needforspeed: stringlib refactoring, continued. added count and

find helpers; updated unicodeobject to use stringlib_count

5 files changed, 132 insertions, 53 deletions

diff --git a/Objects/stringlib/count.h b/Objects/stringlib/count.h
new file mode 100644
index 0000000..0036f63
--- /dev/null
+++ b/Objects/stringlib/count.h
@@ -0,0 +1,34 @@
+/* stringlib: count implementation */
+
+#ifndef STRINGLIB_COUNT_H
+#define STRINGLIB_COUNT_H
+
+#ifndef STRINGLIB_FASTSEARCH_H
+#error must include "stringlib/fastsearch.h" before including this module
+#endif
+
+Py_LOCAL(Py_ssize_t)
+stringlib_count(const STRINGLIB_CHAR* str, Py_ssize_t str_len,
+                const STRINGLIB_CHAR* sub, Py_ssize_t sub_len)
+{
+    Py_ssize_t count;
+
+    if (sub_len == 0)
+        return str_len + 1;
+
+    count = fastsearch(str, str_len, sub, sub_len, FAST_COUNT);
+
+    if (count < 0)
+        count = 0; /* no match */
+
+    return count;
+}
+
+#endif
+
+/*
+Local variables:
+c-basic-offset: 4
+indent-tabs-mode: nil
+End:
+*/
diff --git a/Objects/stringlib/find.h b/Objects/stringlib/find.h
new file mode 100644
index 0000000..c2399ad
--- /dev/null
+++ b/Objects/stringlib/find.h
@@ -0,0 +1,49 @@
+/* stringlib: find/index implementation */
+
+#ifndef STRINGLIB_FIND_H
+#define STRINGLIB_FIND_H
+
+#ifndef STRINGLIB_FASTSEARCH_H
+#error must include "stringlib/fastsearch.h" before including this module
+#endif
+
+Py_LOCAL(Py_ssize_t)
+stringlib_find(const STRINGLIB_CHAR* str, Py_ssize_t str_len,
+               const STRINGLIB_CHAR* sub, Py_ssize_t sub_len)
+{
+    if (sub_len == 0)
+        return 0;
+
+    return fastsearch(str, str_len, sub, sub_len, FAST_SEARCH);
+}
+
+Py_LOCAL(Py_ssize_t)
+stringlib_rfind(const STRINGLIB_CHAR* str, Py_ssize_t str_len,
+                const STRINGLIB_CHAR* sub, Py_ssize_t sub_len)
+{
+    Py_ssize_t pos;
+
+    /* XXX - create reversefastsearch helper! */
+    if (sub_len == 0)
+	pos = str_len;
+    else {
+	Py_ssize_t j;
+        pos = -1;
+	for (j = str_len - sub_len; j >= 0; --j)
+            if (STRINGLIB_CMP(str+j, sub, sub_len) == 0) {
+                pos = j;
+                break;
+            }
+    }
+
+    return pos;
+}
+
+#endif
+
+/*
+Local variables:
+c-basic-offset: 4
+indent-tabs-mode: nil
+End:
+*/
diff --git a/Objects/stringlib/partition.h b/Objects/stringlib/partition.h
index 71e80a9..8cc7abe 100644
--- a/Objects/stringlib/partition.h
+++ b/Objects/stringlib/partition.h
@@ -3,9 +3,15 @@
 #ifndef STRINGLIB_PARTITION_H
 #define STRINGLIB_PARTITION_H
 
+#ifndef STRINGLIB_FASTSEARCH_H
+#error must include "stringlib/fastsearch.h" before including this module
+#endif
+
 Py_LOCAL(PyObject*)
-partition(PyObject* str_obj, const STRINGLIB_CHAR* str, Py_ssize_t str_len,
-          PyObject* sep_obj, const STRINGLIB_CHAR* sep, Py_ssize_t sep_len)
+stringlib_partition(
+    PyObject* str_obj, const STRINGLIB_CHAR* str, Py_ssize_t str_len,
+    PyObject* sep_obj, const STRINGLIB_CHAR* sep, Py_ssize_t sep_len
+    )
 {
     PyObject* out;
     Py_ssize_t pos;
@@ -46,8 +52,10 @@ partition(PyObject* str_obj, const STRINGLIB_CHAR* str, Py_ssize_t str_len,
 }
 
 Py_LOCAL(PyObject*)
-rpartition(PyObject* str_obj, const STRINGLIB_CHAR* str, Py_ssize_t str_len,
-	   PyObject* sep_obj, const STRINGLIB_CHAR* sep, Py_ssize_t sep_len)
+stringlib_rpartition(
+    PyObject* str_obj, const STRINGLIB_CHAR* str, Py_ssize_t str_len,
+    PyObject* sep_obj, const STRINGLIB_CHAR* sep, Py_ssize_t sep_len
+    )
 {
     PyObject* out;
     Py_ssize_t pos;
diff --git a/Objects/stringobject.c b/Objects/stringobject.c
index 5d57f15..75325ab 100644
--- a/Objects/stringobject.c
+++ b/Objects/stringobject.c
@@ -1548,7 +1548,7 @@ string_partition(PyStringObject *self, PyObject *sep_obj)
 	else if (PyObject_AsCharBuffer(sep_obj, &sep, &sep_len))
 		return NULL;
 
-	return partition(
+	return stringlib_partition(
 		(PyObject*) self,
 		PyString_AS_STRING(self), PyString_GET_SIZE(self),
 		sep_obj, sep, sep_len
@@ -1579,7 +1579,7 @@ string_rpartition(PyStringObject *self, PyObject *sep_obj)
 	else if (PyObject_AsCharBuffer(sep_obj, &sep, &sep_len))
 		return NULL;
 
-	return rpartition(
+	return stringlib_rpartition(
 		(PyObject*) self,
 		PyString_AS_STRING(self), PyString_GET_SIZE(self),
 		sep_obj, sep, sep_len
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 7d644d3..596bd10 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -3869,63 +3869,47 @@ STRINGLIB_CMP(const Py_UNICODE* str, const Py_UNICODE* other, Py_ssize_t len)
 #define STRINGLIB_EMPTY unicode_empty
 
 #include "stringlib/fastsearch.h"
-#include "stringlib/partition.h"
 
+#include "stringlib/count.h"
+#include "stringlib/find.h"
+#include "stringlib/partition.h"
 
-Py_LOCAL(Py_ssize_t) count(PyUnicodeObject *self,
+Py_ssize_t PyUnicode_Count(PyObject *str,
+                           PyObject *substr,
                            Py_ssize_t start,
-                           Py_ssize_t end,
-                           PyUnicodeObject *substring)
+                           Py_ssize_t end)
 {
-    Py_ssize_t count = 0;
+    Py_ssize_t result;
+    PyUnicodeObject* str_obj;
+    PyUnicodeObject* sub_obj;
+
+    str_obj = (PyUnicodeObject*) PyUnicode_FromObject(str);
+    if (!str_obj)
+	return -1;
+    sub_obj = (PyUnicodeObject*) PyUnicode_FromObject(substr);
+    if (!sub_obj) {
+	Py_DECREF(str_obj);
+	return -1;
+    }
 
     if (start < 0)
-        start += self->length;
+        start += str_obj->length;
     if (start < 0)
         start = 0;
-    if (end > self->length)
-        end = self->length;
+    if (end > str_obj->length)
+        end = str_obj->length;
     if (end < 0)
-        end += self->length;
+        end += str_obj->length;
     if (end < 0)
         end = 0;
 
-    if (substring->length == 0)
-	return (end - start + 1);
-
-    count = fastsearch(
-        PyUnicode_AS_UNICODE(self) + start, end - start,
-        substring->str, substring->length, FAST_COUNT
+    result = stringlib_count(
+        str_obj->str + start, end - start, sub_obj->str, sub_obj->length
         );
 
-    if (count < 0)
-        count = 0; /* no match */
-
-    return count;
-}
-
-Py_ssize_t PyUnicode_Count(PyObject *str,
-		    PyObject *substr,
-		    Py_ssize_t start,
-		    Py_ssize_t end)
-{
-    Py_ssize_t result;
-
-    str = PyUnicode_FromObject(str);
-    if (str == NULL)
-	return -1;
-    substr = PyUnicode_FromObject(substr);
-    if (substr == NULL) {
-	Py_DECREF(str);
-	return -1;
-    }
-
-    result = count((PyUnicodeObject *)str,
-		   start, end,
-		   (PyUnicodeObject *)substr);
+    Py_DECREF(sub_obj);
+    Py_DECREF(str_obj);
 
-    Py_DECREF(str);
-    Py_DECREF(substr);
     return result;
 }
 
@@ -4767,7 +4751,7 @@ PyObject *replace(PyUnicodeObject *self,
         Py_UNICODE *p;
 
         /* replace strings */
-        n = count(self, 0, self->length, str1);
+        n = stringlib_count(self->str, self->length, str1->str, str1->length);
         if (n > maxcount)
             n = maxcount;
         if (n == 0)
@@ -5162,7 +5146,7 @@ unicode_count(PyUnicodeObject *self, PyObject *args)
         return NULL;
 
     substring = (PyUnicodeObject *)PyUnicode_FromObject(
-						(PyObject *)substring);
+        (PyObject *)substring);
     if (substring == NULL)
 	return NULL;
 
@@ -5177,9 +5161,13 @@ unicode_count(PyUnicodeObject *self, PyObject *args)
     if (end < 0)
         end = 0;
 
-    result = PyInt_FromSsize_t(count(self, start, end, substring));
+    result = PyInt_FromSsize_t(
+        stringlib_count(self->str + start, end - start,
+                        substring->str, substring->length)
+        );
 
     Py_DECREF(substring);
+
     return result;
 }
 
@@ -6222,7 +6210,7 @@ PyUnicode_Partition(PyObject *str_in, PyObject *sep_in)
         return NULL;
     }
 
-    out = partition(
+    out = stringlib_partition(
         str_obj, PyUnicode_AS_UNICODE(str_obj), PyUnicode_GET_SIZE(str_obj),
         sep_obj, PyUnicode_AS_UNICODE(sep_obj), PyUnicode_GET_SIZE(sep_obj)
         );
@@ -6250,7 +6238,7 @@ PyUnicode_RPartition(PyObject *str_in, PyObject *sep_in)
         return NULL;
     }
 
-    out = rpartition(
+    out = stringlib_rpartition(
         str_obj, PyUnicode_AS_UNICODE(str_obj), PyUnicode_GET_SIZE(str_obj),
         sep_obj, PyUnicode_AS_UNICODE(sep_obj), PyUnicode_GET_SIZE(sep_obj)
         );