Issue #28808: PyUnicode_CompareWithASCIIString() now never raises exceptions.

serhiy-storchaka · serhiy-storchaka · commit d0f9e955bce4 · 2016-12-06T00:20:26.000+02:00
diff --git a/Doc/c-api/unicode.rst b/Doc/c-api/unicode.rst
@@ -1657,8 +1657,7 @@ They all return *NULL* or ``-1`` if an exception occurs.
    ASCII-encoded strings, but the function interprets the input string as
    ISO-8859-1 if it contains non-ASCII characters.
 
-   This function returns ``-1`` upon failure, so one should call
-   :c:func:`PyErr_Occurred` to check for errors.
+   This function does not raise exceptions.
 
 
 .. c:function:: PyObject* PyUnicode_RichCompare(PyObject *left,  PyObject *right,  int op)
diff --git a/Include/unicodeobject.h b/Include/unicodeobject.h
@@ -2055,7 +2055,7 @@ PyAPI_FUNC(int) _PyUnicode_EqualToASCIIId(
    equal, and greater than, respectively.  It is best to pass only
    ASCII-encoded strings, but the function interprets the input string as
    ISO-8859-1 if it contains non-ASCII characters.
-   Raise an exception and return -1 on error. */
+   This function does not raise exceptions. */
 
 PyAPI_FUNC(int) PyUnicode_CompareWithASCIIString(
     PyObject *left,
diff --git a/Misc/NEWS b/Misc/NEWS
@@ -480,6 +480,8 @@ Windows
 C API
 -----
 
+- Issue #28808: PyUnicode_CompareWithASCIIString() now never raises exceptions.
+
 - Issue #28761: The fields name and doc of structures PyMemberDef, PyGetSetDef,
   PyStructSequence_Field, PyStructSequence_Desc, and wrapperbase are now of
   type ``const char *`` rather of ``char *``.
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
@@ -10977,10 +10977,24 @@ PyUnicode_CompareWithASCIIString(PyObject* uni, const char* str)
     Py_ssize_t i;
     int kind;
     Py_UCS4 chr;
+    const unsigned char *ustr = (const unsigned char *)str;
 
     assert(_PyUnicode_CHECK(uni));
-    if (PyUnicode_READY(uni) == -1)
-        return -1;
+    if (!PyUnicode_IS_READY(uni)) {
+        const wchar_t *ws = _PyUnicode_WSTR(uni);
+        /* Compare Unicode string and source character set string */
+        for (i = 0; (chr = ws[i]) && ustr[i]; i++) {
+            if (chr != ustr[i])
+                return (chr < ustr[i]) ? -1 : 1;
+        }
+        /* This check keeps Python strings that end in '\0' from comparing equal
+         to C strings identical up to that point. */
+        if (_PyUnicode_WSTR_LENGTH(uni) != i || chr)
+            return 1; /* uni is longer */
+        if (ustr[i])
+            return -1; /* str is longer */
+        return 0;
+    }
     kind = PyUnicode_KIND(uni);
     if (kind == PyUnicode_1BYTE_KIND) {
         const void *data = PyUnicode_1BYTE_DATA(uni);
