cpython/Python/marshal.c

1248 lines
26 KiB
C
Raw Normal View History

1991-06-04 16:42:30 -03:00
/* Write Python objects to files and read them back.
This is intended for writing and reading compiled Python code only;
a true persistent storage facility would be much harder, since
it would have to take circular links and sharing into account. */
2006-03-01 19:49:13 -04:00
#define PY_SSIZE_T_CLEAN
#include "Python.h"
1991-06-04 16:42:30 -03:00
#include "longintrepr.h"
#include "code.h"
1991-06-04 16:42:30 -03:00
#include "marshal.h"
/* High water mark to determine when the marshalled object is dangerously deep
* and risks coring the interpreter. When the object stack gets this deep,
* raise an exception instead of continuing.
*/
#define MAX_MARSHAL_STACK_DEPTH 2000
#define TYPE_NULL '0'
#define TYPE_NONE 'N'
#define TYPE_FALSE 'F'
#define TYPE_TRUE 'T'
#define TYPE_STOPITER 'S'
#define TYPE_ELLIPSIS '.'
#define TYPE_INT 'i'
#define TYPE_INT64 'I'
#define TYPE_FLOAT 'f'
#define TYPE_BINARY_FLOAT 'g'
#define TYPE_COMPLEX 'x'
#define TYPE_BINARY_COMPLEX 'y'
#define TYPE_LONG 'l'
#define TYPE_STRING 's'
#define TYPE_INTERNED 't'
#define TYPE_STRINGREF 'R'
#define TYPE_TUPLE '('
#define TYPE_LIST '['
#define TYPE_DICT '{'
#define TYPE_CODE 'c'
#define TYPE_UNICODE 'u'
#define TYPE_UNKNOWN '?'
#define TYPE_SET '<'
#define TYPE_FROZENSET '>'
1991-06-04 16:42:30 -03:00
typedef struct {
FILE *fp;
int error;
int depth;
/* If fp == NULL, the following are valid: */
PyObject *str;
char *ptr;
char *end;
PyObject *strings; /* dict on marshal, list on unmarshal */
int version;
} WFILE;
1991-06-04 16:42:30 -03:00
#define w_byte(c, p) if (((p)->fp)) putc((c), (p)->fp); \
else if ((p)->ptr != (p)->end) *(p)->ptr++ = (c); \
else w_more(c, p)
static void
w_more(int c, WFILE *p)
{
2006-02-15 13:27:45 -04:00
Py_ssize_t size, newsize;
if (p->str == NULL)
return; /* An error already occurred */
size = PyString_Size(p->str);
newsize = size + size + 1024;
if (newsize > 32*1024*1024) {
newsize = size + (size >> 3); /* 12.5% overallocation */
}
if (_PyString_Resize(&p->str, newsize) != 0) {
p->ptr = p->end = NULL;
}
else {
p->ptr = PyString_AS_STRING((PyStringObject *)p->str) + size;
p->end =
PyString_AS_STRING((PyStringObject *)p->str) + newsize;
*p->ptr++ = Py_SAFE_DOWNCAST(c, int, char);
}
}
static void
w_string(char *s, int n, WFILE *p)
{
if (p->fp != NULL) {
fwrite(s, 1, n, p->fp);
}
else {
while (--n >= 0) {
w_byte(*s, p);
s++;
}
}
}
static void
w_short(int x, WFILE *p)
1991-06-04 16:42:30 -03:00
{
w_byte((char)( x & 0xff), p);
w_byte((char)((x>> 8) & 0xff), p);
1991-06-04 16:42:30 -03:00
}
static void
w_long(long x, WFILE *p)
1991-06-04 16:42:30 -03:00
{
2002-07-30 08:44:44 -03:00
w_byte((char)( x & 0xff), p);
w_byte((char)((x>> 8) & 0xff), p);
w_byte((char)((x>>16) & 0xff), p);
w_byte((char)((x>>24) & 0xff), p);
1991-06-04 16:42:30 -03:00
}
#if SIZEOF_LONG > 4
static void
w_long64(long x, WFILE *p)
{
w_long(x, p);
w_long(x>>32, p);
}
#endif
static void
w_object(PyObject *v, WFILE *p)
1991-06-04 16:42:30 -03:00
{
2006-02-15 13:27:45 -04:00
Py_ssize_t i, n;
p->depth++;
if (p->depth > MAX_MARSHAL_STACK_DEPTH) {
p->error = 2;
}
else if (v == NULL) {
w_byte(TYPE_NULL, p);
1998-04-10 19:27:42 -03:00
}
else if (v == Py_None) {
w_byte(TYPE_NONE, p);
1998-04-10 19:27:42 -03:00
}
else if (v == PyExc_StopIteration) {
w_byte(TYPE_STOPITER, p);
}
1998-04-10 19:27:42 -03:00
else if (v == Py_Ellipsis) {
w_byte(TYPE_ELLIPSIS, p);
}
else if (v == Py_False) {
w_byte(TYPE_FALSE, p);
}
else if (v == Py_True) {
w_byte(TYPE_TRUE, p);
}
else if (PyInt_Check(v)) {
long x = PyInt_AS_LONG((PyIntObject *)v);
#if SIZEOF_LONG > 4
long y = Py_ARITHMETIC_RIGHT_SHIFT(long, x, 31);
if (y && y != -1) {
w_byte(TYPE_INT64, p);
w_long64(x, p);
}
else
#endif
{
w_byte(TYPE_INT, p);
w_long(x, p);
}
1991-06-04 16:42:30 -03:00
}
else if (PyLong_Check(v)) {
PyLongObject *ob = (PyLongObject *)v;
w_byte(TYPE_LONG, p);
1991-06-04 16:42:30 -03:00
n = ob->ob_size;
w_long((long)n, p);
1991-06-04 16:42:30 -03:00
if (n < 0)
n = -n;
for (i = 0; i < n; i++)
w_short(ob->ob_digit[i], p);
1991-06-04 16:42:30 -03:00
}
else if (PyFloat_Check(v)) {
if (p->version > 1) {
unsigned char buf[8];
if (_PyFloat_Pack8(PyFloat_AsDouble(v),
buf, 1) < 0) {
p->error = 1;
return;
}
w_byte(TYPE_BINARY_FLOAT, p);
w_string((char*)buf, 8, p);
}
else {
char buf[256]; /* Plenty to format any double */
PyFloat_AsReprString(buf, (PyFloatObject *)v);
2006-02-16 10:37:48 -04:00
n = strlen(buf);
w_byte(TYPE_FLOAT, p);
2006-02-16 10:37:48 -04:00
w_byte((int)n, p);
w_string(buf, (int)n, p);
}
1991-06-04 16:42:30 -03:00
}
1996-01-11 21:09:56 -04:00
#ifndef WITHOUT_COMPLEX
else if (PyComplex_Check(v)) {
if (p->version > 1) {
unsigned char buf[8];
if (_PyFloat_Pack8(PyComplex_RealAsDouble(v),
buf, 1) < 0) {
p->error = 1;
return;
}
w_byte(TYPE_BINARY_COMPLEX, p);
w_string((char*)buf, 8, p);
if (_PyFloat_Pack8(PyComplex_ImagAsDouble(v),
buf, 1) < 0) {
p->error = 1;
return;
}
w_string((char*)buf, 8, p);
}
else {
char buf[256]; /* Plenty to format any double */
PyFloatObject *temp;
w_byte(TYPE_COMPLEX, p);
temp = (PyFloatObject*)PyFloat_FromDouble(
PyComplex_RealAsDouble(v));
if (!temp) {
p->error = 1;
return;
}
PyFloat_AsReprString(buf, temp);
Py_DECREF(temp);
n = strlen(buf);
w_byte((int)n, p);
w_string(buf, (int)n, p);
temp = (PyFloatObject*)PyFloat_FromDouble(
PyComplex_ImagAsDouble(v));
if (!temp) {
p->error = 1;
return;
}
PyFloat_AsReprString(buf, temp);
Py_DECREF(temp);
n = strlen(buf);
w_byte((int)n, p);
w_string(buf, (int)n, p);
}
1996-01-11 21:09:56 -04:00
}
#endif
else if (PyString_Check(v)) {
if (p->strings && PyString_CHECK_INTERNED(v)) {
PyObject *o = PyDict_GetItem(p->strings, v);
if (o) {
long w = PyInt_AsLong(o);
w_byte(TYPE_STRINGREF, p);
w_long(w, p);
goto exit;
}
else {
int ok;
2006-02-15 13:27:45 -04:00
o = PyInt_FromSsize_t(PyDict_Size(p->strings));
ok = o &&
PyDict_SetItem(p->strings, v, o) >= 0;
Py_XDECREF(o);
if (!ok) {
p->depth--;
p->error = 1;
return;
}
w_byte(TYPE_INTERNED, p);
}
}
else {
w_byte(TYPE_STRING, p);
}
n = PyString_GET_SIZE(v);
if (n > INT_MAX) {
/* huge strings are not supported */
p->depth--;
p->error = 1;
return;
}
w_long((long)n, p);
w_string(PyString_AS_STRING(v), (int)n, p);
}
#ifdef Py_USING_UNICODE
else if (PyUnicode_Check(v)) {
PyObject *utf8;
utf8 = PyUnicode_AsUTF8String(v);
if (utf8 == NULL) {
p->depth--;
p->error = 1;
return;
}
w_byte(TYPE_UNICODE, p);
n = PyString_GET_SIZE(utf8);
if (n > INT_MAX) {
p->depth--;
p->error = 1;
return;
}
w_long((long)n, p);
w_string(PyString_AS_STRING(utf8), (int)n, p);
Py_DECREF(utf8);
1991-06-04 16:42:30 -03:00
}
#endif
else if (PyTuple_Check(v)) {
w_byte(TYPE_TUPLE, p);
n = PyTuple_Size(v);
w_long((long)n, p);
1991-06-04 16:42:30 -03:00
for (i = 0; i < n; i++) {
w_object(PyTuple_GET_ITEM(v, i), p);
1991-06-04 16:42:30 -03:00
}
}
else if (PyList_Check(v)) {
w_byte(TYPE_LIST, p);
n = PyList_GET_SIZE(v);
w_long((long)n, p);
1991-06-04 16:42:30 -03:00
for (i = 0; i < n; i++) {
w_object(PyList_GET_ITEM(v, i), p);
1991-06-04 16:42:30 -03:00
}
}
else if (PyDict_Check(v)) {
2006-02-15 13:27:45 -04:00
Py_ssize_t pos;
PyObject *key, *value;
w_byte(TYPE_DICT, p);
1991-06-07 10:58:22 -03:00
/* This one is NULL object terminated! */
pos = 0;
while (PyDict_Next(v, &pos, &key, &value)) {
w_object(key, p);
w_object(value, p);
1991-06-07 10:58:22 -03:00
}
w_object((PyObject *)NULL, p);
1991-06-07 10:58:22 -03:00
}
else if (PyAnySet_Check(v)) {
PyObject *value, *it;
if (PyObject_TypeCheck(v, &PySet_Type))
w_byte(TYPE_SET, p);
else
w_byte(TYPE_FROZENSET, p);
n = PyObject_Size(v);
if (n == -1) {
p->depth--;
p->error = 1;
return;
}
w_long((long)n, p);
it = PyObject_GetIter(v);
if (it == NULL) {
p->depth--;
p->error = 1;
return;
}
while ((value = PyIter_Next(it)) != NULL) {
w_object(value, p);
Py_DECREF(value);
}
Py_DECREF(it);
if (PyErr_Occurred()) {
p->depth--;
p->error = 1;
return;
}
}
else if (PyCode_Check(v)) {
PyCodeObject *co = (PyCodeObject *)v;
w_byte(TYPE_CODE, p);
w_long(co->co_argcount, p);
w_long(co->co_nlocals, p);
w_long(co->co_stacksize, p);
w_long(co->co_flags, p);
w_object(co->co_code, p);
w_object(co->co_consts, p);
w_object(co->co_names, p);
1995-07-18 11:51:37 -03:00
w_object(co->co_varnames, p);
PEP 227 implementation The majority of the changes are in the compiler. The mainloop changes primarily to implement the new opcodes and to pass a function's closure to eval_code2(). Frames and functions got new slots to hold the closure. Include/compile.h Add co_freevars and co_cellvars slots to code objects. Update PyCode_New() to take freevars and cellvars as arguments Include/funcobject.h Add func_closure slot to function objects. Add GetClosure()/SetClosure() functions (and corresponding macros) for getting at the closure. Include/frameobject.h PyFrame_New() now takes a closure. Include/opcode.h Add four new opcodes: MAKE_CLOSURE, LOAD_CLOSURE, LOAD_DEREF, STORE_DEREF. Remove comment about old requirement for opcodes to fit in 7 bits. compile.c Implement changes to code objects for co_freevars and co_cellvars. Modify symbol table to use st_cur_name (string object for the name of the current scope) and st_cur_children (list of nested blocks). Also define st_nested, which might more properly be called st_cur_nested. Add several DEF_XXX flags to track def-use information for free variables. New or modified functions of note: com_make_closure(struct compiling *, PyCodeObject *) Emit LOAD_CLOSURE opcodes as needed to pass cells for free variables into nested scope. com_addop_varname(struct compiling *, int, char *) Emits opcodes for LOAD_DEREF and STORE_DEREF. get_ref_type(struct compiling *, char *name) Return NAME_CLOSURE if ref type is FREE or CELL symtable_load_symbols(struct compiling *) Decides what variables are cell or free based on def-use info. Can now raise SyntaxError if nested scopes are mixed with exec or from blah import *. make_scope_info(PyObject *, PyObject *, int, int) Helper functions for symtable scope stack. symtable_update_free_vars(struct symtable *) After a code block has been analyzed, it must check each of its children for free variables that are not defined in the block. If a variable is free in a child and not defined in the parent, then it is defined by block the enclosing the current one or it is a global. This does the right logic. symtable_add_use() is now a macro for symtable_add_def() symtable_assign(struct symtable *, node *) Use goto instead of for (;;) Fixed bug in symtable where name of keyword argument in function call was treated as assignment in the scope of the call site. Ex: def f(): g(a=2) # a was considered a local of f ceval.c eval_code2() now take one more argument, a closure. Implement LOAD_CLOSURE, LOAD_DEREF, STORE_DEREF, MAKE_CLOSURE> Also: When name error occurs for global variable, report that the name was global in the error mesage. Objects/frameobject.c Initialize f_closure to be a tuple containing space for cellvars and freevars. f_closure is NULL if neither are present. Objects/funcobject.c Add support for func_closure. Python/import.c Change the magic number. Python/marshal.c Track changes to code objects.
2001-01-25 16:06:59 -04:00
w_object(co->co_freevars, p);
w_object(co->co_cellvars, p);
w_object(co->co_filename, p);
w_object(co->co_name, p);
w_long(co->co_firstlineno, p);
w_object(co->co_lnotab, p);
1991-06-04 16:42:30 -03:00
}
2001-11-09 18:02:48 -04:00
else if (PyObject_CheckReadBuffer(v)) {
/* Write unknown buffer-style objects as a string */
char *s;
2001-11-09 18:02:48 -04:00
PyBufferProcs *pb = v->ob_type->tp_as_buffer;
w_byte(TYPE_STRING, p);
n = (*pb->bf_getreadbuffer)(v, 0, (void **)&s);
if (n > INT_MAX) {
p->depth--;
p->error = 1;
return;
}
w_long((long)n, p);
w_string(s, (int)n, p);
}
1991-06-04 16:42:30 -03:00
else {
w_byte(TYPE_UNKNOWN, p);
p->error = 1;
1991-06-04 16:42:30 -03:00
}
exit:
p->depth--;
1991-06-04 16:42:30 -03:00
}
/* version currently has no effect for writing longs. */
void
PyMarshal_WriteLongToFile(long x, FILE *fp, int version)
{
WFILE wf;
wf.fp = fp;
wf.error = 0;
wf.depth = 0;
wf.strings = NULL;
wf.version = version;
w_long(x, &wf);
}
1991-06-04 16:42:30 -03:00
void
PyMarshal_WriteObjectToFile(PyObject *x, FILE *fp, int version)
{
WFILE wf;
wf.fp = fp;
wf.error = 0;
wf.depth = 0;
wf.strings = (version > 0) ? PyDict_New() : NULL;
wf.version = version;
w_object(x, &wf);
Py_XDECREF(wf.strings);
}
typedef WFILE RFILE; /* Same struct with different invariants */
#define rs_byte(p) (((p)->ptr < (p)->end) ? (unsigned char)*(p)->ptr++ : EOF)
#define r_byte(p) ((p)->fp ? getc((p)->fp) : rs_byte(p))
static int
r_string(char *s, int n, RFILE *p)
{
if (p->fp != NULL)
2006-02-15 13:27:45 -04:00
/* The result fits into int because it must be <=n. */
return (int)fread(s, 1, n, p->fp);
if (p->end - p->ptr < n)
2006-02-15 13:27:45 -04:00
n = (int)(p->end - p->ptr);
memcpy(s, p->ptr, n);
p->ptr += n;
return n;
}
static int
r_short(RFILE *p)
1991-06-04 16:42:30 -03:00
{
register short x;
x = r_byte(p);
x |= r_byte(p) << 8;
/* Sign-extension, in case short greater than 16 bits */
x |= -(x & 0x8000);
1991-06-04 16:42:30 -03:00
return x;
}
static long
r_long(RFILE *p)
1991-06-04 16:42:30 -03:00
{
register long x;
register FILE *fp = p->fp;
if (fp) {
x = getc(fp);
x |= (long)getc(fp) << 8;
x |= (long)getc(fp) << 16;
x |= (long)getc(fp) << 24;
}
else {
x = rs_byte(p);
x |= (long)rs_byte(p) << 8;
x |= (long)rs_byte(p) << 16;
x |= (long)rs_byte(p) << 24;
}
#if SIZEOF_LONG > 4
/* Sign extension for 64-bit machines */
x |= -(x & 0x80000000L);
#endif
return x;
}
/* r_long64 deals with the TYPE_INT64 code. On a machine with
sizeof(long) > 4, it returns a Python int object, else a Python long
object. Note that w_long64 writes out TYPE_INT if 32 bits is enough,
so there's no inefficiency here in returning a PyLong on 32-bit boxes
for everything written via TYPE_INT64 (i.e., if an int is written via
TYPE_INT64, it *needs* more than 32 bits).
*/
static PyObject *
r_long64(RFILE *p)
{
long lo4 = r_long(p);
long hi4 = r_long(p);
#if SIZEOF_LONG > 4
long x = (hi4 << 32) | (lo4 & 0xFFFFFFFFL);
return PyInt_FromLong(x);
#else
unsigned char buf[8];
int one = 1;
int is_little_endian = (int)*(char*)&one;
if (is_little_endian) {
memcpy(buf, &lo4, 4);
memcpy(buf+4, &hi4, 4);
}
else {
memcpy(buf, &hi4, 4);
memcpy(buf+4, &lo4, 4);
}
return _PyLong_FromByteArray(buf, 8, is_little_endian, 1);
#endif
1991-06-04 16:42:30 -03:00
}
static PyObject *
r_object(RFILE *p)
1991-06-04 16:42:30 -03:00
{
2004-03-26 11:09:27 -04:00
/* NULL is a valid return value, it does not necessarily means that
an exception is set. */
PyObject *v, *v2, *v3;
1991-06-04 16:42:30 -03:00
long i, n;
int type = r_byte(p);
PyObject *retval;
p->depth++;
if (p->depth > MAX_MARSHAL_STACK_DEPTH) {
p->depth--;
PyErr_SetString(PyExc_ValueError, "recursion limit exceeded");
return NULL;
}
1991-06-04 16:42:30 -03:00
switch (type) {
1991-06-04 16:42:30 -03:00
case EOF:
PyErr_SetString(PyExc_EOFError,
"EOF read where object expected");
retval = NULL;
break;
1991-06-04 16:42:30 -03:00
case TYPE_NULL:
retval = NULL;
break;
1991-06-04 16:42:30 -03:00
case TYPE_NONE:
Py_INCREF(Py_None);
retval = Py_None;
break;
case TYPE_STOPITER:
Py_INCREF(PyExc_StopIteration);
retval = PyExc_StopIteration;
break;
case TYPE_ELLIPSIS:
Py_INCREF(Py_Ellipsis);
retval = Py_Ellipsis;
break;
case TYPE_FALSE:
Py_INCREF(Py_False);
retval = Py_False;
break;
case TYPE_TRUE:
Py_INCREF(Py_True);
retval = Py_True;
break;
1991-06-04 16:42:30 -03:00
case TYPE_INT:
retval = PyInt_FromLong(r_long(p));
break;
case TYPE_INT64:
retval = r_long64(p);
break;
1991-06-04 16:42:30 -03:00
case TYPE_LONG:
{
int size;
PyLongObject *ob;
n = r_long(p);
if (n < -INT_MAX || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError,
"bad marshal data");
retval = NULL;
break;
}
1991-06-04 16:42:30 -03:00
size = n<0 ? -n : n;
ob = _PyLong_New(size);
if (ob == NULL) {
retval = NULL;
break;
}
1991-06-04 16:42:30 -03:00
ob->ob_size = n;
2004-03-26 11:09:27 -04:00
for (i = 0; i < size; i++) {
int digit = r_short(p);
if (digit < 0) {
Py_DECREF(ob);
PyErr_SetString(PyExc_ValueError,
"bad marshal data");
ob = NULL;
break;
2004-03-26 11:09:27 -04:00
}
if (ob != NULL)
ob->ob_digit[i] = digit;
2004-03-26 11:09:27 -04:00
}
retval = (PyObject *)ob;
break;
1991-06-04 16:42:30 -03:00
}
1991-06-04 16:42:30 -03:00
case TYPE_FLOAT:
{
char buf[256];
double dx;
n = r_byte(p);
2004-03-26 11:09:27 -04:00
if (n == EOF || r_string(buf, (int)n, p) != n) {
PyErr_SetString(PyExc_EOFError,
1991-06-04 16:42:30 -03:00
"EOF read where object expected");
retval = NULL;
break;
1991-06-04 16:42:30 -03:00
}
buf[n] = '\0';
retval = NULL;
PyFPE_START_PROTECT("atof", break)
dx = PyOS_ascii_atof(buf);
1997-03-14 00:32:50 -04:00
PyFPE_END_PROTECT(dx)
retval = PyFloat_FromDouble(dx);
break;
1991-06-04 16:42:30 -03:00
}
case TYPE_BINARY_FLOAT:
{
unsigned char buf[8];
double x;
if (r_string((char*)buf, 8, p) != 8) {
PyErr_SetString(PyExc_EOFError,
"EOF read where object expected");
retval = NULL;
break;
}
x = _PyFloat_Unpack8(buf, 1);
if (x == -1.0 && PyErr_Occurred()) {
retval = NULL;
break;
}
retval = PyFloat_FromDouble(x);
break;
}
1996-01-11 21:09:56 -04:00
#ifndef WITHOUT_COMPLEX
case TYPE_COMPLEX:
{
char buf[256];
Py_complex c;
1996-01-11 21:09:56 -04:00
n = r_byte(p);
2004-03-26 11:09:27 -04:00
if (n == EOF || r_string(buf, (int)n, p) != n) {
PyErr_SetString(PyExc_EOFError,
1996-01-11 21:09:56 -04:00
"EOF read where object expected");
retval = NULL;
break;
1996-01-11 21:09:56 -04:00
}
buf[n] = '\0';
retval = NULL;
PyFPE_START_PROTECT("atof", break;)
c.real = PyOS_ascii_atof(buf);
1997-03-14 00:32:50 -04:00
PyFPE_END_PROTECT(c)
1996-01-11 21:09:56 -04:00
n = r_byte(p);
2004-03-26 11:09:27 -04:00
if (n == EOF || r_string(buf, (int)n, p) != n) {
PyErr_SetString(PyExc_EOFError,
1996-01-11 21:09:56 -04:00
"EOF read where object expected");
retval = NULL;
break;
1996-01-11 21:09:56 -04:00
}
buf[n] = '\0';
PyFPE_START_PROTECT("atof", break)
c.imag = PyOS_ascii_atof(buf);
1997-03-14 00:32:50 -04:00
PyFPE_END_PROTECT(c)
retval = PyComplex_FromCComplex(c);
break;
1996-01-11 21:09:56 -04:00
}
case TYPE_BINARY_COMPLEX:
{
unsigned char buf[8];
Py_complex c;
if (r_string((char*)buf, 8, p) != 8) {
PyErr_SetString(PyExc_EOFError,
"EOF read where object expected");
retval = NULL;
break;
}
c.real = _PyFloat_Unpack8(buf, 1);
if (c.real == -1.0 && PyErr_Occurred()) {
retval = NULL;
break;
}
if (r_string((char*)buf, 8, p) != 8) {
PyErr_SetString(PyExc_EOFError,
"EOF read where object expected");
retval = NULL;
break;
}
c.imag = _PyFloat_Unpack8(buf, 1);
if (c.imag == -1.0 && PyErr_Occurred()) {
retval = NULL;
break;
}
retval = PyComplex_FromCComplex(c);
break;
}
1996-01-11 21:09:56 -04:00
#endif
case TYPE_INTERNED:
1991-06-04 16:42:30 -03:00
case TYPE_STRING:
n = r_long(p);
if (n < 0 || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
v = PyString_FromStringAndSize((char *)NULL, n);
if (v == NULL) {
retval = NULL;
break;
}
if (r_string(PyString_AS_STRING(v), (int)n, p) != n) {
Py_DECREF(v);
PyErr_SetString(PyExc_EOFError,
1991-06-04 16:42:30 -03:00
"EOF read where object expected");
retval = NULL;
break;
1991-06-04 16:42:30 -03:00
}
if (type == TYPE_INTERNED) {
PyString_InternInPlace(&v);
if (PyList_Append(p->strings, v) < 0) {
retval = NULL;
break;
}
}
retval = v;
break;
case TYPE_STRINGREF:
n = r_long(p);
if (n < 0 || n >= PyList_GET_SIZE(p->strings)) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
v = PyList_GET_ITEM(p->strings, n);
Py_INCREF(v);
retval = v;
break;
#ifdef Py_USING_UNICODE
case TYPE_UNICODE:
{
char *buffer;
n = r_long(p);
if (n < 0 || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
buffer = PyMem_NEW(char, n);
if (buffer == NULL) {
retval = PyErr_NoMemory();
break;
}
if (r_string(buffer, (int)n, p) != n) {
PyMem_DEL(buffer);
PyErr_SetString(PyExc_EOFError,
"EOF read where object expected");
retval = NULL;
break;
}
v = PyUnicode_DecodeUTF8(buffer, n, NULL);
PyMem_DEL(buffer);
retval = v;
break;
}
#endif
1991-06-04 16:42:30 -03:00
case TYPE_TUPLE:
n = r_long(p);
if (n < 0 || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
v = PyTuple_New((int)n);
if (v == NULL) {
retval = NULL;
break;
}
for (i = 0; i < n; i++) {
v2 = r_object(p);
if ( v2 == NULL ) {
2004-03-26 11:09:27 -04:00
if (!PyErr_Occurred())
PyErr_SetString(PyExc_TypeError,
"NULL object in marshal data");
Py_DECREF(v);
v = NULL;
break;
}
PyTuple_SET_ITEM(v, (int)i, v2);
}
retval = v;
break;
1991-06-04 16:42:30 -03:00
case TYPE_LIST:
n = r_long(p);
if (n < 0 || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
v = PyList_New((int)n);
if (v == NULL) {
retval = NULL;
break;
}
for (i = 0; i < n; i++) {
v2 = r_object(p);
if ( v2 == NULL ) {
2004-03-26 11:09:27 -04:00
if (!PyErr_Occurred())
PyErr_SetString(PyExc_TypeError,
"NULL object in marshal data");
Py_DECREF(v);
v = NULL;
break;
}
PyList_SET_ITEM(v, (int)i, v2);
}
retval = v;
break;
1991-06-07 10:58:22 -03:00
case TYPE_DICT:
v = PyDict_New();
if (v == NULL) {
retval = NULL;
break;
}
1991-06-07 10:58:22 -03:00
for (;;) {
PyObject *key, *val;
key = r_object(p);
1991-06-07 10:58:22 -03:00
if (key == NULL)
2004-03-26 11:09:27 -04:00
break;
val = r_object(p);
if (val != NULL)
PyDict_SetItem(v, key, val);
Py_DECREF(key);
Py_XDECREF(val);
1991-06-07 10:58:22 -03:00
}
2004-03-26 11:09:27 -04:00
if (PyErr_Occurred()) {
Py_DECREF(v);
v = NULL;
}
retval = v;
break;
case TYPE_SET:
case TYPE_FROZENSET:
n = r_long(p);
if (n < 0 || n > INT_MAX) {
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
}
v = PyTuple_New((int)n);
if (v == NULL) {
retval = NULL;
break;
}
for (i = 0; i < n; i++) {
v2 = r_object(p);
if ( v2 == NULL ) {
if (!PyErr_Occurred())
PyErr_SetString(PyExc_TypeError,
"NULL object in marshal data");
Py_DECREF(v);
v = NULL;
break;
}
PyTuple_SET_ITEM(v, (int)i, v2);
}
if (v == NULL) {
retval = NULL;
break;
}
if (type == TYPE_SET)
2005-08-16 00:47:52 -03:00
v3 = PySet_New(v);
else
2005-08-16 00:47:52 -03:00
v3 = PyFrozenSet_New(v);
Py_DECREF(v);
retval = v3;
break;
1991-06-04 16:42:30 -03:00
case TYPE_CODE:
if (PyEval_GetRestricted()) {
PyErr_SetString(PyExc_RuntimeError,
"cannot unmarshal code objects in "
"restricted execution mode");
retval = NULL;
break;
}
else {
int argcount;
int nlocals;
int stacksize;
int flags;
PyObject *code = NULL;
PyObject *consts = NULL;
PyObject *names = NULL;
PyObject *varnames = NULL;
PyObject *freevars = NULL;
PyObject *cellvars = NULL;
PyObject *filename = NULL;
PyObject *name = NULL;
int firstlineno;
PyObject *lnotab = NULL;
v = NULL;
/* XXX ignore long->int overflows for now */
argcount = (int)r_long(p);
nlocals = (int)r_long(p);
stacksize = (int)r_long(p);
flags = (int)r_long(p);
code = r_object(p);
if (code == NULL)
goto code_error;
consts = r_object(p);
if (consts == NULL)
goto code_error;
names = r_object(p);
if (names == NULL)
goto code_error;
varnames = r_object(p);
if (varnames == NULL)
goto code_error;
freevars = r_object(p);
if (freevars == NULL)
goto code_error;
cellvars = r_object(p);
if (cellvars == NULL)
goto code_error;
filename = r_object(p);
if (filename == NULL)
goto code_error;
name = r_object(p);
if (name == NULL)
goto code_error;
firstlineno = (int)r_long(p);
lnotab = r_object(p);
if (lnotab == NULL)
goto code_error;
v = (PyObject *) PyCode_New(
argcount, nlocals, stacksize, flags,
1995-07-18 11:51:37 -03:00
code, consts, names, varnames,
freevars, cellvars, filename, name,
firstlineno, lnotab);
code_error:
Py_XDECREF(code);
Py_XDECREF(consts);
Py_XDECREF(names);
Py_XDECREF(varnames);
PEP 227 implementation The majority of the changes are in the compiler. The mainloop changes primarily to implement the new opcodes and to pass a function's closure to eval_code2(). Frames and functions got new slots to hold the closure. Include/compile.h Add co_freevars and co_cellvars slots to code objects. Update PyCode_New() to take freevars and cellvars as arguments Include/funcobject.h Add func_closure slot to function objects. Add GetClosure()/SetClosure() functions (and corresponding macros) for getting at the closure. Include/frameobject.h PyFrame_New() now takes a closure. Include/opcode.h Add four new opcodes: MAKE_CLOSURE, LOAD_CLOSURE, LOAD_DEREF, STORE_DEREF. Remove comment about old requirement for opcodes to fit in 7 bits. compile.c Implement changes to code objects for co_freevars and co_cellvars. Modify symbol table to use st_cur_name (string object for the name of the current scope) and st_cur_children (list of nested blocks). Also define st_nested, which might more properly be called st_cur_nested. Add several DEF_XXX flags to track def-use information for free variables. New or modified functions of note: com_make_closure(struct compiling *, PyCodeObject *) Emit LOAD_CLOSURE opcodes as needed to pass cells for free variables into nested scope. com_addop_varname(struct compiling *, int, char *) Emits opcodes for LOAD_DEREF and STORE_DEREF. get_ref_type(struct compiling *, char *name) Return NAME_CLOSURE if ref type is FREE or CELL symtable_load_symbols(struct compiling *) Decides what variables are cell or free based on def-use info. Can now raise SyntaxError if nested scopes are mixed with exec or from blah import *. make_scope_info(PyObject *, PyObject *, int, int) Helper functions for symtable scope stack. symtable_update_free_vars(struct symtable *) After a code block has been analyzed, it must check each of its children for free variables that are not defined in the block. If a variable is free in a child and not defined in the parent, then it is defined by block the enclosing the current one or it is a global. This does the right logic. symtable_add_use() is now a macro for symtable_add_def() symtable_assign(struct symtable *, node *) Use goto instead of for (;;) Fixed bug in symtable where name of keyword argument in function call was treated as assignment in the scope of the call site. Ex: def f(): g(a=2) # a was considered a local of f ceval.c eval_code2() now take one more argument, a closure. Implement LOAD_CLOSURE, LOAD_DEREF, STORE_DEREF, MAKE_CLOSURE> Also: When name error occurs for global variable, report that the name was global in the error mesage. Objects/frameobject.c Initialize f_closure to be a tuple containing space for cellvars and freevars. f_closure is NULL if neither are present. Objects/funcobject.c Add support for func_closure. Python/import.c Change the magic number. Python/marshal.c Track changes to code objects.
2001-01-25 16:06:59 -04:00
Py_XDECREF(freevars);
Py_XDECREF(cellvars);
Py_XDECREF(filename);
Py_XDECREF(name);
Py_XDECREF(lnotab);
1991-06-04 16:42:30 -03:00
}
retval = v;
break;
1991-06-04 16:42:30 -03:00
default:
/* Bogus data got written, which isn't ideal.
This will let you keep working and recover. */
PyErr_SetString(PyExc_ValueError, "bad marshal data");
retval = NULL;
break;
1991-06-04 16:42:30 -03:00
}
p->depth--;
return retval;
1991-06-04 16:42:30 -03:00
}
2004-06-13 17:31:49 -03:00
static PyObject *
2004-03-26 11:09:27 -04:00
read_object(RFILE *p)
{
PyObject *v;
if (PyErr_Occurred()) {
fprintf(stderr, "XXX readobject called with exception set\n");
return NULL;
}
v = r_object(p);
if (v == NULL && !PyErr_Occurred())
PyErr_SetString(PyExc_TypeError, "NULL object in marshal data");
return v;
}
int
PyMarshal_ReadShortFromFile(FILE *fp)
{
RFILE rf;
assert(fp);
rf.fp = fp;
rf.strings = NULL;
rf.end = rf.ptr = NULL;
return r_short(&rf);
}
long
PyMarshal_ReadLongFromFile(FILE *fp)
{
RFILE rf;
rf.fp = fp;
rf.strings = NULL;
rf.ptr = rf.end = NULL;
return r_long(&rf);
}
#ifdef HAVE_FSTAT
/* Return size of file in bytes; < 0 if unknown. */
static off_t
getfilesize(FILE *fp)
{
struct stat st;
if (fstat(fileno(fp), &st) != 0)
return -1;
else
return st.st_size;
}
#endif
/* If we can get the size of the file up-front, and it's reasonably small,
* read it in one gulp and delegate to ...FromString() instead. Much quicker
* than reading a byte at a time from file; speeds .pyc imports.
* CAUTION: since this may read the entire remainder of the file, don't
* call it unless you know you're done with the file.
*/
PyObject *
PyMarshal_ReadLastObjectFromFile(FILE *fp)
{
/* 75% of 2.1's .pyc files can exploit SMALL_FILE_LIMIT.
* REASONABLE_FILE_LIMIT is by defn something big enough for Tkinter.pyc.
*/
#define SMALL_FILE_LIMIT (1L << 14)
#define REASONABLE_FILE_LIMIT (1L << 18)
#ifdef HAVE_FSTAT
off_t filesize;
#endif
#ifdef HAVE_FSTAT
filesize = getfilesize(fp);
if (filesize > 0) {
char buf[SMALL_FILE_LIMIT];
char* pBuf = NULL;
if (filesize <= SMALL_FILE_LIMIT)
pBuf = buf;
else if (filesize <= REASONABLE_FILE_LIMIT)
pBuf = (char *)PyMem_MALLOC(filesize);
if (pBuf != NULL) {
PyObject* v;
2006-02-15 13:27:45 -04:00
size_t n;
/* filesize must fit into an int, because it
is smaller than REASONABLE_FILE_LIMIT */
n = fread(pBuf, 1, (int)filesize, fp);
v = PyMarshal_ReadObjectFromString(pBuf, n);
if (pBuf != buf)
PyMem_FREE(pBuf);
return v;
}
}
#endif
/* We don't have fstat, or we do but the file is larger than
* REASONABLE_FILE_LIMIT or malloc failed -- read a byte at a time.
*/
return PyMarshal_ReadObjectFromFile(fp);
#undef SMALL_FILE_LIMIT
#undef REASONABLE_FILE_LIMIT
}
PyObject *
PyMarshal_ReadObjectFromFile(FILE *fp)
{
RFILE rf;
PyObject *result;
rf.fp = fp;
rf.strings = PyList_New(0);
rf.depth = 0;
rf.ptr = rf.end = NULL;
result = r_object(&rf);
Py_DECREF(rf.strings);
return result;
}
PyObject *
2006-02-15 13:27:45 -04:00
PyMarshal_ReadObjectFromString(char *str, Py_ssize_t len)
{
RFILE rf;
PyObject *result;
rf.fp = NULL;
rf.ptr = str;
rf.end = str + len;
rf.strings = PyList_New(0);
rf.depth = 0;
result = r_object(&rf);
Py_DECREF(rf.strings);
return result;
}
PyObject *
PyMarshal_WriteObjectToString(PyObject *x, int version)
{
WFILE wf;
wf.fp = NULL;
wf.str = PyString_FromStringAndSize((char *)NULL, 50);
if (wf.str == NULL)
return NULL;
wf.ptr = PyString_AS_STRING((PyStringObject *)wf.str);
wf.end = wf.ptr + PyString_Size(wf.str);
wf.error = 0;
wf.depth = 0;
wf.version = version;
wf.strings = (version > 0) ? PyDict_New() : NULL;
w_object(x, &wf);
Py_XDECREF(wf.strings);
if (wf.str != NULL) {
char *base = PyString_AS_STRING((PyStringObject *)wf.str);
if (wf.ptr - base > PY_SSIZE_T_MAX) {
Py_DECREF(wf.str);
PyErr_SetString(PyExc_OverflowError,
"too much marshall data for a string");
return NULL;
}
_PyString_Resize(&wf.str, (Py_ssize_t)(wf.ptr - base));
}
if (wf.error) {
Py_XDECREF(wf.str);
PyErr_SetString(PyExc_ValueError,
(wf.error==1)?"unmarshallable object"
:"object too deeply nested to marshal");
return NULL;
}
return wf.str;
}
1991-06-07 10:58:22 -03:00
/* And an interface for Python programs... */
1991-06-04 16:42:30 -03:00
static PyObject *
marshal_dump(PyObject *self, PyObject *args)
1991-06-04 16:42:30 -03:00
{
WFILE wf;
PyObject *x;
PyObject *f;
int version = Py_MARSHAL_VERSION;
if (!PyArg_ParseTuple(args, "OO|i:dump", &x, &f, &version))
1991-06-04 16:42:30 -03:00
return NULL;
if (!PyFile_Check(f)) {
PyErr_SetString(PyExc_TypeError,
"marshal.dump() 2nd arg must be file");
1991-06-04 16:42:30 -03:00
return NULL;
}
wf.fp = PyFile_AsFile(f);
wf.str = NULL;
wf.ptr = wf.end = NULL;
wf.error = 0;
wf.depth = 0;
wf.strings = (version > 0) ? PyDict_New() : 0;
2005-11-16 01:04:51 -04:00
wf.version = version;
w_object(x, &wf);
Py_XDECREF(wf.strings);
if (wf.error) {
PyErr_SetString(PyExc_ValueError,
(wf.error==1)?"unmarshallable object"
:"object too deeply nested to marshal");
return NULL;
}
Py_INCREF(Py_None);
return Py_None;
1991-06-04 16:42:30 -03:00
}
static PyObject *
2006-05-29 18:58:42 -03:00
marshal_load(PyObject *self, PyObject *f)
1991-06-04 16:42:30 -03:00
{
RFILE rf;
2006-05-29 18:58:42 -03:00
PyObject *result;
if (!PyFile_Check(f)) {
PyErr_SetString(PyExc_TypeError,
"marshal.load() arg must be file");
1991-06-04 16:42:30 -03:00
return NULL;
}
rf.fp = PyFile_AsFile(f);
rf.strings = PyList_New(0);
rf.depth = 0;
result = read_object(&rf);
Py_DECREF(rf.strings);
return result;
}
static PyObject *
marshal_dumps(PyObject *self, PyObject *args)
{
PyObject *x;
int version = Py_MARSHAL_VERSION;
if (!PyArg_ParseTuple(args, "O|i:dumps", &x, &version))
return NULL;
return PyMarshal_WriteObjectToString(x, version);
}
static PyObject *
marshal_loads(PyObject *self, PyObject *args)
{
RFILE rf;
char *s;
2006-03-01 19:49:13 -04:00
Py_ssize_t n;
PyObject* result;
if (!PyArg_ParseTuple(args, "s#:loads", &s, &n))
return NULL;
rf.fp = NULL;
rf.ptr = s;
rf.end = s + n;
rf.strings = PyList_New(0);
rf.depth = 0;
result = read_object(&rf);
Py_DECREF(rf.strings);
return result;
1991-06-04 16:42:30 -03:00
}
static PyMethodDef marshal_methods[] = {
{"dump", marshal_dump, METH_VARARGS},
2006-05-29 18:58:42 -03:00
{"load", marshal_load, METH_O},
{"dumps", marshal_dumps, METH_VARARGS},
{"loads", marshal_loads, METH_VARARGS},
1991-06-04 16:42:30 -03:00
{NULL, NULL} /* sentinel */
};
PyMODINIT_FUNC
PyMarshal_Init(void)
1991-06-04 16:42:30 -03:00
{
PyObject *mod = Py_InitModule("marshal", marshal_methods);
if (mod == NULL)
return;
PyModule_AddIntConstant(mod, "version", Py_MARSHAL_VERSION);
1991-06-04 16:42:30 -03:00
}