C implementation of functools.lru_cache

Modules/_functoolsmodule.c

 
 #include "Python.h"
 #include "structmember.h"
 
 /* _functools module written and maintained
    by Hye-Shik Chang <perky@FreeBSD.org>
    with adaptations by Raymond Hettinger <python@rcn.com>
    Copyright (c) 2004, 2005, 2006 Python Software Foundation.
    All rights reserved.
 */
@@ skipping 522 matching lines @@
 PyDoc_STRVAR(functools_reduce_doc,
 "reduce(function, sequence[, initial]) -> value\n\
 \n\
 Apply a function of two arguments cumulatively to the items of a sequence,\n\
 from left to right, so as to reduce the sequence to a single value.\n\
 For example, reduce(lambda x, y: x+y, [1, 2, 3, 4, 5]) calculates\n\
 ((((1+2)+3)+4)+5).  If initial is present, it is placed before the items\n\
 of the sequence in the calculation, and serves as a default when the\n\
 sequence is empty.");
 
+/* lru_cache object **********************************************************/
+
+/* this object is used delimit args and keywords in the cache keys */
+static PyObject *kwd_mark;
+
+typedef struct lru_list_elem lru_list_elem;
+
+typedef struct lru_list_elem {
+    lru_list_elem *prev, *next;
+    PyObject *key, *result;
+} lru_list_elem;
+
+typedef struct lru_cache_object lru_cache_object;
+
+typedef PyObject *(*lru_cache_ternaryfunc)(lru_cache_object *, PyObject *, PyObject *);
+
+typedef struct lru_cache_object {
+    PyObject_HEAD
+    Py_ssize_t maxsize;
+    PyObject *maxsize_O;
+    PyObject *func;
+    lru_cache_ternaryfunc wrapper;
+    PyObject *cache;
+    Py_ssize_t misses, hits;
+    lru_list_elem root;
+    int typed;
+    PyObject *dict;
+} lru_cache_object;
+
+static PyTypeObject lru_cache_type;
+
+static PyObject *
+lru_cache_make_key(PyObject *args, PyObject *kwds, int typed)
+{
+    PyObject *key, *sorted_items;
+    Py_ssize_t key_size, pos, key_pos;
+
+    /* short path, key will match args anyway, which is a tuple */
+    if (!typed && !kwds) {
+        Py_INCREF(args);
+        return args;
+    }
+
+    if (kwds) {
+        assert(PyDict_Size(kwds));
+        if (!(sorted_items = PyDict_Items(kwds)))
+            return NULL;
+        if (0 > PyList_Sort(sorted_items)) {
+            Py_DECREF(sorted_items);
+            return NULL;
+        }
+    } else
+        sorted_items = NULL;
+
+    key_size = PyTuple_GET_SIZE(args);
+    if (kwds)
+        key_size += PyList_GET_SIZE(sorted_items);
+    if (typed)
+        key_size *= 2;
+    if (kwds)
+        key_size++;
+
+    key = PyTuple_New(key_size);
+    key_pos = 0;
+
+    for (pos = 0; pos < PyTuple_GET_SIZE(args); ++pos) {
+        PyObject *item = PyTuple_GET_ITEM(args, pos);
+        Py_INCREF(item);
+        PyTuple_SET_ITEM(key, key_pos++, item);
+    }
+    if (kwds) {
+        Py_INCREF(kwd_mark);
+        PyTuple_SET_ITEM(key, key_pos++, kwd_mark);
+        for (pos = 0; pos < PyList_GET_SIZE(sorted_items); ++pos) {
+            PyObject *item = PyList_GET_ITEM(sorted_items, pos);
+            Py_INCREF(item);
+            PyTuple_SET_ITEM(key, key_pos++, item);
+        }
+    }
+    if (typed) {
+        for (pos = 0; pos < PyTuple_GET_SIZE(args); ++pos) {
+            PyObject *item = (PyObject *)Py_TYPE(PyTuple_GET_ITEM(args, pos));
+            Py_INCREF(item);
+            PyTuple_SET_ITEM(key, key_pos++, item);
+        }
+        if (kwds) {
+            for (pos = 0; pos < PyList_GET_SIZE(sorted_items); ++pos) {
+                PyObject *item = (PyObject *)Py_TYPE(PyTuple_GET_ITEM(PyList_GET_ITEM(sorted_items, pos), 1));
+                Py_INCREF(item);
+                PyTuple_SET_ITEM(key, key_pos++, item);
+            }
+        }
+    }
+    assert(key_pos == key_size);
+    if (kwds)
+        Py_DECREF(sorted_items);
+    return key;
+}
+
+static PyObject *
+uncached_lru_cache_wrapper(lru_cache_object *self, PyObject *args, PyObject *kwds)
+{
+    PyObject *result = PyObject_Call(self->func, args, kwds);
+    if (!result)
+        return NULL;
+    self->misses++;
+    return result;
+}
+
+static PyObject *
+infinite_lru_cache_wrapper(lru_cache_object *self, PyObject *args, PyObject *kwds)
+{
+    PyObject *result;
+    PyObject *key = lru_cache_make_key(args, kwds, self->typed);
+    if (!key)
+        return NULL;
+    result = PyDict_GetItemWithError(self->cache, key);
+    if (result) {
+        Py_INCREF(result);
+        self->hits++;
+        Py_DECREF(key);
+        return result;
+    }
+    if (PyErr_Occurred()) {
+        Py_DECREF(key);
+        return NULL;
+    }
+    result = PyObject_Call(self->func, args, kwds);
+    if (!result) {
+        Py_DECREF(key);
+        return NULL;
+    }
+    if (PyDict_SetItem(self->cache, key, result) < 0) {
+        Py_DECREF(result);
+        Py_DECREF(key);
+        return NULL;
+    }
+    Py_DECREF(key);
+    self->misses++;
+    return result;
+}
+
+static void
+lru_cache_list_extricate(lru_list_elem *link)
+{
+    link->prev->next = link->next;
+    link->next->prev = link->prev;
+}
+
+static void
+lru_cache_list_append(lru_list_elem *root, lru_list_elem *link)
+{
+    lru_list_elem *last = root->prev;
+    last->next = root->prev = link;
+    link->prev = last;
+    link->next = root;
+}
+
+static PyObject *
+bounded_lru_cache_wrapper(lru_cache_object *self, PyObject *args, PyObject *kwds)
+{
+    PyObject *key = lru_cache_make_key(args, kwds, self->typed);
+    if (!key)
+        return NULL;
+    PyObject *value = PyDict_GetItemWithError(self->cache, key);
+    if (value) {
+        lru_list_elem *link = PyCapsule_GetPointer(value, NULL);
+        lru_cache_list_extricate(link);
+        lru_cache_list_append(&self->root, link);
+        self->hits++;
+        Py_DECREF(key);
+        Py_INCREF(link->result);
+        return link->result;
+    }
+    if (PyErr_Occurred()) {
+        Py_DECREF(key);
+        return NULL;
+    }
+    PyObject *result = PyObject_Call(self->func, args, kwds);
+    if (!result) {
+        Py_DECREF(key);
+        return NULL;
+    }
+    lru_list_elem *link;
+    if (PyDict_Size(self->cache) == self->maxsize) {
+        /* extricate the oldest item */
+        link = self->root.next;
+        lru_cache_list_extricate(link);
+        /* grab its capsule */
+        value = PyDict_GetItem(self->cache, link->key);
+        Py_INCREF(value);
+        /* remove its key from the cache */
+        if (0 > PyDict_DelItem(self->cache, link->key))
+            abort();
+        /* scrub the result from the link */
+        Py_DECREF(link->result);
+    } else {
+        link = PyMem_New(lru_list_elem, 1);
+        value = PyCapsule_New(link, NULL, NULL);
+    }
+    lru_cache_list_append(&self->root, link);
+    link->key = key;
+    link->result = result;
+    Py_INCREF(result);
+    if (0 > PyDict_SetItem(self->cache, key, value)) abort();
+    Py_DECREF(key);
+    Py_DECREF(value);
+    self->misses++;
+    return result;
+}
+
+static PyObject *
+lru_cache_new(PyTypeObject *type, PyObject *args, PyObject *kw)
+{
+    PyObject *func;
+    PyObject *maxsize_O = NULL;
+    PyObject *typed_O = NULL;
+    int typed;
+    lru_cache_object *obj;
+    Py_ssize_t maxsize = 100;
+    PyObject *(*wrapper)(lru_cache_object *, PyObject *, PyObject *);
+    static char *keywords[] = {"func", "maxsize", "typed", NULL};
+    
+    if (!PyArg_ParseTupleAndKeywords(args, kw, "O|OO:lru_cache", keywords,
+                                     &func, &maxsize_O, &typed_O)) {
+        return NULL;
+    }
+
+    if (!PyCallable_Check(func)) {
+        PyErr_SetString(PyExc_TypeError,
+                        "the first argument must be callable");
+        return NULL;
+    }
+
+    if (maxsize_O == NULL) {
+        maxsize_O = PyLong_FromSsize_t(maxsize);
+        wrapper = bounded_lru_cache_wrapper;
+    } else if (maxsize_O == Py_None) {
+        wrapper = infinite_lru_cache_wrapper;
+        Py_INCREF(maxsize_O);
+    } else if (PyNumber_Check(maxsize_O)) {
+        maxsize = PyNumber_AsSsize_t(maxsize_O, PyExc_OverflowError);
+        if (maxsize == -1 && PyErr_Occurred())
+            return NULL;
+        if (maxsize == 0)
+            wrapper = uncached_lru_cache_wrapper;
+        else
+            wrapper = bounded_lru_cache_wrapper;
+        Py_INCREF(maxsize_O);
+    } else {
+        PyErr_SetString(PyExc_TypeError, "maxsize should be integer or None");
+        return NULL;
+    }
+
+    if (typed_O) {
+        int err = PyObject_IsTrue(typed_O);
+        if (err < 0) {
+            Py_DECREF(maxsize_O);
+            return NULL;
+        }
+        typed = err;
+    } else
+        typed = 0;
+
+    obj = (lru_cache_object *)type->tp_alloc(type, 0);
+    if (obj == NULL) {
+        Py_DECREF(maxsize_O);
+        return NULL;
+    }
+
+    obj->root.prev = &obj->root;
+    obj->root.next = &obj->root;
+    obj->maxsize = maxsize;
+    obj->maxsize_O = maxsize_O;
+    if (!(obj->cache = PyDict_New())) {
+        Py_DECREF(obj);
+        Py_DECREF(maxsize_O);
+        return NULL;
+    }
+    obj->func = func;
+    Py_INCREF(func);
+    obj->wrapper = wrapper;
+    obj->misses = obj->hits = 0;
+    obj->typed = typed;
+
+    return (PyObject *)obj;
+}
+
+static void
+lru_cache_clear_list(lru_list_elem *root)
+{
+    lru_list_elem *link = root->next;
+    while (link != root) {
+        lru_list_elem *next = link->next;
+        Py_DECREF(link->result);
+        PyMem_Free(link);
+        link = next;
+    }
+}
+
+static void
+lru_cache_dealloc(lru_cache_object *obj)
+{
+    Py_XDECREF(obj->maxsize_O);
+    Py_XDECREF(obj->func);
+    Py_XDECREF(obj->cache);
+    Py_XDECREF(obj->dict);
+    lru_cache_clear_list(&obj->root);
+    Py_TYPE(obj)->tp_free(obj);
+}
+
+static PyObject *
+lru_cache_call(lru_cache_object *self, PyObject *args, PyObject *kwds)
+{
+    return self->wrapper(self, args, kwds);
+}
+
+static PyObject *
+lru_cache_cache_info(lru_cache_object *self, PyObject *unused)
+{
+    PyObject *ret, *functools = PyImport_ImportModuleNoBlock("functools");
+    if (functools == NULL)
+        return NULL;
+    ret = PyObject_CallMethod(functools, "_CacheInfo", "nnOn",
+                              self->hits, self->misses, self->maxsize_O,
+                              PyDict_Size(self->cache));
+    Py_DECREF(functools);
+    return ret;
+}
+
+static PyObject *
+lru_cache_cache_clear(lru_cache_object *self, PyObject *unused)
+{
+    PyDict_Clear(self->cache);
+    self->hits = self->misses = 0;
+    lru_cache_clear_list(&self->root);
+    self->root.next = self->root.prev = &self->root;
+    Py_RETURN_NONE;
+}
+
+static PyMethodDef lru_cache_methods[] = {
+    {"cache_info", (PyCFunction)lru_cache_cache_info, METH_NOARGS},
+    {"cache_clear", (PyCFunction)lru_cache_cache_clear, METH_NOARGS},
+    {NULL}
+};
+
+static PyGetSetDef lru_cache_getsetlist[] = {
+    {"__dict__", PyObject_GenericGetDict, PyObject_GenericSetDict},
+    {NULL}
+};
+
+static PyTypeObject lru_cache_type = {
+    PyVarObject_HEAD_INIT(NULL, 0)
+    "functools._lru_cache",             /* tp_name */
+    sizeof(lru_cache_object),           /* tp_basicsize */
+    0,                                  /* tp_itemsize */
+    /* methods */
+    (destructor)lru_cache_dealloc,      /* tp_dealloc */
+    0,                                  /* tp_print */
+    0,                                  /* tp_getattr */
+    0,                                  /* tp_setattr */
+    0,                                  /* tp_reserved */
+    0,                                  /* tp_repr */
+    0,                                  /* tp_as_number */
+    0,                                  /* tp_as_sequence */
+    0,                                  /* tp_as_mapping */
+    0,                                  /* tp_hash */
+    (ternaryfunc)lru_cache_call,        /* tp_call */
+    0,                                  /* tp_str */
+    0,                                  /* tp_getattro */
+    0,                                  /* tp_setattro */
+    0,                                  /* tp_as_buffer */
+    Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
+                                        /* tp_flags */
+    0,                                  /* tp_doc */
+    0,                                  /* tp_traverse */
+    0,                                  /* tp_clear */
+    0,                                  /* tp_richcompare */
+    0,                                  /* tp_weaklistoffset */
+    0,                                  /* tp_iter */
+    0,                                  /* tp_iternext */
+    lru_cache_methods,                  /* tp_methods */
+    0,                                  /* tp_members */
+    lru_cache_getsetlist,               /* tp_getset */
+    0,                                  /* tp_base */
+    0,                                  /* tp_dict */
+    0,                                  /* tp_descr_get */
+    0,                                  /* tp_descr_set */
+    offsetof(lru_cache_object, dict),   /* tp_dictoffset */
+    0,                                  /* tp_init */
+    0,                                  /* tp_alloc */
+    lru_cache_new,                      /* tp_new */
+};
+
 /* module level code ********************************************************/
 
 PyDoc_STRVAR(module_doc,
 "Tools that operate on functions.");
 
 static PyMethodDef module_methods[] = {
     {"reduce",          functools_reduce,     METH_VARARGS, functools_reduce_doc},
     {"cmp_to_key",      (PyCFunction)functools_cmp_to_key,
      METH_VARARGS | METH_KEYWORDS, functools_cmp_to_key_doc},
     {NULL,              NULL}           /* sentinel */
 };
 
+static void
+module_free(void *m)
+{
+    Py_DECREF(kwd_mark);
+}
 
 static struct PyModuleDef _functoolsmodule = {
     PyModuleDef_HEAD_INIT,
     "_functools",
     module_doc,
     -1,
     module_methods,
     NULL,
     NULL,
     NULL,
-    NULL
+    module_free,
 };
 
 PyMODINIT_FUNC
 PyInit__functools(void)
 {
     int i;
     PyObject *m;
     char *name;
     PyTypeObject *typelist[] = {
         &partial_type,
+        &lru_cache_type,
         NULL
     };
 
     m = PyModule_Create(&_functoolsmodule);
     if (m == NULL)
         return NULL;
+
+    if (!(kwd_mark = PyObject_CallObject((PyObject *)&PyBaseObject_Type, NULL))) {
+        Py_DECREF(m);
+        return NULL;
+    }
 
     for (i=0 ; typelist[i] != NULL ; i++) {
         if (PyType_Ready(typelist[i]) < 0) {
             Py_DECREF(m);
             return NULL;
         }
         name = strchr(typelist[i]->tp_name, '.');
         assert (name != NULL);
         Py_INCREF(typelist[i]);
         PyModule_AddObject(m, name+1, (PyObject *)typelist[i]);
     }
     return m;
 }