Issue #5108: Handle %s like %S and %R in PyUnicode_FromFormatV(): Call

PyUnicode_DecodeUTF8() once, remember the result and output it in a second
step. This avoids problems with counting UTF-8 bytes that ignores the effect
of using the replace error handler in PyUnicode_DecodeUTF8().

diff --git a/Misc/NEWS b/Misc/NEWS
index 579dcb49aaa1d16a66580eaff3b4d57364736150..64e473ac58989425ac5e0ae7794ca14af3dfaf01 100644 (file)
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -258,6 +258,11 @@ Core and Builtins
 - Issue #5705: os.setuid() would not accept values > 2**31-1 but pwd.getpwnam()
   returned them on 64bit platforms.
 
+- Issue #5108: Handle %s like %S and %R in PyUnicode_FromFormatV(): Call
+  PyUnicode_DecodeUTF8() once, remember the result and output it in a second
+  step. This avoids problems with counting UTF-8 bytes that ignores the effect
+  of using the replace error handler in PyUnicode_DecodeUTF8().
+
 Library
 -------
 

diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 62191ade0ab815aefaf13c7d442b1f8bc76ccc7b..8e18302a24bc5767ff2e3d07f6ad50778d12a5a6 100644 (file)
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -674,15 +674,25 @@ PyUnicode_FromFormatV(const char *format, va_list vargs)
     count = vargs;
 #endif
 #endif
-    /* step 1: count the number of %S/%R format specifications
-     * (we call PyObject_Str()/PyObject_Repr() for these objects
-     * once during step 3 and put the result in an array) */
+     /* step 1: count the number of %S/%R/%s format specifications
+      * (we call PyObject_Str()/PyObject_Repr()/PyUnicode_DecodeUTF8() for these
+      * objects once during step 3 and put the result in an array) */
     for (f = format; *f; f++) {
-        if (*f == '%' && (*(f+1)=='S' || *(f+1)=='R'))
-            ++callcount;
+         if (*f == '%') {
+             if (*(f+1)=='%')
+                 continue;
+             if (*(f+1)=='S' || *(f+1)=='R' || *(f+1)=='A')
+                 ++callcount;
+             while (isdigit((unsigned)*f))
+                 width = (width*10) + *f++ - '0';
+             while (*++f && *f != '%' && !isalpha((unsigned)*f))
+                 ;
+             if (*f == 's')
+                 ++callcount;
+         }
     }
     /* step 2: allocate memory for the results of
-     * PyObject_Str()/PyObject_Repr() calls */
+     * PyObject_Str()/PyObject_Repr()/PyUnicode_DecodeUTF8() calls */
     if (callcount) {
         callresults = PyObject_Malloc(sizeof(PyObject *)*callcount);
         if (!callresults) {
@@ -731,35 +741,13 @@ PyUnicode_FromFormatV(const char *format, va_list vargs)
             case 's':
             {
                 /* UTF-8 */
-                unsigned char*s;
-                s = va_arg(count, unsigned char*);
-                while (*s) {
-                    if (*s < 128) {
-                        n++; s++;
-                    } else if (*s < 0xc0) {
-                        /* invalid UTF-8 */
-                        n++; s++;
-                    } else if (*s < 0xc0) {
-                        n++;
-                        s++; if(!*s)break;
-                        s++;
-                    } else if (*s < 0xe0) {
-                        n++;
-                        s++; if(!*s)break;
-                        s++; if(!*s)break;
-                        s++;
-                    } else {
-#ifdef Py_UNICODE_WIDE
-                        n++;
-#else
-                        n+=2;
-#endif
-                        s++; if(!*s)break;
-                        s++; if(!*s)break;
-                        s++; if(!*s)break;
-                        s++;
-                    }
-                }
+                unsigned char *s = va_arg(count, unsigned char*);
+                PyObject *str = PyUnicode_DecodeUTF8(s, strlen(s), "replace");
+                if (!str)
+                    goto fail;
+                n += PyUnicode_GET_SIZE(str);
+                /* Remember the str and switch to the next slot */
+                *callresult++ = str;
                 break;
             }
             case 'U':
@@ -915,19 +903,15 @@ PyUnicode_FromFormatV(const char *format, va_list vargs)
                 break;
             case 's':
             {
-                /* Parameter must be UTF-8 encoded.
-                   In case of encoding errors, use
-                   the replacement character. */
-                PyObject *u;
-                p = va_arg(vargs, char*);
-                u = PyUnicode_DecodeUTF8(p, strlen(p),
-                                         "replace");
-                if (!u)
-                    goto fail;
-                Py_UNICODE_COPY(s, PyUnicode_AS_UNICODE(u),
-                                PyUnicode_GET_SIZE(u));
-                s += PyUnicode_GET_SIZE(u);
-                Py_DECREF(u);
+                /* unused, since we already have the result */
+                (void) va_arg(vargs, char *);
+                Py_UNICODE_COPY(s, PyUnicode_AS_UNICODE(*callresult),
+                                PyUnicode_GET_SIZE(*callresult));
+                s += PyUnicode_GET_SIZE(*callresult);
+                /* We're done with the unicode()/repr() => forget it */
+                Py_DECREF(*callresult);
+                /* switch to next unicode()/repr() result */
+                ++callresult;
                 break;
             }
             case 'U':