Issue #10557: Fixed error messages from float() and other numeric

types.  Added a new API function, PyUnicode_TransformDecimalToASCII(),
which transforms non-ASCII decimal digits in a Unicode string to their
ASCII equivalents.

4 files changed, 106 insertions, 48 deletions

diff --git a/Objects/complexobject.c b/Objects/complexobject.c
index 5999796..ec529d5 100644
--- a/Objects/complexobject.c
+++ b/Objects/complexobject.c
@@ -766,20 +766,26 @@ complex_subtype_from_string(PyTypeObject *type, PyObject *v)
     char *end;
     double x=0.0, y=0.0, z;
     int got_bracket=0;
-    char *s_buffer = NULL;
+    PyObject *s_buffer = NULL;
     Py_ssize_t len;
 
     if (PyUnicode_Check(v)) {
-        s_buffer = (char *)PyMem_MALLOC(PyUnicode_GET_SIZE(v) + 1);
+        Py_ssize_t i, buflen = PyUnicode_GET_SIZE(v);
+        Py_UNICODE *bufptr;
+        s_buffer = PyUnicode_TransformDecimalToASCII(
+            PyUnicode_AS_UNICODE(v), buflen);
         if (s_buffer == NULL)
-            return PyErr_NoMemory();
-        if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v),
-                                    PyUnicode_GET_SIZE(v),
-                                    s_buffer,
-                                    NULL))
+            return NULL;
+        /* Replace non-ASCII whitespace with ' ' */
+        bufptr = PyUnicode_AS_UNICODE(s_buffer);
+        for (i = 0; i < buflen; i++) {
+            Py_UNICODE ch = bufptr[i];
+            if (ch > 127 && Py_UNICODE_ISSPACE(ch))
+                bufptr[i] = ' ';
+        }
+        s = _PyUnicode_AsStringAndSize(s_buffer, &len);
+        if (s == NULL)
             goto error;
-        s = s_buffer;
-        len = strlen(s);
     }
     else if (PyObject_AsCharBuffer(v, &s, &len)) {
         PyErr_SetString(PyExc_TypeError,
@@ -894,16 +900,14 @@ complex_subtype_from_string(PyTypeObject *type, PyObject *v)
     if (s-start != len)
         goto parse_error;
 
-    if (s_buffer)
-        PyMem_FREE(s_buffer);
+    Py_XDECREF(s_buffer);
     return complex_subtype_from_doubles(type, x, y);
 
   parse_error:
     PyErr_SetString(PyExc_ValueError,
                     "complex() arg is a malformed string");
   error:
-    if (s_buffer)
-        PyMem_FREE(s_buffer);
+    Py_XDECREF(s_buffer);
     return NULL;
 }
 
diff --git a/Objects/floatobject.c b/Objects/floatobject.c
index 4decb0b..8409f0a 100644
--- a/Objects/floatobject.c
+++ b/Objects/floatobject.c
@@ -174,22 +174,30 @@ PyFloat_FromString(PyObject *v)
 {
     const char *s, *last, *end;
     double x;
-    char buffer[256]; /* for errors */
-    char *s_buffer = NULL;
+    PyObject *s_buffer = NULL;
     Py_ssize_t len;
     PyObject *result = NULL;
 
     if (PyUnicode_Check(v)) {
-        s_buffer = (char *)PyMem_MALLOC(PyUnicode_GET_SIZE(v)+1);
+        Py_ssize_t i, buflen = PyUnicode_GET_SIZE(v);
+        Py_UNICODE *bufptr;
+        s_buffer = PyUnicode_TransformDecimalToASCII(
+            PyUnicode_AS_UNICODE(v), buflen);
         if (s_buffer == NULL)
-            return PyErr_NoMemory();
-        if (PyUnicode_EncodeDecimal(PyUnicode_AS_UNICODE(v),
-                                    PyUnicode_GET_SIZE(v),
-                                    s_buffer,
-                                    NULL))
-            goto error;
-        s = s_buffer;
-        len = strlen(s);
+            return NULL;
+        /* Replace non-ASCII whitespace with ' ' */
+        bufptr = PyUnicode_AS_UNICODE(s_buffer);
+        for (i = 0; i < buflen; i++) {
+            Py_UNICODE ch = bufptr[i];
+            if (ch > 127 && Py_UNICODE_ISSPACE(ch))
+                bufptr[i] = ' ';
+        }
+        s = _PyUnicode_AsStringAndSize(s_buffer, &len);
+        if (s == NULL) {
+            Py_DECREF(s_buffer);
+            return NULL;
+        }
+        last = s + len;
     }
     else if (PyObject_AsCharBuffer(v, &s, &len)) {
         PyErr_SetString(PyExc_TypeError,
@@ -197,29 +205,27 @@ PyFloat_FromString(PyObject *v)
         return NULL;
     }
     last = s + len;
-
-    while (Py_ISSPACE(*s))
+    /* strip space */
+    while (s < last && Py_ISSPACE(*s))
         s++;
+    while (s < last - 1 && Py_ISSPACE(last[-1]))
+        last--;
     /* We don't care about overflow or underflow.  If the platform
      * supports them, infinities and signed zeroes (on underflow) are
      * fine. */
     x = PyOS_string_to_double(s, (char **)&end, NULL);
-    if (x == -1.0 && PyErr_Occurred())
-        goto error;
-    while (Py_ISSPACE(*end))
-        end++;
-    if (end == last)
-        result = PyFloat_FromDouble(x);
-    else {
-        PyOS_snprintf(buffer, sizeof(buffer),
-                      "invalid literal for float(): %.200s", s);
-        PyErr_SetString(PyExc_ValueError, buffer);
+    if (end != last) {
+        PyErr_Format(PyExc_ValueError,
+                     "could not convert string to float: "
+                     "%R", v);
         result = NULL;
     }
+    else if (x == -1.0 && PyErr_Occurred())
+        result = NULL;
+    else
+        result = PyFloat_FromDouble(x);
 
-  error:
-    if (s_buffer)
-        PyMem_FREE(s_buffer);
+    Py_XDECREF(s_buffer);
     return result;
 }
 
diff --git a/Objects/longobject.c b/Objects/longobject.c
index e8a7284..534e52d 100644
--- a/Objects/longobject.c
+++ b/Objects/longobject.c
@@ -2133,17 +2133,34 @@ PyObject *
 PyLong_FromUnicode(Py_UNICODE *u, Py_ssize_t length, int base)
 {
     PyObject *result;
-    char *buffer = (char *)PyMem_MALLOC(length+1);
+    PyObject *asciidig;
+    char *buffer, *end;
+    Py_ssize_t i, buflen;
+    Py_UNICODE *ptr;
 
-    if (buffer == NULL)
+    asciidig = PyUnicode_TransformDecimalToASCII(u, length);
+    if (asciidig == NULL)
         return NULL;
-
-    if (PyUnicode_EncodeDecimal(u, length, buffer, NULL)) {
-        PyMem_FREE(buffer);
+    /* Replace non-ASCII whitespace with ' ' */
+    ptr = PyUnicode_AS_UNICODE(asciidig);
+    for (i = 0; i < length; i++) {
+      Py_UNICODE ch = ptr[i];
+      if (ch > 127 && Py_UNICODE_ISSPACE(ch))
+        ptr[i] = ' ';
+    }
+    buffer = _PyUnicode_AsStringAndSize(asciidig, &buflen);
+    if (buffer == NULL) {
+        Py_DECREF(asciidig);
         return NULL;
     }
-    result = PyLong_FromString(buffer, NULL, base);
-    PyMem_FREE(buffer);
+    result = PyLong_FromString(buffer, &end, base);
+    if (result != NULL && end != buffer + buflen) {
+        PyErr_SetString(PyExc_ValueError,
+                        "null byte in argument for int()");
+        Py_DECREF(result);
+        result = NULL;
+    }
+    Py_DECREF(asciidig);
     return result;
 }
 
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index d3a2d1b..751da30 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -6206,6 +6206,30 @@ PyObject *PyUnicode_Translate(PyObject *str,
     return NULL;
 }
 
+PyObject *
+PyUnicode_TransformDecimalToASCII(Py_UNICODE *s,
+                                  Py_ssize_t length)
+{
+    PyObject *result;
+    Py_UNICODE *p; /* write pointer into result */
+    Py_ssize_t i;
+    /* Copy to a new string */
+    result = (PyObject *)_PyUnicode_New(length);
+    Py_UNICODE_COPY(PyUnicode_AS_UNICODE(result), s, length);
+    if (result == NULL)
+        return result;
+    p = PyUnicode_AS_UNICODE(result);
+    /* Iterate over code points */
+    for (i = 0; i < length; i++) {
+        Py_UNICODE ch =s[i];
+        if (ch > 127) {
+            int decimal = Py_UNICODE_TODECIMAL(ch);
+            if (decimal >= 0)
+                p[i] = '0' + decimal;
+        }
+    }
+    return result;
+}
 /* --- Decimal Encoder ---------------------------------------------------- */
 
 int PyUnicode_EncodeDecimal(Py_UNICODE *s,
@@ -8967,6 +8991,13 @@ unicode_freelistsize(PyUnicodeObject *self)
 {
     return PyLong_FromLong(numfree);
 }
+
+static PyObject *
+unicode__decimal2ascii(PyObject *self)
+{
+    return PyUnicode_TransformDecimalToASCII(PyUnicode_AS_UNICODE(self),
+                                             PyUnicode_GET_SIZE(self));
+}
 #endif
 
 PyDoc_STRVAR(startswith__doc__,
@@ -9108,7 +9139,6 @@ unicode_getnewargs(PyUnicodeObject *v)
     return Py_BuildValue("(u#)", v->str, v->length);
 }
 
-
 static PyMethodDef unicode_methods[] = {
 
     /* Order is according to common usage: often used methods should
@@ -9170,8 +9200,9 @@ static PyMethodDef unicode_methods[] = {
 #endif
 
 #if 0
-    /* This one is just used for debugging the implementation. */
+    /* These methods are just used for debugging the implementation. */
     {"freelistsize", (PyCFunction) unicode_freelistsize, METH_NOARGS},
+    {"_decimal2ascii", (PyCFunction) unicode__decimal2ascii, METH_NOARGS},
 #endif
 
     {"__getnewargs__",  (PyCFunction)unicode_getnewargs, METH_NOARGS},