More coding by random modification.
Encoding now return bytes instead of str8.
eval(), exec(), compile() now accept unicode or bytes.
diff --git a/Python/ast.c b/Python/ast.c
index 821a5ad..5845076 100644
--- a/Python/ast.c
+++ b/Python/ast.c
@@ -3101,8 +3101,9 @@
Py_DECREF(u);
return NULL;
}
- r = PyString_AsString(w);
- rn = PyString_Size(w);
+ assert(PyBytes_Check(w));
+ r = PyBytes_AsString(w);
+ rn = PyBytes_Size(w);
assert(rn % 2 == 0);
for (i = 0; i < rn; i += 2) {
sprintf(p, "\\u%02x%02x",
diff --git a/Python/bltinmodule.c b/Python/bltinmodule.c
index 828cb5d..7d52e2b 100644
--- a/Python/bltinmodule.c
+++ b/Python/bltinmodule.c
@@ -412,6 +412,36 @@
\n\
Return negative if x<y, zero if x==y, positive if x>y.");
+
+static char *
+source_as_string(PyObject *cmd)
+{
+ char *str;
+ Py_ssize_t size;
+
+ if (!PyObject_CheckReadBuffer(cmd) &&
+ !PyUnicode_Check(cmd)) {
+ PyErr_SetString(PyExc_TypeError,
+ "eval()/exec() arg 1 must be a string, bytes or code object");
+ return NULL;
+ }
+
+ if (PyUnicode_Check(cmd)) {
+ cmd = _PyUnicode_AsDefaultEncodedString(cmd, NULL);
+ if (cmd == NULL)
+ return NULL;
+ }
+ if (PyObject_AsReadBuffer(cmd, (const void **)&str, &size) < 0) {
+ return NULL;
+ }
+ if (strlen(str) != size) {
+ PyErr_SetString(PyExc_TypeError,
+ "source code string cannot contain null bytes");
+ return NULL;
+ }
+ return str;
+}
+
static PyObject *
builtin_compile(PyObject *self, PyObject *args, PyObject *kwds)
{
@@ -422,8 +452,7 @@
int dont_inherit = 0;
int supplied_flags = 0;
PyCompilerFlags cf;
- PyObject *result = NULL, *cmd, *tmp = NULL;
- Py_ssize_t length;
+ PyObject *cmd;
static char *kwlist[] = {"source", "filename", "mode", "flags",
"dont_inherit", NULL};
@@ -432,22 +461,11 @@
&supplied_flags, &dont_inherit))
return NULL;
- cf.cf_flags = supplied_flags;
+ cf.cf_flags = supplied_flags | PyCF_SOURCE_IS_UTF8;
- if (PyUnicode_Check(cmd)) {
- tmp = PyUnicode_AsUTF8String(cmd);
- if (tmp == NULL)
- return NULL;
- cmd = tmp;
- cf.cf_flags |= PyCF_SOURCE_IS_UTF8;
- }
- if (PyObject_AsReadBuffer(cmd, (const void **)&str, &length))
+ str = source_as_string(cmd);
+ if (str == NULL)
return NULL;
- if ((size_t)length != strlen(str)) {
- PyErr_SetString(PyExc_TypeError,
- "compile() expected string without null bytes");
- goto cleanup;
- }
if (strcmp(startstr, "exec") == 0)
start = Py_file_input;
@@ -458,7 +476,7 @@
else {
PyErr_SetString(PyExc_ValueError,
"compile() arg 3 must be 'exec' or 'eval' or 'single'");
- goto cleanup;
+ return NULL;
}
if (supplied_flags &
@@ -466,17 +484,14 @@
{
PyErr_SetString(PyExc_ValueError,
"compile(): unrecognised flags");
- goto cleanup;
+ return NULL;
}
/* XXX Warn if (supplied_flags & PyCF_MASK_OBSOLETE) != 0? */
if (!dont_inherit) {
PyEval_MergeCompilerFlags(&cf);
}
- result = Py_CompileStringFlags(str, filename, start, &cf);
-cleanup:
- Py_XDECREF(tmp);
- return result;
+ return Py_CompileStringFlags(str, filename, start, &cf);
}
PyDoc_STRVAR(compile_doc,
@@ -584,28 +599,14 @@
return PyEval_EvalCode((PyCodeObject *) cmd, globals, locals);
}
- if (!PyString_Check(cmd) &&
- !PyUnicode_Check(cmd)) {
- PyErr_SetString(PyExc_TypeError,
- "eval() arg 1 must be a string or code object");
+ str = source_as_string(cmd);
+ if (str == NULL)
return NULL;
- }
- cf.cf_flags = 0;
- if (PyUnicode_Check(cmd)) {
- tmp = PyUnicode_AsUTF8String(cmd);
- if (tmp == NULL)
- return NULL;
- cmd = tmp;
- cf.cf_flags |= PyCF_SOURCE_IS_UTF8;
- }
- if (PyString_AsStringAndSize(cmd, &str, NULL)) {
- Py_XDECREF(tmp);
- return NULL;
- }
while (*str == ' ' || *str == '\t')
str++;
+ cf.cf_flags = PyCF_SOURCE_IS_UTF8;
(void)PyEval_MergeCompilerFlags(&cf);
result = PyRun_StringFlags(str, Py_eval_input, globals, locals, &cf);
Py_XDECREF(tmp);
@@ -694,25 +695,16 @@
locals);
}
else {
- PyObject *tmp = NULL;
- char *str;
+ char *str = source_as_string(prog);
PyCompilerFlags cf;
- cf.cf_flags = 0;
- if (PyUnicode_Check(prog)) {
- tmp = PyUnicode_AsUTF8String(prog);
- if (tmp == NULL)
- return NULL;
- prog = tmp;
- cf.cf_flags |= PyCF_SOURCE_IS_UTF8;
- }
- if (PyString_AsStringAndSize(prog, &str, NULL))
+ if (str == NULL)
return NULL;
+ cf.cf_flags = PyCF_SOURCE_IS_UTF8;
if (PyEval_MergeCompilerFlags(&cf))
v = PyRun_StringFlags(str, Py_file_input, globals,
locals, &cf);
else
v = PyRun_String(str, Py_file_input, globals, locals);
- Py_XDECREF(tmp);
}
if (v == NULL)
return NULL;
diff --git a/Python/getargs.c b/Python/getargs.c
index 2a02a89..7d55bae 100644
--- a/Python/getargs.c
+++ b/Python/getargs.c
@@ -883,7 +883,9 @@
char **buffer;
const char *encoding;
PyObject *s;
- int size, recode_strings;
+ int recode_strings;
+ Py_ssize_t size;
+ char *ptr;
/* Get 'e' parameter: the encoding name */
encoding = (const char *)va_arg(*p_va, const char *);
@@ -912,6 +914,8 @@
if (!recode_strings && PyString_Check(arg)) {
s = arg;
Py_INCREF(s);
+ size = PyString_GET_SIZE(s);
+ ptr = PyString_AS_STRING(s);
}
else {
PyObject *u;
@@ -931,14 +935,15 @@
if (s == NULL)
return converterr("(encoding failed)",
arg, msgbuf, bufsize);
- if (!PyString_Check(s)) {
+ if (!PyBytes_Check(s)) {
Py_DECREF(s);
return converterr(
- "(encoder failed to return a string)",
+ "(encoder failed to return bytes)",
arg, msgbuf, bufsize);
}
+ size = PyBytes_GET_SIZE(s);
+ ptr = PyBytes_AS_STRING(s);
}
- size = PyString_GET_SIZE(s);
/* Write output; output is guaranteed to be 0-terminated */
if (*format == '#') {
@@ -994,9 +999,7 @@
arg, msgbuf, bufsize);
}
}
- memcpy(*buffer,
- PyString_AS_STRING(s),
- size + 1);
+ memcpy(*buffer, ptr, size+1);
STORE_SIZE(size);
} else {
/* Using a 0-terminated buffer:
@@ -1012,8 +1015,7 @@
PyMem_Free()ing it after usage
*/
- if ((Py_ssize_t)strlen(PyString_AS_STRING(s))
- != size) {
+ if ((Py_ssize_t)strlen(ptr) != size) {
Py_DECREF(s);
return converterr(
"(encoded string without NULL bytes)",
@@ -1030,9 +1032,7 @@
return converterr("(cleanup problem)",
arg, msgbuf, bufsize);
}
- memcpy(*buffer,
- PyString_AS_STRING(s),
- size + 1);
+ memcpy(*buffer, ptr, size+1);
}
Py_DECREF(s);
break;
diff --git a/Python/import.c b/Python/import.c
index 7e3d2f4..2e1f894 100644
--- a/Python/import.c
+++ b/Python/import.c
@@ -1254,6 +1254,9 @@
for (i = 0; i < npath; i++) {
PyObject *copy = NULL;
PyObject *v = PyList_GetItem(path, i);
+ PyObject *origv = v;
+ char *base;
+ Py_ssize_t size;
if (!v)
return NULL;
if (PyUnicode_Check(v)) {
@@ -1263,15 +1266,24 @@
return NULL;
v = copy;
}
- else
- if (!PyString_Check(v))
+ if (PyString_Check(v)) {
+ base = PyString_AS_STRING(v);
+ size = PyString_GET_SIZE(v);
+ }
+ else if (PyBytes_Check(v)) {
+ base = PyBytes_AS_STRING(v);
+ size = PyBytes_GET_SIZE(v);
+ }
+ else {
+ Py_XDECREF(copy);
continue;
- len = PyString_GET_SIZE(v);
+ }
+ len = size;
if (len + 2 + namelen + MAXSUFFIXSIZE >= buflen) {
Py_XDECREF(copy);
continue; /* Too long */
}
- strcpy(buf, PyString_AS_STRING(v));
+ strcpy(buf, base);
if (strlen(buf) != len) {
Py_XDECREF(copy);
continue; /* v contains '\0' */
@@ -1282,7 +1294,7 @@
PyObject *importer;
importer = get_path_importer(path_importer_cache,
- path_hooks, v);
+ path_hooks, origv);
if (importer == NULL) {
Py_XDECREF(copy);
return NULL;
diff --git a/Python/marshal.c b/Python/marshal.c
index 94d73a0..9243798 100644
--- a/Python/marshal.c
+++ b/Python/marshal.c
@@ -263,14 +263,14 @@
return;
}
w_byte(TYPE_UNICODE, p);
- n = PyString_GET_SIZE(utf8);
+ n = PyBytes_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);
+ w_string(PyBytes_AS_STRING(utf8), (int)n, p);
Py_DECREF(utf8);
}
else if (PyTuple_Check(v)) {
@@ -1031,7 +1031,7 @@
if (wf.ptr - base > PY_SSIZE_T_MAX) {
Py_DECREF(wf.str);
PyErr_SetString(PyExc_OverflowError,
- "too much marshall data for a string");
+ "too much marshal data for a string");
return NULL;
}
_PyString_Resize(&wf.str, (Py_ssize_t)(wf.ptr - base));