pass over
the string.

4 files changed, 83 insertions, 59 deletions

diff --git a/preproc.c b/preproc.c
index ac289ca..4052672 100644
--- a/preproc.c
+++ b/preproc.c
@@ -1144,11 +1144,18 @@ static int ppscan(void *private_data, struct tokenval *tokval)
     }
 
     if (tline->type == TOK_STRING) {
+	char bq, *ep;
+	bool errquote;
 	bool rn_warn;
         size_t l;
 
-        l = nasm_unquote(tline->text);
-	/* TOKEN_ERRNUM if improperly quoted... */
+	bq = tline->text[0];
+        l = nasm_unquote(tline->text, &ep);
+	if (ep[0] != bq || ep[1] != '\0')
+	    errquote = true;
+	
+	if (errquote)
+	    return tokval->t_type = TOKEN_ERRNUM;
 
         tokval->t_integer = readstrnum(tline->text, l, &rn_warn);
         if (rn_warn)
@@ -1520,8 +1527,8 @@ static bool if_condition(Token * tline, enum preproc_token ct)
             }
             /* When comparing strings, need to unquote them first */
             if (t->type == TOK_STRING) {
-		size_t l1 = nasm_unquote(t->text);
-		size_t l2 = nasm_unquote(tt->text);
+		size_t l1 = nasm_unquote(t->text, NULL);
+		size_t l2 = nasm_unquote(tt->text, NULL);
 		
 		if (l1 != l2) {
 		    j = false;
@@ -2081,20 +2088,21 @@ static int do_directive(Token * tline)
         return DIRECTIVE_FOUND;
 
     case PP_DEPEND:
-	tline = expand_smacro(tline->next);
-        skip_white_(tline);
-        if (!tline || (tline->type != TOK_STRING &&
-                       tline->type != TOK_INTERNAL_STRING)) {
+	t = tline->next = expand_smacro(tline->next);
+        skip_white_(t);
+        if (!t || (t->type != TOK_STRING &&
+                       t->type != TOK_INTERNAL_STRING)) {
             error(ERR_NONFATAL, "`%%depend' expects a file name");
+	    free_tlist(t);
             free_tlist(origline);
             return DIRECTIVE_FOUND;     /* but we did _something_ */
         }
-        if (tline->next)
+        if (t->next)
             error(ERR_WARNING,
                   "trailing garbage after `%%depend' ignored");
-	p = tline->text;
-        if (tline->type != TOK_INTERNAL_STRING)
-	    nasm_unquote(p);
+	p = t->text;
+        if (t->type != TOK_INTERNAL_STRING)
+	    nasm_unquote(p, NULL);
 	if (dephead && !in_list(*dephead, p)) {
 	    StrList *sl = nasm_malloc(strlen(p)+1+sizeof sl->next);
 	    sl->next = NULL;
@@ -2102,25 +2110,26 @@ static int do_directive(Token * tline)
 	    *deptail = sl;
 	    deptail = &sl->next;
 	}
+	free_tlist(t);
 	free_tlist(origline);
         return DIRECTIVE_FOUND;
 
     case PP_INCLUDE:
-	tline = expand_smacro(tline->next);
-        skip_white_(tline);
+	t = tline->next = expand_smacro(tline->next);
+        skip_white_(t);
 	
-        if (!tline || (tline->type != TOK_STRING &&
-                       tline->type != TOK_INTERNAL_STRING)) {
+        if (!t || (t->type != TOK_STRING &&
+                       t->type != TOK_INTERNAL_STRING)) {
             error(ERR_NONFATAL, "`%%include' expects a file name");
             free_tlist(origline);
             return DIRECTIVE_FOUND;     /* but we did _something_ */
         }
-        if (tline->next)
+        if (t->next)
             error(ERR_WARNING,
                   "trailing garbage after `%%include' ignored");
-	p = tline->text;
-        if (tline->type != TOK_INTERNAL_STRING)
-	    nasm_unquote(p);
+	p = t->text;
+        if (t->type != TOK_INTERNAL_STRING)
+	    nasm_unquote(p, NULL);
         inc = nasm_malloc(sizeof(Include));
         inc->next = istk;
         inc->conds = NULL;
@@ -2129,7 +2138,7 @@ static int do_directive(Token * tline)
 	    /* -MG given but file not found */
 	    nasm_free(inc);
 	} else {
-	    inc->fname = src_set_fname(p);
+	    inc->fname = src_set_fname(nasm_strdup(p));
 	    inc->lineno = src_set_linnum(0);
 	    inc->lineinc = 1;
 	    inc->expansion = NULL;
@@ -2196,7 +2205,7 @@ static int do_directive(Token * tline)
         skip_white_(tline);
         if (tok_type_(tline, TOK_STRING)) {
 	    p = tline->text;
-	    nasm_unquote(p);
+	    nasm_unquote(p, NULL);
 	    expand_macros_in_string(&p); /* WHY? */
             error(ERR_NONFATAL, "%s", p);
             nasm_free(p);
@@ -2681,7 +2690,7 @@ static int do_directive(Token * tline)
                   "trailing garbage after `%%pathsearch' ignored");
 	p = t->text;
         if (t->type != TOK_INTERNAL_STRING)
-	    nasm_unquote(p);
+	    nasm_unquote(p, NULL);
 
 	fp = inc_fopen(p, &xsl, &xsl, true);
 	if (fp) {
@@ -2742,7 +2751,7 @@ static int do_directive(Token * tline)
 
         macro_start = nasm_malloc(sizeof(*macro_start));
         macro_start->next = NULL;
-        make_tok_num(macro_start, nasm_unquote(t->text));
+        make_tok_num(macro_start, nasm_unquote(t->text, NULL));
         macro_start->mac = NULL;
 
         /*
@@ -2831,7 +2840,7 @@ static int do_directive(Token * tline)
 	    a2 = evalresult->value;
 	}
 
-	len = nasm_unquote(t->text);
+	len = nasm_unquote(t->text, NULL);
 	if (a2 < 0)
 	    a2 = a2+1+len-a1;
 	if (a1+a2 > (int64_t)len)
diff --git a/quote.c b/quote.c
index 6bfded3..5a04d36 100644
--- a/quote.c
+++ b/quote.c
@@ -183,12 +183,13 @@ static char *emit_utf8(char *q, int32_t v)
  *
  * In-place replacement is possible since the unquoted length is always
  * shorter than or equal to the quoted length.
+ *
+ * *ep points to the final quote, or to the null if improperly quoted.
  */
-size_t nasm_unquote(char *str)
+size_t nasm_unquote(char *str, char **ep)
 {
-    size_t ln;
-    char bq, eq;
-    char *p, *q, *ep;
+    char bq;
+    char *p, *q;
     char *escp = NULL;
     char c;
     enum unq_state {
@@ -201,33 +202,42 @@ size_t nasm_unquote(char *str)
     int ndig = 0;
     int32_t nval = 0;
 
-    bq = str[0];
+    p = q = str;
+    
+    bq = *p++;
     if (!bq)
 	return 0;
-    ln = strlen(str);
-    eq = str[ln-1];
 
-    if ((bq == '\'' || bq == '\"') && bq == eq) {
+    switch (bq) {
+    case '\'':
+    case '\"':
 	/* '...' or "..." string */
-	memmove(str, str+1, ln-2);
-	str[ln-2] = '\0';
-	return ln-2;
-    }
-    if (bq == '`' || eq == '`') {
+	while ((c = *p) && c != bq) {
+	    p++;
+	    *q++ = c;
+	}
+	*q = '\0';
+	break;
+
+    case '`':
 	/* `...` string */
-	q = str;
-	p = str+1;
-	ep = str+ln-1;
 	state = st_start;
 
-	while (p < ep) {
-	    c = *p++;
+	while ((c = *p)) {
+	    p++;
 	    switch (state) {
 	    case st_start:
-		if (c == '\\')
+		switch (c) {
+		case '\\':
 		    state = st_backslash;
-		else
+		    break;
+		case '`':
+		    p--;
+		    goto out;
+		default:
 		    *q++ = c;
+		    break;
+		}
 		break;
 
 	    case st_backslash:
@@ -357,12 +367,18 @@ size_t nasm_unquote(char *str)
 		*q++ = escp[-1];
 	    break;
 	}
-	*q = '\0';
-	return q-str;
+    out:
+	break;
+
+    default:
+	/* Not a quoted string, just return the input... */
+	p = q = strchr(str, '\0');
+	break;
     }
 
-    /* Otherwise, just return the input... */
-    return ln;
+    if (ep)
+	*ep = p;
+    return q-str;
 }
 
 /*
diff --git a/quote.h b/quote.h
index 501f735..5f96159 100644
--- a/quote.h
+++ b/quote.h
@@ -4,7 +4,7 @@
 #include "compiler.h"
 
 char *nasm_quote(char *str, size_t len);
-size_t nasm_unquote(char *str);
+size_t nasm_unquote(char *str, char **endptr);
 char *nasm_skip_string(char *str);
 
 #endif /* NASM_QUOTE_H */
diff --git a/stdscan.c b/stdscan.c
index 0fdd7af..4db1368 100644
--- a/stdscan.c
+++ b/stdscan.c
@@ -47,7 +47,7 @@ static char *stdscan_copy(char *p, int len)
     char *text;
 
     text = nasm_malloc(len + 1);
-    strncpy(text, p, len);
+    memcpy(text, p, len);
     text[len] = '\0';
 
     if (stdscan_templen >= stdscan_tempsize) {
@@ -176,15 +176,14 @@ int stdscan(void *private_data, struct tokenval *tv)
 	}
     } else if (*stdscan_bufptr == '\'' || *stdscan_bufptr == '"' ||
 	       *stdscan_bufptr == '`') {
-	/* a char constant */
-	char s;
+	/* a quoted string */
         bool rn_warn;
-	stdscan_bufptr = nasm_skip_string(tv->t_charptr = stdscan_bufptr);
-	s = *stdscan_bufptr;
-	tv->t_inttwo = nasm_unquote(tv->t_charptr);
-        if (!s)
-            return tv->t_type = TOKEN_ERRNUM;   /* unmatched quotes */
-        stdscan_bufptr++;       /* skip over final quote */
+	char start_quote = *stdscan_bufptr;
+	tv->t_charptr = stdscan_bufptr;
+	tv->t_inttwo = nasm_unquote(tv->t_charptr, &stdscan_bufptr);
+	if (*stdscan_bufptr != start_quote)
+	    return tv->t_type = TOKEN_ERRNUM;
+	stdscan_bufptr++;	/* Skip final quote */
         tv->t_integer = readstrnum(tv->t_charptr, tv->t_inttwo, &rn_warn);
 	/* Issue: can't readily check rn_warn, because we might be in
 	   a db family context... */