[Zope-Checkins] CVS: Zope/lib/python/Products/PluginIndexes/TextIndex/Splitter/UnicodeSplitter/src - UnicodeSplitter.c:1.2

[Andreas Jung]

Update of /cvs-repository/Zope/lib/python/Products/PluginIndexes/TextIndex/Splitter/UnicodeSplitter/src
In directory cvs.zope.org:/tmp/cvs-serv25855/src

Modified Files:
	UnicodeSplitter.c 
Log Message:
code cleanup


=== Zope/lib/python/Products/PluginIndexes/TextIndex/Splitter/UnicodeSplitter/src/UnicodeSplitter.c 1.1 => 1.2 ===
         } else res = word;
 
-    return res;
+        return res;
 }
 
 static void
@@ -67,15 +67,9 @@
                 return NULL;
         }
 
-        item=PyList_GetItem(self->list , i);
+        item=PyList_GET_ITEM(self->list , i);
         Py_INCREF(item);
 
-
-#ifdef DEBUG
-        printf("\n\tItem %d",i);
-        PyObject_Print(item,stdout,0);
-        fflush(stdout);
-#endif
         return item;
 }
 
@@ -90,10 +84,10 @@
         if (! (r=PyList_New(0))) return NULL;
 
         for (i=0;i<PyList_Size(self->list);i++) {
-                item=PyList_GetItem(self->list,i);
+                item=PyList_GET_ITEM(self->list,i);
 
                 if (PyUnicode_Compare(word,item)==0) {
-                        index=PyInt_FromLong(i); 
+                        index=PyInt_FromLong(i);
                         if(!index) return NULL;
                         Py_INCREF(item);
                         PyList_Append(r,index);
@@ -194,8 +188,6 @@
                 register Py_UNICODE ch;
 
                 ch = *s;
-                *s = Py_UNICODE_TOLOWER(ch);
-
 #ifdef DEBUG
                 printf("%d %c %d\n",i,ch,ch);
                 fflush(stdout);
@@ -256,6 +248,39 @@
 
 }
 
+
+static
+void fixlower(PyUnicodeObject *self)
+{
+        int len = self->length;
+        Py_UNICODE *s = self->str;
+
+        while (len-- > 0) {
+                register Py_UNICODE ch;
+
+                ch = Py_UNICODE_TOLOWER(*s);
+                if (ch != *s) *s = ch;
+                s++;
+        }
+}
+
+
+static
+PyUnicodeObject *prepareString(PyUnicodeObject *o)
+
+{
+        PyUnicodeObject *u;
+
+        u = (PyUnicodeObject*) PyUnicode_FromUnicode(NULL, o->length);
+        if (u == NULL) return NULL;
+
+        Py_UNICODE_COPY(u->str, o->str, o->length);
+        fixlower(u);
+
+        return  u;
+}
+
+
 static PyObject *
 get_Splitter(PyObject *modinfo, PyObject *args,PyObject *keywds)
 {
@@ -274,7 +299,7 @@
         if (PyString_Check(doc)) {
                 // This sux a bit. The default encoding should be ascii or latin1.
                 // But there must be better support to pass an optional encoding parameter
-                
+
                 unicodedoc = PyUnicode_FromEncodedObject(doc,"latin1","strict");
                 if (! unicodedoc) goto err;
 
@@ -291,7 +316,7 @@
                 Py_INCREF(synstop);
         } else  self->synstop=NULL;
 
-        splitUnicodeString(self,(PyUnicodeObject *)unicodedoc);
+        splitUnicodeString(self,prepareString((PyUnicodeObject *) unicodedoc));
 
         return (PyObject*)self;