As discussed on python-dev, really fix the PyMem_*/PyObject_* memory API

mismatches.  At least I hope this fixes them all.

This reverts part of my change from yesterday that converted everything
in Parser/*.c to use PyObject_* API.  The encoding doesn't really need
to use PyMem_*, however, it uses new_string() which must return PyMem_*
for handling the result of PyOS_Readline() which returns PyMem_* memory.

If there were 2 versions of new_string() one that returned PyMem_*
for tokens and one that return PyObject_* for encodings that could
also fix this problem.  I'm not sure which version would be clearer.
This seems to fix both Guido's and Phillip's problems, so it's good enough
for now.  After this change, it would be good to review Parser/*.c
for consistent use of the 2 memory APIs.



git-svn-id: http://svn.python.org/projects/python/trunk@45273 6015fed2-1504-0410-9fe1-9d1591cc4771

Author: neal.norwitz

Parser/myreadline.c

@@ -111,7 +111,7 @@ PyOS_StdioReadline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 	size_t n;
 	char *p;
 	n = 100;
-	if ((p = (char *)PyObject_MALLOC(n)) == NULL)
+	if ((p = (char *)PyMem_MALLOC(n)) == NULL)
 		return NULL;
 	fflush(sys_stdout);
 #ifndef RISCOS
@@ -130,7 +130,7 @@ PyOS_StdioReadline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 	case 0: /* Normal case */
 		break;
 	case 1: /* Interrupt */
-		PyObject_FREE(p);
+		PyMem_FREE(p);
 		return NULL;
 	case -1: /* EOF */
 	case -2: /* Error */
@@ -141,7 +141,7 @@ PyOS_StdioReadline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 	n = strlen(p);
 	while (n > 0 && p[n-1] != '\n') {
 		size_t incr = n+2;
-		p = (char *)PyObject_REALLOC(p, n + incr);
+		p = (char *)PyMem_REALLOC(p, n + incr);
 		if (p == NULL)
 			return NULL;
 		if (incr > INT_MAX) {
@@ -151,14 +151,14 @@ PyOS_StdioReadline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 			break;
 		n += strlen(p+n);
 	}
-	return (char *)PyObject_REALLOC(p, n+1);
+	return (char *)PyMem_REALLOC(p, n+1);
 }
 
 
 /* By initializing this function pointer, systems embedding Python can
    override the readline function.
 
-   Note: Python expects in return a buffer allocated with PyObject_Malloc. */
+   Note: Python expects in return a buffer allocated with PyMem_Malloc. */
 
 char *(*PyOS_ReadlineFunctionPointer)(FILE *, FILE *, char *);
 

Parser/pgenmain.c

@@ -136,7 +136,7 @@ char *
 PyOS_Readline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 {
 	size_t n = 1000;
-	char *p = PyObject_MALLOC(n);
+	char *p = PyMem_MALLOC(n);
 	char *q;
 	if (p == NULL)
 		return NULL;
@@ -149,7 +149,7 @@ PyOS_Readline(FILE *sys_stdin, FILE *sys_stdout, char *prompt)
 	n = strlen(p);
 	if (n > 0 && p[n-1] != '\n')
 		p[n-1] = '\n';
-	return PyObject_REALLOC(p, n+1);
+	return PyMem_REALLOC(p, n+1);
 }
 
 /* No-nonsense fgets */

Parser/tokenizer.c

@@ -164,15 +164,15 @@ error_ret(struct tok_state *tok) /* XXX */
 {
 	tok->decoding_erred = 1;
 	if (tok->fp != NULL && tok->buf != NULL) /* see PyTokenizer_Free */
-		PyObject_FREE(tok->buf);
+		PyMem_FREE(tok->buf);
 	tok->buf = NULL;
 	return NULL;		/* as if it were EOF */
 }
 
 static char *
 new_string(const char *s, Py_ssize_t len)
 {
-	char* result = (char *)PyObject_MALLOC(len + 1);
+	char* result = (char *)PyMem_MALLOC(len + 1);
 	if (result != NULL) {
 		memcpy(result, s, len);
 		result[len] = '\0';
@@ -237,7 +237,7 @@ get_coding_spec(const char *s, Py_ssize_t size)
 				char* r = new_string(begin, t - begin);
 				char* q = get_normal_name(r);
 				if (r != q) {
-					PyObject_FREE(r);
+					PyMem_FREE(r);
 					r = new_string(q, strlen(q));
 				}
 				return r;
@@ -278,18 +278,18 @@ check_coding_spec(const char* line, Py_ssize_t size, struct tok_state *tok,
 					tok->decoding_state = -1;
 				}
 				else
-					PyObject_FREE(cs);
+					PyMem_FREE(cs);
 #else
                                 /* Without Unicode support, we cannot
                                    process the coding spec. Since there
                                    won't be any Unicode literals, that
                                    won't matter. */
-				PyObject_FREE(cs);
+				PyMem_FREE(cs);
 #endif
 			}
 		} else {	/* then, compare cs with BOM */
 			r = (strcmp(tok->encoding, cs) == 0);
-			PyObject_FREE(cs);
+			PyMem_FREE(cs);
 		}
 	}
 	if (!r) {
@@ -335,7 +335,7 @@ check_bom(int get_char(struct tok_state *),
 		return 1;
 	}
 	if (tok->encoding != NULL)
-		PyObject_FREE(tok->encoding);
+		PyMem_FREE(tok->encoding);
 	tok->encoding = new_string("utf-8", 5);	/* resulting is in utf-8 */
 	return 1;
   NON_BOM:
@@ -657,7 +657,7 @@ PyTokenizer_FromFile(FILE *fp, char *ps1, char *ps2)
 	struct tok_state *tok = tok_new();
 	if (tok == NULL)
 		return NULL;
-	if ((tok->buf = (char *)PyObject_MALLOC(BUFSIZ)) == NULL) {
+	if ((tok->buf = (char *)PyMem_MALLOC(BUFSIZ)) == NULL) {
 		PyTokenizer_Free(tok);
 		return NULL;
 	}
@@ -676,13 +676,13 @@ void
 PyTokenizer_Free(struct tok_state *tok)
 {
 	if (tok->encoding != NULL)
-		PyObject_FREE(tok->encoding);
+		PyMem_FREE(tok->encoding);
 #ifndef PGEN
 	Py_XDECREF(tok->decoding_readline);
 	Py_XDECREF(tok->decoding_buffer);
 #endif
 	if (tok->fp != NULL && tok->buf != NULL)
-		PyObject_FREE(tok->buf);
+		PyMem_FREE(tok->buf);
 	PyMem_FREE(tok);
 }
 
@@ -722,10 +722,10 @@ tok_stdin_decode(struct tok_state *tok, char **inp)
 	if (converted == NULL)
 		goto error_nomem;
 
-	PyObject_FREE(*inp);
+	PyMem_FREE(*inp);
 	*inp = converted;
 	if (tok->encoding != NULL)
-		PyObject_FREE(tok->encoding);
+		PyMem_FREE(tok->encoding);
 	tok->encoding = new_string(encoding, strlen(encoding));
 	if (tok->encoding == NULL)
 		goto error_nomem;
@@ -782,40 +782,40 @@ tok_nextc(register struct tok_state *tok)
 			if (newtok == NULL)
 				tok->done = E_INTR;
 			else if (*newtok == '\0') {
-				PyObject_FREE(newtok);
+				PyMem_FREE(newtok);
 				tok->done = E_EOF;
 			}
 #if !defined(PGEN) && defined(Py_USING_UNICODE)
 			else if (tok_stdin_decode(tok, &newtok) != 0)
-				PyObject_FREE(newtok);
+				PyMem_FREE(newtok);
 #endif
 			else if (tok->start != NULL) {
 				size_t start = tok->start - tok->buf;
 				size_t oldlen = tok->cur - tok->buf;
 				size_t newlen = oldlen + strlen(newtok);
 				char *buf = tok->buf;
-				buf = (char *)PyObject_REALLOC(buf, newlen+1);
+				buf = (char *)PyMem_REALLOC(buf, newlen+1);
 				tok->lineno++;
 				if (buf == NULL) {
-					PyObject_FREE(tok->buf);
+					PyMem_FREE(tok->buf);
 					tok->buf = NULL;
-					PyObject_FREE(newtok);
+					PyMem_FREE(newtok);
 					tok->done = E_NOMEM;
 					return EOF;
 				}
 				tok->buf = buf;
 				tok->cur = tok->buf + oldlen;
 				tok->line_start = tok->cur;
 				strcpy(tok->buf + oldlen, newtok);
-				PyObject_FREE(newtok);
+				PyMem_FREE(newtok);
 				tok->inp = tok->buf + newlen;
 				tok->end = tok->inp + 1;
 				tok->start = tok->buf + start;
 			}
 			else {
 				tok->lineno++;
 				if (tok->buf != NULL)
-					PyObject_FREE(tok->buf);
+					PyMem_FREE(tok->buf);
 				tok->buf = newtok;
 				tok->line_start = tok->buf;
 				tok->cur = tok->buf;
@@ -831,7 +831,7 @@ tok_nextc(register struct tok_state *tok)
 			if (tok->start == NULL) {
 				if (tok->buf == NULL) {
 					tok->buf = (char *)
-						PyObject_MALLOC(BUFSIZ);
+						PyMem_MALLOC(BUFSIZ);
 					if (tok->buf == NULL) {
 						tok->done = E_NOMEM;
 						return EOF;
@@ -866,8 +866,8 @@ tok_nextc(register struct tok_state *tok)
 				Py_ssize_t curvalid = tok->inp - tok->buf;
 				Py_ssize_t newsize = curvalid + BUFSIZ;
 				char *newbuf = tok->buf;
-				newbuf = (char *)PyObject_REALLOC(newbuf,
-								  newsize);
+				newbuf = (char *)PyMem_REALLOC(newbuf,
+							       newsize);
 				if (newbuf == NULL) {
 					tok->done = E_NOMEM;
 					tok->cur = tok->inp;